Citation
Task specific neural correlates of the cued stroop interference effect

Material Information

Title:
Task specific neural correlates of the cued stroop interference effect an MEG study
Creator:
Monahan, Colleen Bridget ( author )
Language:
English
Physical Description:
1 electronic file : ;

Subjects

Subjects / Keywords:
Learning, Psychology of ( lcsh )
Attention ( lcsh )
Neural networks (Neurobiology) ( lcsh )
Magnetoencephalography ( lcsh )
Attention ( fast )
Learning, Psychology of ( fast )
Magnetoencephalography ( fast )
Neural networks (Neurobiology) ( fast )
Genre:
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )

Notes

Review:
The neural correlates associated with the Stroop interference effect remain a topic of debate despite being the subject of multiple neuroimaging studies over the past three decades. This study aimed to spatially and temporally localize the neural networks behind the Stroop interference effect in a cued, task switching Stroop task where incongruent trials were compared to congruent or neutral trials in the context of the specific task performed (color-naming or word-reading). This allowed a look at the influence that task-type and conflict type had on the pattern of activation when interference was occurring and whether interference is resolved via a top-down biasing method. Comparisons of source estimations revealed that certain occipital and parietal regions were recruited in all four variations of conflict while activation in the inferior temporal and prefrontal cortices was limited to the color-naming trials. In comparing conflict types (incongruent>neutral versus incongruent>congruent), we see distinct differences in activation within the prefrontal and occipital cortices, but no distinct differences in the temporal or parietal regions. The color-naming, incongruent>neutral condition behaviorally displayed the strongest instance of conflict and had the most significant, sustained prefrontal activation of the four contrasts. The neural pattern of activation observed in this contrast provides evidence for a top-down biasing model in times of high conflict and attentionial demands. Each of the four contrasts examined in this study has its own distinct temporal and spatial pattern of neural activation. This supports a model of selective attentional and conflict resolution that recruits distinct, but parallel neural networks that are both task and conflict condition specific.
Bibliography:
Includes bibliographical references.
System Details:
System requirements: Adobe Reader.
Statement of Responsibility:
by Colleen Bridget Monahan.

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
982959921 ( OCLC )
ocn982959921
Classification:
LD1193.E56 2016m M66 ( lcc )

Downloads

This item has the following downloads:


Full Text
TASK SPECIFIC NEURAL CORRELATES OF THE CUED STROOP INTERFERENCE
EFFECT: AN MEG STUDY by
COLLEEN BRIDGET MONAHAN B.S., Loyola University Maryland, 2011 M.Ed., University of Missouri St. Louis, 2013
A thesis submitted to the Faculty of the Graduate School of the University of Colorado in partial fulfillment of the requirements for the degree of Master of Science Bioengineering Program
2016


This thesis for the Master of Science degree by Colleen Bridget Monahan has been approved for the Bioengineering Program by
Richard Weir, Chair Benzi Kluger, Advisor
Robin Shandas


Monahan, Colleen Bridget (M.S., Bioengineering)
Task Specific Neural Correlates Of The Cued Stroop Interference Effect: An MEG Study Thesis directed by Associate Professor Benzi Kluger
ABSTRACT
The neural correlates associated with the Stroop interference effect remain a topic of debate despite being the subject of multiple neuroimaging studies over the past three decades. This study aimed to spatially and temporally localize the neural networks behind the Stroop interference effect in a cued, task switching Stroop task where incongruent trials were compared to congruent or neutral trials in the context of the specific task performed (colornaming or word-reading). This allowed a look at the influence that task-type and conflict type had on the pattern of activation when interference was occurring and whether interference is resolved via a top-down biasing method. Comparisons of source estimations revealed that certain occipital and parietal regions were recruited in all four variations of conflict while activation in the inferior temporal and prefrontal cortices was limited to the color-naming trials. In comparing conflict types (incongruent>neutral versus incongruent>congruent), we see distinct differences in activation within the prefrontal and occipital cortices, but no distinct differences in the temporal or parietal regions. The colornaming, incongruent>neutral condition behaviorally displayed the strongest instance of conflict and had the most significant, sustained prefrontal activation of the four contrasts.
The neural pattern of activation observed in this contrast provides evidence for a top-down biasing model in times of high conflict and attentionial demands. Each of the four contrasts examined in this study has its own distinct temporal and spatial pattern of neural activation.
111


This supports a model of selective attentional and conflict resolution that recruits distinct, but parallel neural networks that are both task and conflict condition specific.
The form and content of this abstract are approved. I recommend its publication.
Approved: Benzi Kluger
IV


ACKNOWLEDGEMENTS
This material is based upon work supported by, or in part by, the U.S. Army Research Laboratory and the U. S. Army Research Office under contract number W91 INF-10-1-0192. I would also like to acknowledge COMIRB for approval of this study under number 10-0539.
v


TABLE OF CONTENTS
CHAPTER
I. BACKGROUND AND SIGNIFICANCE...................................1
II. REVIEW OF EXISITING LITERATURE................................4
Introduction..................................................4
Neuroimaging Findings.........................................7
(MRI and PET............................................8
EEG....................................................10
MEG....................................................11
Gaps in Existing Research....................................13
Conclusion...................................................14
III. SPECIFIC AIMS AND HYPOTHESES.................................15
IV. MATERIALS AND METHODS........................................17
Ethics Statement.............................................17
Participants.................................................17
Study Design: the Computerized Stroop Task...................17
Magnetoencephalography (MEG) Recording.......................18
MEG Pre-Processing...........................................19
Source Localization Estimation...............................20
V. RESULTS......................................................22
Behavioral Results...........................................22
Verification of the Stroop Interference Effect with Reaction Time.. .22
Verification of the Stroop Interference Effect with Response Accuracy............................................................23
vi


Source Space Analysis Results
24
Color-Naming: Incongruent>Neutral...........................24
C ol or-Naming: Incongruent>C ongruent......................25
Word-Reading: Incongruent>Neutral...........................25
Word-Reading: Incongruent>Congruent.........................26
VI. DISCUSSION.........................................................35
Behavioral Implications............................................35
Task-Specific Prefrontal Activation................................36
Task-Specific Temporal Lobe Activation.............................39
Conflict Specific Activation.......................................41
Common Posterior Activations.......................................42
Comparison to EEG Findings.........................................43
Conclusion.........................................................44
Future Work........................................................45
\ II. AUTHOR CONTRIBUTIONS...............................................46
Affirmation of Original Work.......................................46
BIBLIOGRAPHY....................................................................47
APPENDIX........................................................................54
A. Steps Taken within Brainstorm and Matlab..............................54
A.l. Pre-Processing (for Each Subject).............................54
A.2. Source Estimation (for Each Condition within Each Subject) and Statistical Tests..................................................55
A.3. Statistical Tests Workflow....................................56
vii


B. Custom Matlab Scripts..........................................................57
B.l. Calculate Reaction Time...............................................57
B.2. Calculate Center of Weight and Mean t Values..........................74
viii


CHAPTER I
BACKGROUND AND SIGNIFICANCE
Humans exhibit enormous flexibility in our ability to react and adapt to the changing environment of everyday life. We are able to take in vast amounts of visual information at any point in time and selectively place our attention on whatever information is relevant to what we are doing. This is especially important when the object of focus has attributes that contribute conflicting information. Our brains need to identify and process the conflict in order to make sense of it. For the past three decades, researchers have been trying to localize the areas of our brains that work to detect and resolve visual conflict, as well as determine the chronological pattern of activation (Larson, Clayson, & Clawson, 2014; Nee, Wager, & Jonides, 2007). Research in this area is important because people with cognitive dysfunction from dementia or brain injury can experience decline in these executive functions. In understanding the neural correlates behind these functions, researchers can begin to find solutions to delay or ameliorate the adverse effects associated with a decline in executive functioning.
In order to better understand the neural processes associated with cognitive flexibility and conflict monitoring, researchers use the Stroop task by contrasting conditions of high conflict (incongruent stimuli) with conditions of low conflict (congruent or neutral stimuli) (MacLeod & MacDonald, 2000). However, inconsistencies exist among the research with regard to where these neural processes are localized and which neural processes are connected to which behaviors (Nee et al., 2007). These inconsistences most likely result from a number of methodological differences between studies that include how the neural signals were measured, what tasks were used, and which behaviors were examined.
1


The ability to detect and resolve the conflict behind the Stroop effect is thought to occur via a top-down biasing between the prefrontal cortex and task-relevant posterior regions of the brain (Banich et al., 2000b; Cohen, Dunbar, & McClelland, 1990; Herd, Banich, & OReilly, 2006). Yet current measures of neural activity are either not spatially or not temporally resolved enough to provide meaningful results on the chronological patterns of activation following the onset of a conflicting, incongruent visual stimulus.
The current measures of neural activity used to identify and localize the neural signals behind the Stroop effect are electroencephalograph (EEG), magnetoencephalograph (MEG), and functional MRI (fMRI). Results from fMRI have high spatial resolution but low temporal resolution when compared to EEG and MEG. Both MEG and EEG have temporal resolution on the millisecond time scale but MEG has the advantage of being more spatially resolved than EEG (Hamalainen, 1992). Thus, of these three measures, MEG emerges as the greatest potential measure to provide localized, chronological data about brain activity associated with the Stroop effect.
Multiple brain areas have been implicated with the Stroop interference effect using fMRI results (Nee et al., 2007). However, these results tell us nothing about the order in which areas of the brain activate in the short time frame between the stimulus and response of the Stroop task. Thus, fMRI studies regarding the Stroop effect, of which there are dozens, cannot definitively distinguish if conflict is being resolved in a town-down or bottom up manner. Contrary to fMRI, EEG studies provide information about time periods of activation but cannot localize the activation beyond general regions of the brain (i.e. centro-frontal, temporo-parietal). There is a gap in the current research on the Stroop interference
2


effect regarding whether the Stroop effect is actually resolved via top-down biasing due to the lack of highly spatially, temporally resolved results. MEG studies can help fill that gap.
The goal of this study is to contribute to the understanding of the neural correlates behind executive functions such as selective attention and conflict resolution by examining task-specific neural processes measured using MEG in a cued, task switching computerized version of the Stroop task. The results of this research will add clarity to the complex processes that the brain undergoes while a person selectively attends to a visually conflicting stimulus. By using MEG, this research will contribute temporal patterns of activation with high spatial localization in a way that has yet to be seen in the fMRI and EEG literature. If successful, the results of this research could incrementally alter how clinical treatment is applied to people who are experiencing cognitive decline associated aging, illness, or brain injury.
3


CHAPTER II
REVIEW OF EXISTING LITERATURE Introduction
Over the past few decades, many researchers in the field of cognitive neuroscience have focused their research on determining the underlying neural mechanisms behind the Stroop task (Cohen, Dunbar, & McClelland, 1990; MacLeod, 1991; Minzenberg, Laird, Thelen, Carter, & Glahn, 2009; Nee, Wager, & Jonides, 2007). To complete the Stroop task, typically, subjects are instructed to name the ink color of a list of color words, regardless of the meaning of the word. When the meaning of the word is incongruent with the ink color (e.g. RED written in blue ink), a phenomenon called the Stroop interference effect occurs (Stroop, 1935). The Stroop effect is the decrease in response time or increase in error rate when subjects are asked to name the ink color of incongruent stimuli as compared with congruent (e.g. RED written in red) or neutral stimuli (e.g. CAR written in red) (MacLeod & MacDonald, 2000). Over the years, the Stroop task has been modified and used extensively as a means to test cognitive functioning including mechanisms of control, conflict monitoring, selective attention, and executive functioning (Cohen et al., 1990; MacLeod & MacDonald, 2000). While this research contributes substantially to the overall understanding of the mechanisms behind cognitive control and attention, the use of different neuroimaging techniques and various versions of the Stroop task has led to mixed results on the localization and functionality of these neural processes (Nee et al., 2007).
The goal of this study is to contribute to the understanding of the neural correlates behind the Stroop interference effect by examining task-specific neural processes measured in a cued, task switching computerized version of the Stroop task. By using a cued, task
4


switching version of the Stroop task, a new task set is defined at the beginning of each trial (MacDonald, Cohen, Stenger, & Carter, 2000; West, 2003). For each trial, subjects are initially presented with an instructional cue (WORD OR COLOR) that indicates what attribute of the stimulus upon which to focus their selective attention. The cue is then followed by an incongruent, congruent, or neutral visual stimulus and subjects respond according to the cued instructions. In the original version of the Stroop task, where subjects perform the task of word reading or color naming for the whole block without switching, the interference effect is only observed in the color naming task (MacLeod & MacDonald,
2000). Because word reading is a much more of an automatic practice when presented with the stimulus, the interference effect is not present during that task (Cohen et al., 1990). In the cued Stroop task, however, the interference effect is seen in both color naming and word reading tasks as measured by decreased reaction time during incongruent conditions as compared to the congruent or neutral conditions (MacDonald et al., 2000; Perlstein, Larson, Dotson, & Kelly, 2006; West, 2003). By having to update the task set on a trial by trial basis, the behavioral results associated with automaticity of word reading are somewhat suspended. In this way, using the cued, task switching Stroop task allows the researcher to test the influence of the task type on the interference task.
When presented with the stimulus in the Stroop task, the subject must attend to one attribute of the stimulus over the other. This creates conflict when the different attributes of the stimuli are incongruent and when tasked with identifying the less automatic attribute (MacLeod & MacDonald, 2000). It has been posited that there are two distinct attentional systems within the brain that are recruited in the conflict processing and resolution of the Stroop task through top down processing (Banich et al., 2000b; Fan, Flombaum, McCandliss,
5


Thomas, & Posner, 2003; Miller & Cohen, 2001). The executive attentional functioning system distinguishes the aspect of the stimuli that is task-relevant and requiring attention.
This executive functioning is found in the anterior part of the brain but communicates closely with the regulatory system of attention located in the posterior of the brain. These anterior, prefrontal regions impose the attentional set by alerting or priming the posterior brain areas associated with processing the task relevant information about the stimulus (Banich et al., 2000b). Here, brain areas are activated based on the perceived stimulus attributes. In trials where color naming is the task, brain areas associated with color processing are more active (Corbetta, Miezin, Dobmeyer, Shulman, & Petersen, 1991; MacLeod & MacDonald, 2000).
In the Stroop effect, attentional demands are higher during the incongruent conditions compared to the congruent trials due to higher conflict between stimulus attributes. Banich et al. (2001) proposes that the posterior system regulates attentional resources by processing information associated with the task-irrelevant attribute of the stimuli (the ink color in a word reading task). Indeed, the visual word form area (VFWA) was found to be more activated in a color-naming interference effect (Wallentin, Gravholt, & Skakkebsek, 2015). Interestingly, a decrease in the interference effect via adaptation to conflict may be modulated by increasing activity in task-relevant posterior brain regions (Purmann & Pollmann, 2015). Contrary to this, other research has shown an increase in activity in task-relevant areas and decrease in task-irrelevant areas (Harrison et al., 2005). Thus, clarification needs to be added to the existing literature on how the anterior and posterior attentional systems modulate conflict in a task-specific way.
In trying to understand how activation in different neural areas contributes to the interference effect, it is important to look at how the Stroop effect is measured within the
6


study design. In the literature, the Stroop effect has been achieved by comparing incongruent trials to either neutral or congruent conditions (Nee et al., 2007). Comparing incongruent and congruent conditions has the advantage that physical characteristics and cognitive processing (lexical and semantic) of the stimuli are the same, preventing unanticipated neural activation unrelated to the control and attentional processing of the interference effect (Milham & Banich, 2005; Peterson et al., 1999). This comparison, however, confounds the effect of stimulus encoding conflict with response related conflict (Nee et al., 2007).
Conversely, comparing incongruent and neutral conditions gives the advantage of being able to see the effect that response conflict has on the interference effect (Aarts, Roelofs, & van Turennout, 2009; Zysset, Muller, Lohmann, & von Cramon, 2001). For the color-naming task, the neutral stimulus is a non-color word or non-word (e.g. XXXX) presented in colored ink. In the word-reading task, the neutral stimulus is a color word presented in black ink. In both of these conditions, the task-relevant attribute is the only aspect of the stimulus that can inform a response decision. This allows researchers to investigate the neural correlates of the response conflict that exists in the incongruent condition. However, because physical and cognitive attributes of the stimuli are not the same, additional neural activation may be included in the analysis of the incongruent>neutral comparison (Peterson et al., 1999). By making multiple comparisons between different experimental conditions with the Stroop task, the nuances associated with the conflict and attentional processing of the interference effect can be better understood.
Neuroimaging Findings
As previously mentioned, there have been numerous neuroimaging studies dedicated to the uncovering and understanding of the neural activations behind the Stroop interference
7


effect. There are several areas of brain activation during the Stroop effect that are widely accepted as contributing to attentional or control systems necessary to process the incongruent stimuli. However, the manipulation of study design and the use of different imaging techniques has led to an ongoing debate on the exact spatial and temporal distribution of brain areas of interest (C. Carter & Van Veen, 2007; Larson et al., 2014; Nee et al., 2007). fMRI and PET
Results from fMRI and a few positron emission topography (PET) studies have made use of their high spatial resolution to identify areas associated with various cognitive processes. The anterior cingulate cortex (ACC), left pre-frontal cortex (PFC), and various posterior regions have been repeatedly found in comparing incongruent to congruent or neutral trials in the Stroop task (Nee et al., 2007).
The anterior cingulate cortex (ACC) has been proposed to play a variety of roles in the cognitive processing of the interference effect. In several studies, the ACC has been shown to a have a executive role in detecting and reducing conflict in a top down manner (C. Carter & Van Veen, 2007; C. Carter et al., 2000; Nee et al., 2007; Silton et al., 2010; Van Veen & Carter, 2002) and contributing more generally to performance and conflict monitoring (Adleman et al., 2002; Barch et al., 2001; Botvinick, Cohen, & Carter, 2004; Braver, Barch, Gray, Molfese, & Snyder, 2001; Fan et al., 2003; Hanslmayr et al., 2008; Kerns, 2004; MacDonald et al., 2000; Nordahl et al., 2001; Van Veen & Carter, 2002). Aside from conflict monitoring, the ACC functions in executive attentional control with respect to response conflict alone (Milham et al., 2001; van Veen, Cohen, Botvinick, Stenger, & Carter, 2001), with respect to both task and response conflict (Aarts et al., 2009), and with respect to
8


coordinating the activity of multiple attentional subsystems (Cieslik, Mueller, Eickhoff, Langner, & Eickhoff, 2015; Peterson et al., 1999).
Other areas of the prefrontal cortex (PFC) has also been implicated in task preparation and establishing the attentional set to the present task during the Stroop effect (Banich et al., 2000a; Brass, Ullsperger, Knoesche, von Cramon, & Phillips, 2005; C. Carter & Van Veen, 2007; Harrison et al., 2005; MacDonald et al., 2000; Milham et al., 2001). In accordance with the idea of parallel, coordinated anterior and posterior attentional systems, evidence supports the role of the prefrontal cortex in integrating general task information and informing relevant regions within the parietal cortex (Banich et al., 2001; Banich et al., 2000b; Brass et al., 2005). Particular areas of activation within the PFC during the Stroop effect include the left PFC (Bunge, Hazeltine, Scanlon, Rosen, & Gabrieli, 2002; Harrison et al., 2005; Milham et al., 2001; Van Veen & Carter, 2005; Zysset et al., 2001), left dorsolateral PFC (DLPFC) (Aarts et al., 2009; Banich et al., 2000b; C. Carter & Van Veen, 2007; MacLeod & MacDonald, 2000; Nee et al., 2007; Silton et al., 2010), ventro-lateral PFC (VLPFC) (Aarts et al., 2009; Purmann & Pollmann, 2015), right lateral PFC (Milham et al., 2001), medial PFC (Oehrn et al., 2014; Purmann & Pollmann, 2015), and pre-motor cortex (Zysset, Schroeter, Neumann, & Yves von Cramon, 2007).
The fMRI research on activation within the parietal lobe has been more varied in terms of spatial distribution as compared to ACC and PFC activation during the Stroop effect. Specific areas of activation noted in the research include the posterior parietal cortex (Adleman et al., 2002; Nee et al., 2007; Purmann & Pollmann, 2015; Rushworth, Johansen-Berg, Gobel, & Devlin, 2003), inferior parietal cortex (C. Carter, Mintun, & Cohen, 1995; Rushworth, Paus, & Sipila, 2001) left parietal-occipital lobe (Adleman et al., 2002),
9


intraparietal sulcus (Zysset et al., 2001) and general parietal cortex (Van Veen & Carter, 2005; Wallentin et al., 2015). The parietal cortex, particularly on the left side, is thought to play a role in processing the physical attributes of both task-relevant and task-irrelevant information in order to inform output responses (Banich et al., 2000b; Herd, Banich, & OReilly, 2006). Worth noting are studies that found additional activation in the occipital lobe (Adleman et al., 2002), occipito-temporal cortex (Zysset et al., 2001) and VWFA within the temporal cortex (Wallentin et al., 2015) in the color-naming Stroop effect.
EEG
Complementary to the spatially resolved fMRI research are scalp encephalography (EEG) studies that take advantage of their high temporal resolution to describe chronological activation during the Stroop effect. EEG non-invasively measures electrical activity of the brain using electrodes that are placed evenly over the subjects skull (Luck, 2005). Rooted within this electrical activity are neural responses called event related potentials (ERPs) that are associated with specific sensory, cognitive, and motor events (Barkley & Baumgartner, 2003; Malmivuo, 2012).
Two main ERPs associated with the interference effect have been consistently reported in the literature (Liotti, Woldorff, Perez, & Mayberg, 2000; West, 2003). The first is a medial dorsal or fronto-central negativity, known as the medial frontal negativity (MFN), N450, or Ninc, that is characterized as occurring 350 to 500 ms post stimulus (Donohue, Appelbaum, McKay, & Woldorff, 2016; Larson, Farrer, & Clayson, 2011; Liotti et al., 2000). This wave is thought to have its neural origins within the ACC (Badzakova-Trajkov, Barnett, Waldie, & Kirk, 2009; Hanslmayr et al., 2008; Larson et al., 2014; Liotti et al., 2000; Szucs & Soltesz, 2010; Van Veen & Carter, 2002; West & Alain, 2000) but a few sources have
10


found neural correlates elsewhere such as in the PFC (Markela-Lerenc et al., 2004) and posterior cingulate cortex (West, Bailey, Tiernan, Boonsuk, & Gilbert, 2012; Xiao, Qiu, & Zhang, 2009). The cognitive functioning of the MFN is thought to be in conflict detection and monitoring (Appelbaum, Boehler, Davis, Won, & Woldorff, 2014; Coderre, Conklin, & Van Heuven, 2011; Larson et al., 2014; Szucs & Soltesz, 2010; Van Veen & Carter, 2002; West & Alain, 2000; West et al., 2012).
The second ERP wave is described as a centro-parietal positive slow wave known as the conflict slow potential (conflict SP) (Larson et al., 2014, 2011; Larson, Kaufman, & Perlstein, 2009), but is also referred to the late positive component (LPC) (Appelbaum et al., 2014; Donohue et al., 2016). The conflict SP develops approximately between 500ms and 800 ms post stimulus, but has been shown to extend past 1000ms (West & Alain, 2000). Activation has been documented over centro-parietal (Chen, Bailey, Tiernan, & West, 2011; Larson et al., 2011; West et al., 2012; West, 2003) as well as left temporo-parietal scalp regions (Donohue, Liotti, Perez, & Woldorff, 2012; Liotti et al., 2000; West & Alain, 2000) Source reconstruction techniques have identified its neural generator as the lateral frontal and posterior parietal cortex(Chen et al., 2011; West, 2003) It has been suggested that the conflict SP functions in conflict adjustment and resolution mechanisms (Appelbaum et al., 2014;
Chen et al., 2011; Larson et al., 2014, 2011, 2009; West & Alain, 2000).
MEG
The majority of research that focuses on explaining the neural mechanisms behind Stroop interference effect uses fMRI or EEG data. However, for the past two decades, magnetoencephalography (MEG) has been used as a method of noninvasive functional imaging of the brain (Malmivuo, 2012; Wheless et al., 2004). Applying this measure of
11


neural activity to the Stroop task paradigm has very rarely been seen in past research and would serve to add clarity to the pool of existing fMRI and EEG results on this subject (Galer et al., 2014; Eikai et al., 2002).
Whenever an ERP is generated, a magnetic field, called an event related field (ERF) is also generated around the ERP dipole (Luck, 2005). MEG measures these magnetic signals generated by pyramidal cells that largely lie orthogonal to the cortical surface with millisecond temporal resolution (Hamalainen, 1992). The most important benefit of ERFs is that these magnetic fields are not blurred by contact with the skull (Barkley & Baumgartner, 2003; Hamalainen, 1992). When EEG records electrical voltage at any given moment from electrodes on the skull, it reflects the combination of ERPs generated from several different sources throughout the brain. Electric potentials tend to spread out laterally when they encounter the skull and thus a signal recorded by an EEG electrode reflects a skewed picture of the origin of those potentials (Luck, 2005). Therefore it is believed by many researchers that the spatial resolution of MEG is superior to that of EEG and can be expected to identify activated sources within the millimeter to centimeter range (Barkley & Baumgartner, 2003; Wheless et al., 2004). It is also worth noting that MEG and EEG are not spatially sensitive to the same neural sources (Ahlfors, Han, Belliveau, & Hamalainen, 2010; Sharon, Hamalainen, Tootell, Halgren, & Belliveau, 2007). MEG is highly sensitive to tangential sources close to the surface of the skull, while EEG can detect much deeper sources (Ahlfors et al., 2010; Hamalainen, 1992).
While several studies exist that measure Stroop effect related neural activity with EEG, there are only four studies that use the Stroop task with MEG (Galer et al., 2014; Kawaguchi et al., 2004, 2005; Ukai et al., 2002). In three of these studies, the Stroop task
12


was used to locate the neural origins of the incongruent conditions without any comparison to congruent or neutral conditions (Kawaguchi et al., 2004, 2005; Ukai et al., 2002). Therefore the researchers could only report on the brain activation associated with information processing during the Stroop task, rather than the Stroop effect. In the final study, Galer et al. (2015) used the Stroop test to look at interference-related neural events by comparing incongruent with congruent trials. Using minimum norm source estimation, Galer et al. (2015) found higher activation in the left pre-supplementary motor and left posterior parietal cortex in the incongruent trials from 480 to 700ms post stimulus (Galer et al., 2014).
Gaps in Existing Research
The Stroop interference effect has been documented in many different conditions of conflict. Yet questions still remain about the extent to which the ACC, PFC, parietal, and other posterior cortices chronologically contribute to various instances of the interference effect. Findings from fMRI studies provide evidence that areas within the PFC function in selective attention and conflict monitoring (Nee et al., 2007). Yet due to the poor temporal resolution of fMRI, the extent to which these structures function via top down biasing or sustained feedback loops with posterior brain regions cannot be determined. On the other hand, the poor spatial resolution of EEG results only allows for general brain regions to be identified as active at any point in time post stimulus (Larson et al., 2014). There is a need for spatially resolved studies that show the patterns of neural activation over time. Studies using MEG can help fill that gap.
The majority of studies that examine the Stroop interference effect only examine the neural correlates associated with one contrast (i.e. incongruent>neutral for color-naming) (Nee et al., 2007). Yet the interference effect has been associated with both stimulus and
13


response conflict within different types of tasks (Aarts et al., 2009; Milham & Banich, 2005; Nee et al., 2007). There is a need for interference related studies to examine multiple conflict contrasts within the same study design and subject population. This allows for the critical, meaningful examination of differences in patterns of neural activation for different conflict types that result behaviorally in the interference effect.
Conclusion
The primary goal of the present study is to use an alternative measure of electrophysiological activity, MEG, to add spatial and temporal clarity to the roles that various neural areas play in different conditions of cognitive conflict. In comparing incongruent to congruent stimuli, neural networks contributing to conflict conditions where both stimulus and response conflict are present can be explored. The comparison of incongruent to neutral stimuli allows the role of response conflict to become salient within the interference task. Both of these interference conditions are examined in the context of their task type (color-naming and word-reading) to further determine if the task type influences the conflict and attentional pathways recruited during the different presentations of the Stroop effect. By using MEG, neural generators of the Stroop effect can be explored with millisecond temporal resolution. In this way, the present study will report on the neural correlates of various conditions of conflict with temporal and spatial precision in a manner that is yet to be seen in the literature.
14


CHAPTER III
SPECIFIC AIMS AND HYPOTHESES
Specific Aim 1: Localize the neural generators and describe the time course of activation behind the interference effect with respect to the task type (color naming versus word reading) using MEG as the measure of neural activity.
Hypothesis 1: There will be more significant activation in the prefrontal cortex in the color naming contrasts as compared to the word reading contrasts due to the automacity associated with word reading (MacLeod & MacDonald, 2000). Posterior regions will have task-specific patterns of activation (Corbetta et al., 1991; MacLeod & MacDonald, 2000). Activation in the prefrontal areas will initially precede activation in posterior regions in line with a top down biasing model, but sustained activation in both regions will be seen from 500-1000ms post stimulus (Miller & Cohen, 2001).
Specific Aim 2: Localize the neural generators and describe the time course of activation behind the interference effect with respect to the type of conflict (incongruent>neutral (I>N) versus incongruent>congruent (I>C)) using MEG as the measure of neural activity. Hypothesis 2: Both conflict types (incongruent>neutral and incongruent>congruent) will have significant activation in the prefrontal cortex (Nee et al., 2007). I>N will have greater activation in the left DLPFC and I>C will have greater activation in the ACC (Nee et al., 2007). Posterior regions will have condition specific patterns of activation. Activation in the prefrontal areas will initially precede activation in posterior regions, but sustained activation in both regions will be seen from 500-1000ms post stimulus (Miller & Cohen, 2001).
15


Specific Aim 3: Localize the neural generators and describe the time course of activation behind the interference effect with respect to general conflict processing (i.e. task-independent, conflict type independent) using MEG as the measure of neural activity. Hypothesis 3: There will be common areas of activation in all four contrasts in both the prefrontal cortex (ACC, DLPFC) and posterior parietal cortex (Nee et al., 2007). The timing of the activation will reflect top-down biasing. Significant activation in the ACC will occur first, between 350 and 500ms post stimulus. Sustained activation in other regions of the PFC and posterior regions will occur between 500 and 1000ms post stimulus (Miller & Cohen, 2001).
16


CHAPTER IV
MATERIALS AND METHODS Ethics Statement
COMIRB approved this study protocol (Number 10-0539) and all participants signed informed consent statements.
Participants
Brain activity of thirty participants was recorded using simultaneous MEG and EEG. Data from one participant was excluded from this study due to excessive motion artifact. Of the twenty-nine remaining participants, there were 18 females and 11 males ranging in age from 19 to 36 years old. All participants were right-handed. Prior to the recordings, participants were instructed to get a full night of sleep and to refrain from consuming caffeine and nicotine in order to reduce cognitive changes not associated with the target testing. All recordings started at 10:00 am.
Study Design: Computerized Stroop Test
Participants performed a computerized cued version of the Stroop test originally developed by Cohen et al. (1999), in an acoustically and magnetically sealed room. Each trial of the Stroop test consisted of three parts: the presentation of a cue (either word or color), a 1 or 5 second cue-stimulus interval, and a visual word stimulus (green, red, or blue). Initial cues were presented as auditory instructions through a loudspeaker and subsequent stimuli were presented visually on an LCD screen. The stimulus features (word color and word meaning) were congruent, incongruent, or neutral. Congruent stimuli were color words printed in the same color (i.e. BLUE printed in blue) and incongruent stimuli were color words printed in a different color (i.e. BLUE printed in green). Neutral stimuli


Congruent
Incongruent Neutral (word) Neutral (color)
Figure 1. The Stroop task paradigm consisting of three parts: task specific auditory cue, followed by the visual stimulus to which subjects responded with a keypress. The time period of interest in this study was -200ms to 1000 with 0ms indicating stimulus onset. Reaction time was calculated between the stimulus onset and time of appropriate keypress.
were non-color words printed in red, blue, green for the color-naming task (i.e. CAR printed in red) and color words printed in black for the word reading tasks (i.e. RED printed in black). On trials where the cue was word, participants stated the word meaning. On trials where the cue was color, participants stated the ink color of the word. The trials were evenly divided based on task type (word reading or color naming), transition type (switch or repeat), cue-stimulus interval (1 or 5 seconds) and congruency of stimuli (congruent, incongruent, or neutral). Participants responded to each trial stimulus by pressing one of three buttons on a keypad. After five minutes of practice trials, participants performed this experiment for a single three-hour session. This analysis only looks at the first 360 trials.
Magnetoenchephalography (MEG) Recording
Magnetic evoked fields were recorded using a Magnes 3600 WH whole-head MEG device (4-D Neuroimaging, San Diego, CA, EISA), which comprises 248 first-order axial-
18


gradiometer sensors in a helmet-shaped array. Data was collected at a sampling rate of 291 Hz. Each participants head position was determined with respect to the sensor array using five head position indicator coils attached to the subjects scalp. The locations of the coils with respect to three anatomical landmarks (nasion and pre-auricular points) and two extra non-fiducial points, as well as the scalp surface were determined with a 3D digitizer (Polhemus, Colchester, VT, USA).
MEG Pre-Processing
Pre-processing of the data was performed with Brainstorm (Tadel, Baillet, Mosher, Pantazis, & Leahy, 2011), which is documented and freely available for download online under the GNU general public license (http://neuroimage.usc.edu/brainstorm). Custom scripts written in Matlab r2104a (The MathWorks, Inc., Natick, MA, USA) were also employed. The continuous data was initially visually inspected in order to reject time segments with poor signal quality and excessive noise. The raw data set was band-pass filtered between 0.1 Hz and 30 Hz. Unwanted artifacts, including eye blinks and cardiac signal, were removed using SSP (Signal-Space Projection) functions within Brainstorm. Epochs were set up spanning -200ms to 1000ms relative to cue onset (0ms) and only those trials with correct responses were included in the subsequent analysis. Epochs were baseline corrected to remove any DC offset and trials with excessive amplitude (>4500fT) were rejected. Baseline was defined as -200 to -5ms relative to stimulus onset. For each subject, trials within each experimental condition were then averaged. Pseudo-individual brain anatomies were created for each subject by warping the default anatomy provided by Brainstorm to fit the digitized head points collected with a magnetic tracking system prior to MEG data acquisition.
19


Source Localization Estimation
In order to spatially estimate the neural activity measured by the MEG sensors, two separate modeling steps were taken. First the head model, or forward model, was constructed within Brainstorm using overlapping spheres for each sensor to model the tissue layers of the brain in relation to the sensor instrumentation. This allowed the magnetic fields measured at the sensors to be connected to their electric generators within the brain. Second, the ill posed inverse model was addressed. This step of modeling involves estimating the activity of thousands of cortical dipoles modeled with the head model with the activation from 248 MEG channels. The challenge lies with the possibility of an infinite number of combinations causing the pattern of activation measured at the sensor level. The inverse model was applied via established minimum norm source estimation techniques (Hamalainen, 2010; Liu, Dale, & Belliveau, 2002). Within Brainstorm, a whitened and depth weighted linear L2-minimum norm estimation algorithm was implemented (Hamalainen, 2010). This estimation was calculated assuming one dipole at each vertex of the cortical surface, oriented normal to the surface. These estimates produce a density map of flowing current at the surface of the cortex.
After sources were reconstructed separately for each experimental condition within each subject, source images were spatially smoothed using a Gaussian kernel of 12mm FWHM. Paired t-tests were conducted on the following sets of contrasts: incongruent, color>congruent, color (IC>CC); incongruent, color>incongruent, neutral (IC>IN); incongruent, word>congruent, word (IW>CW); incongruent, word>neutral word (IW>NW). Each paired t-test resulted in a map of voxel values consisting of t-values. For each of the four contrasts, t-maps were generated for 100ms averaged time windows, ranging from 0 to
20


1000 ms post stimulus. Areas of significant difference were reported at p<0.01 corrected for false discovery rate (FDR). On those maps displaying significant difference, regions of interest were defined based on local maximum t-values in areas of significance of at least 40 voxels. For each regions of interest, the center of weight and mean t-value were calculated using custom Matlab scripts. Spatial locations of ROIs were defined by Bradmans Area using MRIcron software (Rorden, Kamath, & Bonilha, 2007).
21


CHAPTER V
RESULTS Behavioral Results
Verification of the Stroop Interference Effect with Reaction Time (RT)
In the color naming condition, there was a significant difference in mean reactions time between incongruent (M= 1.18, SD = 0.33) and neutral (M= .85, SD = 0.29) conditions; t(28) = 13.4, p<0.001 and the incongruent and congruent (M= .91, SD = 0.29) conditions; t(28) = 8.2 p<0.001 (see Table 1). In the word reading condition, there was also a significant difference in mean reactions time between incongruent (M= 1.18, SD = 0.34) and neutral (M= .87, SD = 0.26) conditions; t(28) = 14.7, p<0.001 and the incongruent and congruent (M= .97, SD = 0.32) conditions; t(28) = 10.0 p<0.001. This verifies that during both the color naming and word reading tasks, an interference effect was experienced in the form of decreased reaction time during the incongruent stimulus. Mean reaction time was measured on trials answered correctly. Aside from measuring the traditional interference effect, it is worth noting that in both the color-naming and word-reading conditions, neutral tasks were performed significantly faster than congruent tasks; t(28) = 7.6, p<0.001 for word-
Table 1
Mean Percent Accuracy
Task Type
Congruency Condition Color-Naming Word-Reading
Incongruent 86.5% 91.3%
Congruent 95.3% 95.7%
Neutral 95.3% 94.6%
22


reading; t(28) = 4.4, p<0.001 for color naming. This indicates that more conflict was experienced during the congruent tasks than the neutral tasks, regardless of the stimulus attribute to which attention was given (color or word).
Verification of the Stroop Interference Effect with Response Accuracy
In the color naming condition, a significant difference in response accuracy was found between incongruent (M= 0.87, SD = 0.14) and neutral conditions (M= .95, SD = 0.06); t(28) = 4.41, p<0.001 and the incongruent and congruent (M=.95 SD =0.05) conditions; t(28) = 4.14, p< 0.001 (see Table 2). In the word reading condition, a paired t test between incongruent (M= 0.91, SD = 0.08) and neutral (M= 0.95, SD = 0.07) conditions had a t(28) = 5.11, p<0.001 and the incongruent and congruent (M= 0.96, SD = 0.07) conditions had t(28) =5.83, p<0.001. This shows that in all four contrasts, the interference effect was seen in how accurate subjects were in completing the task at hand. It is also worth noting that there is a significant difference in response accuracy between incongruent conditions of the color naming task and incongruent conditions of the word reading task; t(28) =2.6, p<0.05. This demonstrates that on a behavioral level, there is a higher level of accuracy in responding to the word reading tasks over color naming in the incongruent
Table 2
Mean Reaction Times and Standard Deviations (in seconds)
Task Type
Congruency Condition Color-Naming Word-Reading
Incongruent 1.18(0.33) 1.18(0.34)
Congruent 0.91 (0.29) 0.97 (0.32)
Neutral 0.85 (0.29) 0.87 (0.26)
23


conditions. This provides evidence for the argument that, even within the cued version of the Stroop task, word reading is more automatic than color naming (MacLeod & MacDonald, 2000).
Source Space Analysis Results
For each of the four contrasts (IOCC, IONC, IW>CW, IW>NW), ten paired t-tests were run using the minimum norm estimate source images on 100ms time averaged windows ranging from 0 to 1000ms post stimulus. Areas of significant difference were reported at p <0.01 corrected for false discovery rate (FDR). For all four contrasts, no significance was found from 0 to 500ms. For IONC and IW>CW, significant difference was detected on the five t-value source maps generated from 500ms to 1000ms. For IOCC and IW>NW, significant difference was detected on the four t-value source maps generated from 600ms to 1000ms. On each t-value map displaying significant difference, regions of interest (ROIs) were defined based on local maximum t-values in areas of significance of at least 40 voxels. Color-Naming: Incongruent>Neutral
The two contrasts where color naming was performed resulted in more significantly activated regions of interests than the two contrasts of the word-reading task. For the IC>NC contrast, ten ROIs were identified over the 500 to 1000ms windows. These areas include left frontopolar, dorsomedial prefrontal (DMPFC), and premotor regions, left and right occipital cortices, and left occipito-temporal, temporal, and parietal regions. Activation in the frontal regions was strongest on the front end of the time series (500-700ms) and activation in the posterior areas of the brain (parietal, occipital, temporal) was strongest between 700 and 900ms post stimulus. In Table 3, each significant ROIs for the IC>NC contrast are listed by anatomical and functional region. For each 100ms time-averaged window, the center of
24


weight and mean t-value are recorded for those areas where significant difference was detected within that time window. Figure 2 gives a visual representation of the data described in Table 3.
Color-Naming: Incongruent>Congruent
In the IC>CC Stroop effect, activation was seen in eight distinct regions over the 600 to 1000ms time windows. These ROIs are focused in the left hemisphere in the dorsolateral prefrontal cortex (DLPFC), premotor cortex, inferior temporal gyrus, occipito-temporal cortex, temporal/parietal junction and various occipital regions. Activation is also seen in the right posterior parietal cortex. Notably, throughout the time series (600-1000ms), strong activation was recorded in the inferior temporal gyrus and temporal/parietal junction, areas highly associated with language (Bonner & Price, 2013; R. Carter & Huettel, 2013). Activation in the frontal cortex is more transient with significant activation limited to 100ms time blocks for the DLPFC and premotor areas. Occipital regions are most active later in the time series (700-1000ms). In Table 4, the significant ROIs for the IC>CC contrast are listed by anatomical and functional region. For each 100ms time-averaged window, the center of weight and mean t-value are recorded for those areas where significant difference was detected within that time window. Figure 3 gives a visual representation of the data described in Table 4.
Word-Reading: Incongruent>Neutral
In the IW>NW Stroop effect, activation was seen in five areas over the 600ms to 1000ms time windows. ROIs were restricted to the occipital and parietal lobes in both the right and left hemispheres. Unlike the color-naming contrasts, no activation was seen in the temporal or frontal lobes. Activation across the time series was strongest in the left occipital
25


and parietal cortices. In Table 5, the significant ROIs for the IW>NW contrast are listed by anatomical and functional region. For each 100ms time-averaged window, the center of weight and mean t-value are recorded for those areas where significant difference was detected within that time window. Figure 4 gives a visual representation of the data described in Table 5.
Word-Reading: Incongruent>Congruent
In the IW>CW Stroop effect, activation was seen in four areas over the 500ms to 1000ms time windows. ROIs were identified in the left primary and association visual cortices and right secondary visual cortex. The strongest and most sustained activation was recorded in the temporal/parietal junction. In Table 6, the significant ROIs for the IW>CW contrast are listed by anatomical and functional region. For each 100ms time-averaged window, the center of weight and mean t-value are recorded for those areas where significant difference was detected within that time window. Figure 5 gives a visual representation of the data described in Table 6.
26


Time Color-Naming: Incongruent>Neutral
(post stimulus)
500ms
600ms
700ms
800ms
900ms
^ t 1000ms
L Dorsomedial Prefrontal Cortex
L Visual Association Cortex
L Temporal/Parietal Junction
Color Incongrutnt Color Ntulril
Figure 2. On the left, visual maps of posterior and left sagittal views of significant t-values with p<0.01 FDR for IC>NC contrast in 100ms time averaged windows from 500ms to 1000ms post stimulus. On the right, the current density, time series activation for three regions of interest are displayed. The x-axis is time in ms, spanning the -200 to 1000ms time window, with 0ms as the stimulus onset. The y-axis is the current density amplitude in pA.m.
27


Table 3
Color-Naming: Incongrnent>Nentral Regions of Interest (ROIs)
500-600ms 600-700ms 700-800ms 800-900ms 900-1000ms
Functional Region (Brodmans .Area) MNI t MNI t MNI t MNI t MNI t
L Premotor (6) [-29. 3. 53] 4.5 [-28. 3. 53] 3.6 [-28. 3. 54] 3.5
L DMPFC (8/9) [-8. 39. 49] 5.2 [-9. 38. 49] 5.1 [-10. 38. 48] 3.3 [-10. 39. 48] 3.5
L Wernickes Area [-60. -30. 3] 3.6
(21/22) L Frontopolar (10) [-5. 58. 4] 3.8 [-5. 58. 4] 3.2 [-5. 58. 3] 3.6
L TPJ (21/39) [-50. -60. 23] 4.4 [-50. -60. 23] 4.9 [-50. -59. 23] 4.9 [-50. -60. 23] 4.2
L VI (17) [-7. -88. -2] 4.3 [-6. -88. -1] 5.5 [-7. -88. -3] 5.4 [-7. -88. -3] 5.4
L Vis Assoc (19) [-42. -84. 9] 5.1 [-41. -84. 9] 6.2 [-41. -85. 9] 5.3 [-41. -85. 9] 4.7
RV2 (18) [17. -95. 2] 3.3 [17. -94. 1] 3.4
L Inf Temp (20) [-39. -18. -34] 3.4 [-40. -20. -34] 3.8
L Occ/Temp (37) [-54. -66. -13] 4.8
Note. The ROis identified in the IC>NC contrast with significant activation! p<0.01 with FDR correction. Each ROI was identified by local maximum t-values and is made up of at least 40 voxels. The MNI [x.y.z] coordinates for the center of weight and mean t-value of each ROI are reported for each 100ms time averaged window (ie. 500-600ms) where significant activation was found. Functional Region Abbreviations: DMPFC = dorsomedial prefrontal cortex: TPJ = temporal/parietal junction: VI = primary visual cortex: Vis Assoc = visual association cortex: V2 = secondary visual cortex: Inf Temp = inferior temporal cortex: Occ/Temp = occipital temporal cortex
28


L Dorsolateral Prefrontal Cortex
SaA'/xV
L Occipito/Temporal Cortex
L Temporal/Parietal Junction
Color-Naming: Incongruent>Congruent
Time
(post stimulus)
600nrs
700nrs
800nrs
900nrs
, r 1000ms
Color Congruent ------Color Incongruent
Figure 3. On the left, visual maps of posterior and left sagittal views of significant t-values with p<0.01 FDR for IOCC contrast in 100ms time averaged windows from 600ms to 1000ms post stimulus. On the right, the current density, time series activation for three regions of interest are displayed. The x-axis is time in ms, spanning the -200 to 1000ms time window, with 0ms as the stimulus onset. The y-axis is the current density amplitude in pA.m.
29


Table 4
Color-Naming: Incongruent> Congruent Regions of Interest (ROIs)
600-700ms 700-800ms 800-900ms 900-1000ms
Functional Region (Brodmans .Area) MNI t MNI t MNI t MNI t
R Post Parietal (7) [22. -74. 43] 4.1 [21. -73. 43] 4.2 [21. -73. 43] 3.6
L TPJ (21/39) [-53. -57. 19] 4 [-53. -57. 19] 4.7 [-53. -57. 18] 5.0 [-53. -57. 18] 4.5
L Inf Temp (20) [-40. -18. -34] 4.1 [-39. -18. -34] 4.1 [-39. -18. -34] 3.7 [-39. -18. -35] 3.6
L Occ/Temp (37) [-54. -67. -12] 3.3 [-54. -66. -13] 4.4 [-54. -65. -13] 4.0 [-54. -65. -13] 3.6
L DLPFC (46) [-36. 31. 37] 3.5
L VI (17) [-13. -98. 2] 3.7 [-13. -98. 2] 3.6 [-13. -98. 2] 3.4
L Visual Assoc (19) [-35. -85. 31] 4.2 [-35. -85. 31] 4.6 [-35. -85. 31] 4.1
L Premotor (6) [-49. -6. 41] 4.3
Note. The ROis identified in the IC>CC contrast with significant activation! p<0.01 with FDR correction. Each ROI was identified by local maximum t-values and is made up of at least 40 voxels. The MNI [x.y.z] coordinates for the center of weight and mean t-value of each ROI are reported for each 100ms time averaged window (i.e. 500-600ms) where significant activation was found. Functional Region Abbreviations: Post Parietal = posterior parietal cortex: TPJ = temporal/parietal junction: Inf Temp = inferior temporal cortex: Occ/Temp = occipital temporal cortex: DLPFC = dorsolateral prefrontal cortex: VI = primary visual cortex: Vis Assoc = visual association cortex:
30


L Primary Visual Cortex
Word-Reading: Incongruent>Neutral
Time
(post stimulus)
600ms
L Visual Association Cortex
700ms
800ms
L Temporal/Parietal Junction
900ms
1000ms
Word Incongrutnt ------Word Nutnl
Figure 4. On the left, visual maps of posterior and left sagittal views of significant t-values with p<0.01 FDR for IW>NW contrast in 100ms time averaged windows from 600ms to 1000ms post stimulus. On the right, the current density, time series activation for three regions of interest are displayed. The x-axis is time in ms, spanning the -200 to 1000ms time window, with 0ms as the stimulus onset. The y-axis is the current density amplitude in pA.m. onset.
31


Table 5
Word-Reading: Incongrnent>Nentral Regions of Interest (ROIs)
600-700ms 700-800ms 800-900ms 900-1000ms
Functional Region (Brodmans .Area) MNI t MNI t MNI t MNI t
L VI (17) [-12. -98. 1] 5.1 [-12. -98. 2] 6.0 [-11. -98. 1] 6.1 [-11. -98. 1] 5.4
R V2 (18) [17. -95. 2] 4.0 [17. -94. 2] 3.7 [17. -94. 2] 4.2 [17. -95. 2] 4.2
L Visual Assoc (19) [-25. -85. 32] 5.2
R Post Parietal (7) [22. -74. 43] 4.2
L TPJ (7/21/39) [-37. -71. 38] 4.4 [-37. -71. 38] 5.0 [-37. -71. 38] 5.1
Note. The ROis identified in the IW>NW contrast with significant activation! p<0.01 with FDR correction. Each ROI was identified by local maximum t-values and is made up of at least 40 voxels. The MNI [x.y.z] coordinates for the center of weight and mean t-value of each ROI are reported for each 100ms time averaged window (i.e. 600-700ms) where significant activation was found. Functional Region Abbreviations: VI = primary visual cortex: V2 = secondary visual cortex: Vis Assoc = visual association cortex: Post Parietal = posterior parietal cortex: TPJ = temporal/parietal junction:
32


600ms
700ms
800ms
900ms
, t 1000ms
Time Word-Reading: Incongruent>Congruent
(post stimulus)
500ms
L Visual Association Cortex
Word Congruent ------Word Incongrutnt
L Primary Visual Cortex
L Temporal/Parictal Junction
Figure 5. On the left, visual maps of posterior and left sagittal views of significant t-values with p<0.01 FDR for IW>CW contrast in 100ms time averaged windows from 500ms to 1000ms post stimulus. On the right, the current density, time series activation for three regions of interest are displayed. The x-axis is time in ms, spanning the -200 to 1000ms time window, with 0ms as the stimulus onset. The y-axis is the current density amplitude in pA.m.
33


Table 6
Word-Reading: Incongruent> Congruent Regions of Interest (ROIs)
500-600ms 600-700ms 700-800ms 800-900ms 900-1000ms
Functional Region (Brodmans .Area) MNI t MNI t MNI t MNI t MNI t
L TPJ (21/39) [-48.-64.25] 5.0 [-48.-64.25] 4.7 [-48.-64.25] 6.7 [-48. -64. 25] 6.0 [-48. -64. 25] 5.7
L Visual Assoc (19) [-19.-88.33] 4.1 [-19.-88.33] 5.3 [-35. -89. 0] 4.6 [-35. -89. 0] 4.5
R V2 (18) [19. -94. 16] 3.7
L VI (17) [-14. -98. 13] 5.2 [-14. -98. 13] 4.4
Note. The ROis identified in the IW>NW contrast with significant activations p<0.01 with FDR correction. Each ROI was identified by local maximum t-values and is made up of at least 40 voxels. The MNI [x.y.z] coordinates for the center of weight and mean t-value of each ROI are reported for each 100ms time averaged window (i.e. 500-600ms) where significant activation was found. Functional Region Abbreviations: TPJ = temporal/parietal junctions Vis Assoc = visual association cortexs V2 = secondary visual cortexs VI = primary visual cortexs
34


CHAPTER VI
DISCUSSION
In the present study, we used the spatial and temporal precision of MEG to investigate the neural correlates of the Stroop interference effect in a cued, task switching Stroop task where incongruent trials were compared to congruent and neutral trials separately in the context of the specific task performed (color-naming or word-reading). These different contrasts represent four variations of neural conflict and attentional processing that could be present in an interference effect. The behavioral results reveal that the interference effect was present in all four contrasts in the form of significantly reduced accuracy and reaction time in conditions of incongruent stimuli as compared to either neutral or congruent stimuli. Comparisons of source estimation maps reveal that certain posterior areas of the brain are recruited in all four variations of conflict while activation in the temporal and frontal lobes was limited to the color-naming trials. However, the overall spatial and temporal patterns of activations were distinct for all four conflict effects examined in this study.
Behavioral Implications
In measuring the mean reaction time and accuracy for all subjects across all trials, we found that behaviorally, the Stroop interference effect was present when comparing incongruent trials to either neutral or congruent trials in both the color naming and wordreading conditions. However, in both tasks, neutral trials were performed significantly faster than congruent trials. This indicates that more conflict was experienced during the congruent tasks than the neutral tasks, regardless of the stimulus attribute to which attention was given (color or word). In examining the nuances in response accuracy, subjects were significantly more accurate in responding to incongruent conditions of the word-reading task than
35


incongruent conditions of the color-naming task. This larger interference effect for the color naming trials compared to word-reading trials is consistent with task specific interference effects seen previously in cued Stroop tasks (MacDonald et al., 2000; West, 2003). It provides supporting evidence for the argument that word reading is more automatic than color naming (MacLeod & MacDonald, 2000). From these behavioral results, it follows that the color-naming, incongruent>neutral trials would produce the most robust neural interference effect. Looking at the source estimation results, we indeed see more areas of activation in IC>NC than any other contrast, particularly in the prefrontal cortex, an area of the brain very well documented as playing a prominent role in conflict processing (C. Carter & Van Veen, 2007; Larson et al., 2014; MacLeod & MacDonald, 2000; Nee et al., 2007).
Task-Specific Prefrontal Activation
In the present study, significant activation in the prefrontal cortex was observed in the color-naming trials, but not the word-reading trials. Activation in the IC>NC contrast was seen in the left premotor region (500-800ms), left DMPFC (500-900ms) and left frontopolar cortex (600-900ms). In the IC>CC contrast, we see significant activation in the left premotor cortex (900-1000ms); and left DLPFC (700-800ms). In the context of the Stroop interference effect, these regions have been implicated in selectively attending to and integrating general task information, especially when the attentional needs of the task are the greatest (Banich et al., 2000a; Bunge et al., 2002; Harrison et al., 2005; MacDonald et al., 2000; Milham et al., 2001). This directly reinforces the behavioral results that there is more of an interference effect with incongruent color naming trials than word reading trials. Because word reading is more of an automatic process than color-naming, it follows that selectively attending to the
36


color of the stimulus is more difficult and requires higher activation within the left prefrontal regions associated with directing attention to the task-relevant attributes.
In looking at the specific activated regions within the left PFC in the context of the two different interference effects in the color-naming task, we see different patterns of activation. The premotor cortex is activated in both IONC and IOCC and plays a crucial role in the selection of movements (Rushworth, Kennerley, & Walton, 2005). In the IONC contrast, we also see activation in the DMPFC and frontopolar cortex, seen previously in studies on the interference effect (Aarts et al., 2009; Purmann & Pollmann, 2015; Zysset et al., 2001). Notably, activation in these regions begins prior to the activated posterior areas of the brain. This provides supporting evidence to the top-down model of control, specifically that areas in the PFC function in selective attention and activate associated regions in the posterior regions of the brain (Banich et al., 2000b). Significant activation in the left prefrontal regions extended for hundreds of milliseconds, eventually occurring simultaneously with posterior regions. This extended activation time range points to a mechanism of sustained crosstalk between anterior and posterior attentional systems in resolving high conflict interference.
In the IC>CC contrast, besides activation in the premotor cortex, the DLPFC was activated. This supports the role of the DLPFC in executive aspects of attentional selection (Banich et al., 2009; Banich, Milham, Atchley, Cohen, Webb, Wszalek, Kramer, Liang, Wright, et al., 2000; Mansouri, Tanaka, & Buckley, 2009; Silton et al., 2010) and conflict resolution between competing stimuli attributes (C. Carter & Van Veen, 2007). While activation in the DLPFC was more temporally transient in this condition than prefrontal activation in the IC>NC contrast, examination of the stimulus-locked time series waveforms
37


comparing IC and CC activation within the DLPFC region of interest, we can see that differences between the incongruent and congruent waves begins around 500ms and continues until 1000ms (see Figure 3). Significant differences in activation, however, are only observed during the 700 to 800ms time window.
Based on prior research and the behavioral results of this study, it is not surprising that the color naming conditions saw more activation in the PFC than word-reading conditions. However, the pattern of activation within the color-naming trials differed when incongruent trials were compared to congruent versus neutral trials. The IC>NC contrast saw much stronger medial PFC activation, whereas the IC>CC contrast had lateral activation within the PFC. From the analysis for fMRI studies, activation in both lateral and medial prefrontal regions was observed in both IC>NC and IC>CC contrasts across a number of studies (Nee et al., 2007). The strong early, sustained medial prefrontal activation that we see in the IC>NC contrast in this study provides evidence that this region plays a role in the top-down biasing in times when conflict, particularly response conflict, is high.
Noticeably missing from these color-naming specific prefrontal activations is any significant activation within the anterior cingulate cortex (ACC). This is surprising as the ACC is one of the brain regions most implicated as active during resolution of the Stroop interference effect (C. Carter & Van Veen, 2007; Nee et al., 2007). This absence of activation could be related to methodological or physiological reasons. Methodologically, one of the limitations of using MEG is its hypersensitivity to superficial cortical regions and relative insensitivity to deeper cortical sources (Goldenholz et al., 2009). Another limitation to the MEG approach is that we are not directly measuring the magnetic activity within a specific region but rather estimating where we think the neural generator is most likely to be.
38


Taking this into account, it is possible that the minimum norm estimation was not able to distinguish distinct sources between the DMPFC and the ACC. The one other study that examined the Stroop effect using MEG was also unable to localize any conflict related activity within the ACC (Galer et al., 2014).
Physiologically, the ACC has been linked, not only to conflict monitoring, but also to the anticipation of high conflict stimuli (Wang, Ding, & Kluger, 2015). If we consider the experimental paradigm used in this study, each stimulus was preceded with an instructional cue that told the subject which aspect of the stimuli to attend to. When presented with the cue Color, I propose that subjects anticipated a high conflict situation, knowing that attention had to be placed on the less automatic stimulus attribute. Thus, in all color-naming trials, regardless of congruency condition, the ACC could be highly active in anticipation of a potential incongruent stimulus. In analyzing the source estimation differences between incongruent and congruent or neutral trials, no significant difference was found in the ACC because the ACC had been primed in all of these potentially high conflict trials.
Task-Specific Temporal Lobe Activation
Task-specific significant activation was also observed in the temporal lobe. In colornaming trials, the left anterior inferior temporal lobe and left occipito-temporal cortex were activated in both IC>NC and IC>CC conditions. Both of these regions fall along the ventral visual stream, a network of neural regions within the inferior temporal lobe associated with object and visual identification and recognition (Goodale & Milner, 1992). Looking first at the posterior regions, we see activation in the left occipito-temporal cortex, an area involved in lexical and semantic processing (Simons, Koutstaal, Prince, Wagner, & Schacter, 2003). A recent study examining the color-naming Stroop effect found activation in left occipito-
39


temporal cortex and particularly in the visual word form area (VWFA), a region associated with reading words (Wallentin et al., 2015). As you move anterior along the ventral visual stream, we move away from lower-level visual processing into high level conceptual associations in the anterior ventral temporal regions (Bonner & Price, 2013). Here was also see activation the color-naming incongruent trials as compared to either neutral or congruent trials.
The activation in the left inferior temporal lobe in the color naming instances of the interference effect is in line with the theory of two attentional systems working by top-down biasing. After PFC placed the attentional set on the color form of the stimuli, posterior systems were activated accordingly. Our results that the left occipito-temporal cortex and anterior temporal lobe were active in both color-naming interference effects but not in the word-reading tasks supports the idea that additional neural resources are recruited in conditions of higher conflict (i.e. color-naming over word reading). While there is an ongoing argument as to whether the conflict is modulated by increasing activation in taskrelevant or task-irrelevant posterior regions, this study does not provide definitive evidence one way or the other. In this study, areas along the ventral visual stream associated with lexical and sematic processing were most active during the color-naming interference effect. Yet that lexical and semantic processing cannot be distinguished as either task-relevant or task-irrelevant. However, the increased activation in the ventral visual stream in the color naming contrasts provides evidence that in the word-reading task, the interference effect was not as attentionally demanding and thus did not require the recruitment of additional brain resources to resolve the conflict.
40


Conflict Specific Activation
Even though the interference effect has long been a targeted area of research, distinguishing between the different types of conflict has been researched more often in recent years (Aarts et al., 2009; Milham & Banich, 2005; Szucs & Soltesz, 2010). The interference effect has been seen with both response conflict and stimulus conflict (Nee et al., 2007). In the present study, we can examine the neural correlates behind two different combinations of conflict. Incongruent>congruent is a measure of both stimulus and response conflict. The comparison of incongruent and neutral trials, however, is an examination of the contributions of response conflict to the interference effect. This is because, in the neutral conditions, the task-irrelevant attribute does not contribute information for an appropriate response and thus (Chen et al., 2011; Milham & Banich, 2005).
In examining the patterns of activation between the incongruent>neutral (I>N) and incongruent>congruent (I>C) contrasts, we see similar general patterns of activation within the various brain areas, with potentially conflict-specific variations within each area. In the color-naming tasks, both interference contrasts had prefrontal activation but the stimulus conflict (I>C) conditions saw more lateral prefrontal activation and the response conflict (I>N) condition had a much stronger medial prefrontal activation. In the occipital lobe, all interference contrasts had significant activation, but both the left VI and right V2 cortices had much stronger activation in both of the response conflict (I>N) conditions as compared to the response conflict (I>C) conditions. One possible explanation is that the incongruent and congruent conditions were similar in terms of semantics and form. Because of this, there was less of a difference in lower level visual processing of I>C than I>N. Of note, there was
41


no difference in the general pattern of temporal or parietal activation between the stimulus and response conflict comparisons.
We argue then that while incongruent>neutral and incongruent>congruent both exhibit the Stroop interference effect, they respectively follow slightly different neural paths to successfully resolve the conflict. Within the frontal and occipital lobes, we see distinct differences in strength or region of activation, but no distinct differences are observed in the temporal or parietal lobe. This points to a model of selective attentional and conflict resolution with both conflict dependent and conflict independent areas of activity.
Common Posterior Activations
In all four instances of the interference effect examined in this study, we see common activations in the occipital lobe, parietal, and temporal/parietal junction. Of particular interest is the high activation in the left temporal/parietal junction for all four contrasts. This area of the brain is an association cortex, incorporating information from a variety of subcortical and sensory modalities (R. Carter & Huettel, 2013). The left temporal/parietal junction in particular has been associated with language processing and attentional reorienting (R. Carter & Huettel, 2013). Consistently across all four contrasts, we see strong, sustained activation in the temporal/parietal junction from 500ms to 1000ms post stimulus. This provides evidence that the left temporal/parietal junction is strongly activated to help resolve the interference caused by both color-naming and word-reading incongruent stimuli. In addition, activation in the left anterior inferior parietal lobule and in the posterior superior parietal lobule, both areas activated in this study, has also been linked contributing to motor attention (Behrmann, Geng, & Shomstein, 2004; Rushworth et al., 2003).
42


When studying the interference effect elicited using a single word Stroop stimulus, these posterior areas are activated in a general, task-independent way. While the distinct regions of interest varied slightly between the four contrasts, this provides evidence that there are posterior regions of the brain that are recruited in general conflict processing in the Stroop interference effect.
Comparison to EEG findings
In examining the temporal aspect of activation observed within the four contrasts, we see differences in our activation time series as compared to those seen in previous EEG papers. Very notably, the N450, which is the medial frontal negativity that usually occurs between 350 and 500 ms post stimulus, is not found in the current study. This makes sense in light of our source estimation results, since the neural generator of the N450 is usually attributed to the ACC (Badzakova-Trajkov et al., 2009; Hanslmayr et al., 2008; Liotti et al., 2000; West, 2003). Galer et al. (2015), the only other MEG study to have examined the Stroop effect, was also lacking a distinct N450 wave that is so present in previous EEG Stroop effect studies (Donohue et al., 2016; Ergen et al., 2014; Larson et al., 2011; Szucs & Soltesz, 2010; West & Alain, 2000; West, 2003).
However, in the present study, the conflict SP, which is the centro-parietal or left temporo-parietal slow wave that lasts from 500 to up tolOOOms post stimulus is very salient. What is lacking from current EEG results concerning the conflict SP is an ability to distinguish between what specific areas of the brain are contributing to this sustained positivity. In the present study, we are able to show, using MEG, that the both prefrontal and posterior regions contribute to the conflict SP in conditions of high conflict, namely in the color-naming contrasts. Particularly, in the color-naming, incongruent>neutral contrast, we
43


see a pattern of activation within the conflict SP that supports a method of top down biasing employed to resolve the conflict. Each of the four contrasts, however, show distinct patterns of spatial and temporal activation within the conflict SP that can explain why the timing and localization of this waveform varies slightly from study to study within the EEG literature (Larson et al., 2014).
Concerning the lack of significant N450, one potential contributing factor could be the differential sensitivities that MEG and EEG have toward various regions of the brain (Goldenholz et al., 2009). MEG is most sensitive to superficial and tangentially oriented sources while EEG is more sensitive to deep, radial sources (Ahlfors et al., 2010; Goldenholz et al., 2009). Many researchers have argued that in order to optimize on the temporal and spatial resolution of EEG and MEG, they should be used in conjunction with each other (Goldenholz et al., 2009; Liu et al., 2002; Sharon et al., 2007).
Conclusion
In this study, we see the interference effect present in different task types (color-naming and word reading) and conditions of conflict (stimulus and response). Comparisons of source estimations reveal that certain posterior areas of the brain are recruited in all four variations of conflict while activation in the temporal and frontal lobes was limited to the less automatic, color-naming trials. In comparing conflict types, we see distinct differences in activation within the frontal and occipital lobes, but no distinct differences are observed in the temporal or parietal lobe. Each of the four contrasts examined in this study has its own distinct temporal and spatial pattern of neural activation. This supports a model of selective attentional and conflict resolution that recruits distinct, but parallel neural networks that are both task and conflict condition specific (Fan et al., 2003; Van Veen & Carter, 2005).
44


The color-naming, incongruent>neutral condition behaviorally displayed the strongest instance of conflict and had the most significant, sustained prefrontal activation of the four contrasts. The neural pattern of activation observed in this contrast provides evidence for a top-down biasing model in times of high conflict and attendonal demands.
Future Work
In order to strengthen the spatial and temporal understanding of the neural networks behind the Stroop interference effect, a combination of EEG and MEG should be used to localize sources. A different source estimation method, such as beamforming, should also be used to examine the neural estimated behind these four contrasts. Also, by examining the cue-locked activations in the cued Stroop task, perhaps we can get more insight into the lack of activation in the ACC in the stimulus locked.
45


CHAPTER VII
AUTHOR CONTRIBUTIONS
Conceived and designed the experiments: Benzi Kluger. Performed the experiments: Jo Shattuck. Analyzed the data: Colleen Monahan. Contributed materials/analysis tools: Colleen Monahan. Wrote the paper: Colleen Monahan.
Affirmation of Original Work
I hereby affirm that the work presented in this thesis is original except where due references are made. It does not contain any work for which a degree or diploma has been awarded by any University/Institution.
46


BIBLIOGRAPHY
Aarts, E., Roelofs, A., & van Turennout, M. (2009). Attentional control of task and response in lateral and medial frontal cortex: Brain activity and reaction time distributions. Neuropsychologia, ¥7(10), 2089-2099. http://doi.Org/10.1016/j.neuropsychologia.2009.03.019
Adleman, N. E., Menon, V., Blasey, C. M., White, C. D., Warsofsky, I. S., Glover, G. H., & Reiss, A. L. (2002). A developmental fMRI study of the Stroop color-word task. Neuroimage, 16(1), 61-75. http://doi.org/10.1006/nimg.2001.1046
Ahlfors, S. P., Han, J., Belliveau, J. W., & Hamalainen, M. S. (2010). Sensitivity of MEG and EEG to source orientation. Brain Topography, 23(3), 227-232. http ://doi .org/10.1007/s 10548-010-0154-x
Appelbaum, L., Boehler, C., Davis, L., Won, R., & Woldorff, M. (2014). The dynamics of proactive and reactive cognitive control processes in the human brain. Journal of Cognitive Neuroscience, 25(5), 1021-1038.
Badzakova-Trajkov, G., Barnett, K. J., Waldie, K. E., & Kirk, I. J. (2009). An ERP
investigation of the Stroop task: The role of the cingulate in attentional allocation and conflict resolution. Brain Research, 1253, 139-148. http://doi.Org/10.1016/j.brainres.2008.l 1.069
Banich, M., Burgess, G., Depue, B., Ruzic, L., Bidwell, L., Hitt-Laustsen, S., ... Willcutt, E. (2009). The neural basis of sustained and transient attentional control in young adults with ADHD. Neuropsychologia, ¥7(14), 3095-3104. http://doi.Org/10.1016/j.neuropsychologia.2009.07.005
Banich, M., Milham, M., Atchley, R., Cohen, N., Webb, A., Wszalek, T., ... Brown, C. (2000). Prefrontal regions play a predominant role in imposing an attentional set: Evidence from fMRI. Cognitive Brain Research, 10(1-2), 1-9. http ://doi .org/10.1016/S0926-6410(00)00015-X
Banich, M., Milham, M., Atchley, R., Cohen, N., Webb, A., Wszalek, T., ... Magin, R.
(2000) . fMri studies of Stroop tasks reveal unique roles of anterior and posterior brain systems in attentional selection. Journal of Cognitive Neuroscience, 12(6), 988-1000. http ://doi .org/10.1162/08989290051137521
Banich, M., Milham, M., Jacobson, B., Webb, A., Wszalek, T., Cohen, N., & Kramer, A.
(2001) . Attentional selection and the processing of task-irrelevant information: Insights from fMRI examinations of the stroop task. Progress in Brain Research, 134, 459-470. http://doi.org/10.1016/S0079-6123(01)34030-X
Barch, D. M., Braver, T. S., Akbudak, E., Conturo, T., Ollinger, J., & Snyder, A. (2001). Anterior cingulate cortex and response conflict: Effects of response modality and processing domain. Cerebral Cortex, 11(9), 837-848. http://doi.Org/10.1093/cercor/l 1.9.837
Barkley, G. L., & Baumgartner, C. (2003). MEG and EEG in epilepsy. Journal of Clinical Neurophysiology: Official Publication of the American Electroencephalographic Society, 20(3), 163-178. http://doi.org/10.1097/00004691-200305000-00002
Behrmann, M., Geng, J. J., & Shomstein, S. (2004). Parietal cortex and attention. Current
47


Opinion in Neurobiology, http ://doi. org/10.1016/j. conb .2004.03.012
Bonner, M. F., & Price, A. R. (2013). Where Is the Anterior Temporal Lobe and What Does It Do? Journal of Neuroscience, 33(10), 4213-4215. http://doi.org/10.1523/JNEUROSCI.0041-13.2013
Botvinick, M. M., Cohen, J. D., & Carter, C. S. (2004). Conflict monitoring and anterior cingulate cortex: An update. Trends in Cognitive Sciences. http://doi.Org/10.1016/j.tics.2004.10.003
Brass, M., Ullsperger, M., Knoesche, T. R., von Cramon, D. Y., & Phillips, N. (2005). Who comes first? The role of the prefrontal and parietal cortex in cognitive control. Journal of Cognitive Neuroscience, 17(9), 1367-1375. http://doi.org/10.1162/0898929054985400
Braver, T. S., Barch, D. M., Gray, J. R., Molfese, D. L., & Snyder, a. (2001). Anterior cingulate cortex and response conflict: effects of frequency, inhibition and errors. Cerebral Cortex (New York, N.Y. : 1991), 11(9), 825-36. http://doi.Org/10.1093/cercor/l 1.9.825
Bunge, S. A., Hazeltine, E., Scanlon, M. D., Rosen, A. C., & Gabrieli, J. D. E. (2002). Dissociable contributions of prefrontal and parietal cortices to response selection. Neuroimage, 17(3), 1562-1571. http://doi.org/10.1006/nimg.2002.1252
Carter, C., Macdonald, A., Botvinick, M., Ross, L. L., Stenger, V., Noll, D., & Cohen, J. D. (2000). Parsing executive processes: strategic vs. evaluative functions of the anterior cingulate cortex. Proceedings of the National Academy of Sciences of the United States of America, 97(4), 1944-1948. http://doi.Org/10.1073/pnas.97.4.1944
Carter, C., Mintun, M., & Cohen, J. (1995). Interference and facilitation effects during selective attention: an H2150 PET study of Stroop task performance. Neuroimage. http://doi.org/10.1006/nimg.1995.1034
Carter, C., & Van Veen, V. (2007). Anterior cingulate cortex and conflict detection: An update of theory and data. Cognitive, Affective, & Behavioral Neuroscience, 7(4), 367-379. http://doi.Org/10.3758/CABN.7.4.367
Carter, R., & Huettel, S. (2013). A nexus model of the temporal-parietal junction. Trends in Cognitive Sciences. http://doi.Org/10.1016/j.tics.2013.05.007
Chen, A., Bailey, K., Tiernan, B. N., & West, R. (2011). Neural correlates of stimulus and response interference in a 2-1 mapping stroop task. International Journal of Psychophysiology, 80(2), 129-138. http://doi.Org/10.1016/j.ijpsycho.2011.02.012
Cieslik, E. C., Mueller, V. I., Eickhoff, C. R., Langner, R., & Eickhoff, S. B. (2015). Three key regions for supervisory attendonal control: Evidence from neuroimaging metaanalyses. Neuroscience and Biobehavioral Reviews. http ://doi. org /10.1016/j. neubiorev.2014.11.003
Coderre, E., Conklin, K., & Van Heuven, W. J. B. (2011). Electrophysiological measures of conflict detection and resolution in the Stroop task. Brain Research, 1413, 51-59. http://doi.Org/10.1016/j.brainres.2011.07.017
Cohen, J. D., Dunbar, K., & McClelland, J. L. (1990). On the control of automatic processes:
48


a parallel distributed processing account of the Stroop effect. Psychological Review, 97(3), 332-61. http://doi.Org/10.1037/0033-295X.97.3.332
Corbetta, M., Miezin, F. M., Dobmeyer, S., Shulman, G. L., & Petersen, S. E. (1991).
Selective and divided attention during visual discriminations of shape, color, and speed: functional anatomy by positron emission tomography. The Journal of Neuroscience : The Official Journal of the Society for Neuroscience, 77(8), 2383-2402.
Donohue, S. E., Appelbaum, L. G., McKay, C. C., & Woldorff, M. G. (2016). The neural dynamics of stimulus and response conflict processing as a function of response complexity and task demands. Neuropsychologia, 84, 14-28. http ://doi. org /10.1016/j. neuropsychologia.2016.01.035
Donohue, S. E., Liotti, M., Perez, R., & Woldorff, M. G. (2012). Is conflict monitoring supramodal? Spatiotemporal dynamics of cognitive control processes in an auditory Stroop task. Cognitive, Affective, & Behavioral Neuroscience, 12, 1-15. http://doi.org/10.3758/sl3415-011-0060-z
Ergen, M., Saban, S., Kirmizi-Alsan, E., Uslu, A., Keskin-Ergen, Y., & Demiralp, T. (2014). Time-frequency analysis of the event-related potentials associated with the Stroop test. International Journal of Psychophysiology: Official Journal of the International Organization of Psychophysiology, 94(3), 463-472. http://doi.Org/10.1016/j.ijpsycho.2014.08.177
Fan, J., Flombaum, J. I., McCandliss, B. D., Thomas, K. M., & Posner, M. I. (2003). Cognitive and brain consequences of conflict. Neuroimage, 75(1), 42-57. http://doi.org/10.1006/nimg.2002.1319
Galer, S., Op De Beeck, M., Urbain, C., Bourguignon, M., Ligot, N., Wens, V., ... De Tiege, X. (2014). Investigating the Neural Correlates of the Stroop Effect with Magnetoencephalography. Brain Topography, 25(1), 95-103. http://doi.org/10.1007/sl0548-014-0367-5
Goldenholz, D. M., Ahlfors, S. P., Hamalainen, M. S., Sharon, D., Ishitobi, M., Vaina, L. M., & Stufflebeam, S. M. (2009). Mapping the Signal-To-Noise-Ratios of Cortical Sources in Magnetoencephalography and Electroencephalography. Human Brain Mapping, 30(4), 1077-1086. http://doi.org/10.1002/hbm.20571.Mapping
Goodale, M. A., & Milner, A. D. (1992). Separate visual pathways for perception and action. Trends in Neurosciences, 75(1), 20-25. http://doi.org/10.1016/0166-2236(92)90344-8
Hamalainen, M. (1992). Magnetoencephalography: A tool for functional brain imaging. Brain Topography, 5(2), 95-102. http://doi.org/10.1007/BF01129036
Hamalainen, M. (2010). MNE Software Users guide. MGH/HMS/MITAthinoula A. Martinos Center for Biomedical Imaging, /d(December). Retrieved from http://nmr.mgh.harvard.cdu/meg/manuals/MNE-manual-2.5.pdf
Hanslmayr, S., Pastotter, B., Bauml, K.-H., Gruber, S., Wimber, M., & Klimesch, W. (2008). The electrophysiological dynamics of interference during the Stroop task. Journal of Cognitive Neuroscience, 20(2), 215-225. http://doi.org/10.1162/jocn.2008.20020
Harrison, B., Shaw, M., Yiicel, M., Purcell, R., Brewer, W., Strother, S., ... Pantelis, C. (2005). Functional connectivity during Stroop task performance. Neuroimage, 24(1),
49


181-191. http://doi.Org/10.1016/j.neuroimage.2004.08.033
Herd, S. A., Banich, M. T., & OReilly, R. C. (2006). Neural mechanisms of cognitive control: an integrative model of stroop task performance and FMRI data. Journal of Cognitive Neuroscience, 75( 1), 22-32. http://doi.org/10.1162/089892906775250012
Kawaguchi, S., Ukai, S., Shinosaki, K., Ishii, R., Yamamoto, M., Ogawa, A., ... Takeda, M. (2005). Information Processing Flow and Neural Activations in the Dorsolateral Prefrontal Cortex in the Stroop Task in Schizophrenic Patients. Neuropsychobiology, 57(4), 191-203. http://doi.Org/http://dx.doi.org/10.1159/000085594
Kawaguchi, S., Ukai, S., Shinosaki, K., Yamamoto, M., Ishii, R., Ogawa, A., ... Takeda, M. (2004). Neuroimaging of the information processing flow in schizophrenia during the Stroop task using a spatially filtered MEG analysis. In Frontiers in Human Brain Topography (Vol. 1270, pp. 361-364). http://doi.Org/10.1016/j.ics.2004.05.050
Kerns, J. G. (2004). Anterior Cingulate Conflict Monitoring and Adjustments in Control. Science, 303(5660), 1023-1026. http://doi.org/10.1126/science.1089910
Larson, M. J., Clayson, P. E., & Clawson, A. (2014). Making sense of all the conflict: A theoretical review and critique of conflict-related ERPs. International Journal of Psychophysiology, http://doi.org/10.1016/j.ijpsycho.2014.06.007
Larson, M. J., Farrer, T. J., & Clayson, P. E. (2011). Cognitive control in mild traumatic brain injury: Conflict monitoring and conflict adaptation. International Journal of Psychophysiology, 52(1), 69-78. http://doi.Org/10.1016/j.ijpsycho.2011.02.018
Larson, M. J., Kaufman, D. a S., & Perlstein, W. M. (2009). Conflict adaptation and cognitive control adjustments following traumatic brain injury. Journal of the International Neuropsychological Society: JINS, 75(6), 927-37. http ://doi .org/10.1017/S13 5 5617709990701
Liotti, M., Woldorff, M. G., Perez, R., & Mayberg, H. S. (2000). An ERP study of the
temporal course of the Stroop color-word interference effect. Neuropsychologia, 35(5), 701-711. http://doi.org/10.1016/S0028-3932(99)00106-2
Liu, A. K., Dale, A. M., & Belliveau, J. W. (2002). Monte Carlo simulation studies of EEG and MEG localization accuracy. Human Brain Mapping, 76(1), 47-62. http://doi.org/10.1002/hbm.10024
Luck, S. (2005). An Introduction to the Event-Related Potential Technique. Cambridge, MA: The MIT Press.
MacDonald, A., Cohen, J. D., Stenger, V. a, & Carter, C. S. (2000). Dissociating the role of the dorsolateral prefrontal and anterior cingulate cortex in cognitive control. Science (New York, N.Y.), 255(5472), 1835-1838. http://doi.org/10.1126/science.288.5472.1835
MacLeod, C. M. (1991). Half a century of research on the Stroop effect: an integrative review. Psychological Bulletin, 109(2), 163-203. http://doi.org/Doi 10.1037//0033-2909.109.2.163
MacLeod, C. M., & MacDonald, P. A. (2000). Interdimensional interference in the Stroop effect: Uncovering the cognitive and neural anatomy of attention. Trends in Cognitive Sciences, 4(10), 383-391. http://doi.org/10.1016/S1364-6613(00)01530-8
50


Malmivuo, J. (2012). Comparison of the properties of EEG and MEG in detecting the electric activity of the brain. Brain Topography, 25(1), 1-19. http://doi.org/10.1007/sl0548-011-0202-1
Mansouri, F. a, Tanaka, K., & Buckley, M. J. (2009). Conflict-induced behavioural
adjustment: a clue to the executive functions of the prefrontal cortex. Nature Reviews. Neuroscience, 70(2), 141-52. http://doi.org/10.1038/nm2538
Markela-Lerenc, J., Ille, N., Kaiser, S., Fiedler, P., Mundt, C., & Weisbrod, M. (2004).
Prefrontal-cingulate activation during executive control: Which comes first? Cognitive Brain Research, 75(3), 278-287. http://doi.Org/10.1016/j.cogbrainres.2003.10.013
Milham, M., & Banich, M. (2005). Anterior cingulate cortex: An fMRI analysis of conflict specificity and functional differentiation. Human Brain Mapping, 25(3), 328-335. http://doi.org/10.1002/hbm.20110
Milham, M., Banich, M., Webb, A., Barad, V., Cohen, N., Wszalek, T., & Kramer, A.
(2001). The relative involvement of anterior cingulate and prefrontal cortex in attentional control depends on nature of conflict. Cognitive Brain Research, 72(3), 467-473. http://doi.org/10.1016/S0926-6410(01 )00076-3
Miller, E. K., & Cohen, J. D. (2001). An integrative theory of prefrontal cortex function. Annual Review of Neuroscience, 2-/(FEBRUARY 2001), 167-202. http://doi.org/10.1146/annurev.neuro.24.1.167
Minzenberg, M. J., Laird, A. R., Thelen, S., Carter, C. S., & Glahn, D. C. (2009). Metaanalysis of 41 functional neuroimaging studies of executive function in schizophrenia. Archives of General Psychiatry, 66(8), 811-822. http ://doi. org /10.1001/ archgenpsy chi atry .2009.91
Nee, D. E., Wager, T. D., & Jonides, J. (2007). Interference resolution: insights from a metaanalysis of neuroimaging tasks. Cognitive, Affective & Behavioral Neuroscience, 7(1), 1-17. http://doi.Org/10.3758/CABN.7.l.l
Nordahl, T. E., Carter, C. S., Salo, R. E., Kraft, L., Baldo, J., Salamat, S., ... Kusubov, N. (2001). Anterior cingulate metabolism correlates with Stroop errors in paranoid schizophrenia patients. Neuropsychopharmacology, 25(1), 139-148. http ://doi. org /10.1016/S0893 -13 3 X(00)0023 9-6
Oehm, C. R., Hanslmayr, S., Fell, J., Deuker, L., Kremers, N. a, Do Lam, A. T., ...
Axmacher, N. (2014). Neural communication patterns underlying conflict detection, resolution, and adaptation. The Journal of Neuroscience : The Official Journal of the Society for Neuroscience, 34(31), 10438-52. http://doi.org/10.1523/JNEUROSCI.3099-13.2014
Perlstein, W. M., Larson, M. J., Dotson, V. M., & Kelly, K. G. (2006). Temporal dissociation of components of cognitive control dysfunction in severe TBI: ERPs and the cued-Stroop task. Neuropsychologia, 44(2), 260-274. http://doi.Org/10.1016/j.neuropsychologia.2005.05.009
Peterson, B. S., Skudlarski, P., Gatenby, J. C., Zhang, H., Anderson, A. W., & Gore, J. C. (1999). An fMRI study of stroop word-color interference: Evidence for cingulate subregions subserving multiple distributed attentional systems. Biological Psychiatry,
51


Â¥5(10), 1237-1258. http://doi.org/10.1016/S0006-3223(99)00056-6
Purmann, S., & Pollmann, S. (2015). Adaptation to recent conflict in the classical color-word Stroop-task mainly involves facilitation of processing of task-relevant information. Frontiers in Human Neuroscience, 9(March), 1-11. http://doi.org/10.3389/fnhum.2015.00088
Rorden, C., Kamath, H.-O., & Bonilha, L. (2007). Improving lesion-symptom mapping. Journal of Cognitive Neuroscience, 19(1), 1081-1088. http://doi.Org/10.1162/jocn.2007.19.7.1081
Rushworth, M., Johansen-Berg, H., Gobel, S., & Devlin, J. (2003). The left parietal and premotor cortices: Motor attention and selection. In Neuroimage (Vol. 20). http://doi.Org/10.1016/j.neuroimage.2003.09.011
Rushworth, M., Kennerley, S., & Walton, M. (2005). Cognitive neuroscience: resolving conflict in and over the medial frontal cortex. Current Biology: CB, 15(2), R54-6. http://doi.Org/10.1016/j.cub.2004.12.054
Rushworth, M., Paus, T., & Sipila, P. (2001). Attention systems and the organization of the human parietal cortex. The Journal of Neuroscience : The Official Journal of the Society for Neuroscience, 27(14), 5262-71. http://doi.org/21/14/5262 [pii]
Sharon, D., Hamalainen, M. S., Tootell, R. B. H., Halgren, E., & Belliveau, J. W. (2007).
The advantage of combining MEG and EEG: Comparison to fMRI in focally stimulated visual cortex. Neuroimage, 36(4), 1225-1235. http://doi.Org/10.1016/j.neuroimage.2007.03.066
Silton, R. L., Heller, W., Towers, D. N., Engels, A. S., Spielberg, J. M., Edgar, J. C., ... Miller, G. A. (2010). The time course of activity in dorsolateral prefrontal cortex and anterior cingulate cortex during top-down attentional control. Neuroimage, 50(3), 1292-1302. http://doi.Org/10.1016/j.neuroimage.2009.12.061
Simons, J. S., Koutstaal, W., Prince, S., Wagner, A. D., & Schacter, D. L. (2003). Neural mechanisms of visual object priming: Evidence for perceptual and semantic distinctions in fusiform cortex. Neuroimage, 19(3), 613-626. http://doi.org/10.1016/S1053-8119(03)00096-X
Stroop, J. R. (1935). Studies of interference in serial verbal reactions. Journal of Experimental Psychology, 18(6), 643-662. http://doi.org/10.1037/h0054651
Szucs, D., & Soltesz, F. (2010). Stimulus and response conflict in the color-word Stroop task: A combined electro-myography and event-related potential study. Brain Research,
1325, 63-76. http://doi.Org/10.1016/j.brainres.2010.02.011
Tadel, F., Baillet, S., Mosher, J. C., Pantazis, D., & Leahy, R. M. (2011). Brainstorm: A user-friendly application for MEG/EEG analysis. Computational Intelligence and Neuroscience, 2011. http://doi.org/10.1155/2011/879716
Ukai, S., Shinosaki, K., Ishii, R., Ogawa, A., Mizuno-Matsumoto, Y., Inouye, T., ... Takeda, M. (2002). Parallel distributed processing neuroimaging in the Stroop task using spatially filtered magnetoencephalography analysis. Neuroscience Letters, 334(1), 9-12. http ://doi. org/10.1016/S03 04-3 940(02)01002-9
52


Van Veen, V., & Carter, C. (2002). The anterior cingulate as a conflict monitor: FMRI and ERP studies. Physiology and Behavior, 77(4-5), 477-482. http://doi.org/10.1016/S0031-9384(02)00930-7
Van Veen, V., & Carter, C. (2005). Separating semantic conflict and response conflict in the Stroop task: A functional MRI study. Neuroimage, 27(3), 497-504. http://doi.Org/10.1016/j.neuroimage.2005.04.042
van Veen, V., Cohen, J. D., Botvinick, M. M., Stenger, V. a, & Carter, C. S. (2001). Anterior cingulate cortex, conflict monitoring, and levels of processing. Neuroimage, 74(6),
1302-8. http://doi.org/10.1006/nimg.2001.0923
Wallentin, M., Gravholt, C. H., & Skakkebsek, A. (2015). Brocas region and Visual Word Form Area activation differ during a predictive Stroop task. Cortex, 73, 257-270. http://doi.Org/10.1016/j.cortex.2015.08.023
Wang, C., Ding, M., & Kluger, B. M. (2015). Functional roles of neural preparatory
processes in a cued Stroop task revealed by linking electrophysiology with behavioral performance. PLoS ONE, 10(1), 1-16. http://doi.org/10.1371/joumal.pone.0134686
West, R. (2003). Neural correlates of cognitive control and conflict detection in the Stroop and digit-localisation tasks. Neuropsychologia, 41, 1122-1135.
West, R., & Alain, C. (2000). Effects of task context and fluctuations of attention on neural activity supporting performance of the Stroop task. Brain Research, 573(1), 102-111. http://doi.org/10.1016/S0006-8993(00)02530-0
West, R., Bailey, K., Tieman, B. N., Boonsuk, W., & Gilbert, S. (2012). The temporal dynamics of medial and lateral frontal neural activity related to proactive cognitive control. Neuropsychologia, 50(14), 3450-3460. http ://doi. org /10.1016/j. neurop sy chol ogi a. 2012.10.011
Wheless, J. W., Castillo, E., Maggio, V., Kim, H. L., Breier, J. I., Simos, P. G., &
Papanicolaou, A. C. (2004). Magnetoencephalography (MEG) and magnetic source imaging (MSI). The Neurologist, 70(3), 138-153. http://doi.Org/10.1097/01.nrl.0000126589.21840.al
Xiao, X., Qiu, J., & Zhang, Q. (2009). The dissociation of neural circuits in a Stroop task. Neuroreport, 20(1), 674-8. http://doi.org/10.1097/WNR.0b013e32832a0al0
Zysset, S., Muller, K., Lohmann, G., & von Cramon, D. Y. (2001). Color-Word Matching Stroop Task: Separating Interference and Response Conflict. Neuroimage, 73(1), 29-36. http://doi.org/10.1006/nimg.2000.0665
Zysset, S., Schroeter, M. L., Neumann, J., & Yves von Cramon, D. (2007). Stroop interference, hemodynamic response and aging: An event-related fMRI study. Neurobiology of Aging, 28(6), 937-946. http://doi.Org/10.1016/j.neurobiolaging.2006.05.008
53


APPENDIX B
Custom Matlab Scripts
B.l. Calculate Reaction Time
% Calculate reaction time and accuracy for Transition Type, Task Type and Congruency %well as the combination of those three categories for each subject
% RTs: Differences between the stimuli and the responses D = SUB.F. events;
%Trigger 10
triglO = D(88).times; resplO = D(39).times;
S10 = D(68).times; %Switch Trigger: 10 slO = [zeros(l,(length(triglO))); resplO]; forj = l:length(S10) fori = l:length(triglO)
if trigl0(i)>S10(j) && trig 10(i)<(S 100+5.1) s 10(1 ,i)=trigl 0(i); end end end
S_RT10 = sl0(2,:) s 10(1,:); fori = l:length(S_RT10) if S_RT10(i)>3 S_RT10(i)=0; end end
S RTlOa = nonzeros(S RTlO); %Response time for trigger 10 that is a switch R10 = D(48).times; %Repeat Trigger: 10; rlO = [zeros(l,(length(triglO))); resplO]; forj = l:length(R10) fori = I:length(trigl0)
if trig 10(i)>R 10(j) && trig 10(i)<(Rl 00+5.1) rl0(l,i)=trigl0(i); end end end
R_RT10 = rl0(2,:) rl0(1,:); fori = l:length(R_RT10) if R_RT10(i)>3 R_RT10(i)=0; end end
R RTlOa = nonzeros(R_RT10); %Response time for Trigger 10 that is a repeat


SRT10 = setdiff(S_RT 1 Oa,R_RT 10a);
RRT10 = setdiff(R_RT 1 Oa, S_RT 1 Oa);
RT10 = resplO trig 10; %Response times of all Trigger 10
% Trigger 20
trig20 = D(108).times; resp20 = D(6).times;
S20 = D(77).times; %Switch Trigger:20 s20 = [zeros(l,(length(trig20))); resp20]; forj = l:length(S20) for i = 1 :length(trig20)
if trig20(i)>S20(j) && trig20(i)<(S20(j)+5.1) s20(l,i)=trig20(i); end end end
S_RT20 = s20(2,:) s20(l,:); fori = l:length(S_RT20) if S_RT20(i)>3 S_RT20(i)=0; end end
S_RT20a = nonzeros(S_RT20); %Response time for trigger 20 that is a switch R20 = D(57).times; %Repeat Trigger:20; r20 = [zeros(l,(length(trig20))); resp20]; forj = l:length(R20) for i = 1 :length(trig20)
if trig20(i)>R20(j) && trig20(i)<(R20(j)+5.1) r20(l,i)=trig20(i); end end end
R RT20 = r20(2,:) r20(l,:); for i = 1 :length(R_RT20) if R_RT20(i)>3 R_RT20(i)=0; end end
R_RT20a = nonzeros(R_RT20); %Response time for Trigger 20 that is a repeat SRT20 = setdiff(S_RT20a,R_RT20a);
RRT20 = setdiff(R_RT20a,S_RT20a);
RT20 = resp20 trig20;
% Trigger 30
trig30 = D(110).times; resp30 = D(35).times;
S30 = D(78).times; %Switch Trigger:30 s30 = [zeros(l,(length(trig30))); resp30];
58


forj = l:length(S30) fori = l:length(trig30)
if trig30(i)>S30(j) && trig30(i)<(S30(j)+5.1) s30(l,i)=trig30(i); end end end
S_RT30 = s30(2,:) s30(l,:); fori = l:length(S_RT30) if S_RT30(i)>3 S_RT30(i)=0; end end
S_RT30a = nonzeros(S_RT30); %Response time for trigger 30 that is a switch R30 = D(58).times; %Repeat Trigger:30; r30 = [zeros(l,(length(trig30))); resp30]; forj = l:length(R30) fori = l:length(trig30)
if trig30(i)>R30(j) && trig30(i)<(R300+5.1) r30(l ,i)=trig30(i); end end end
R RT30 = r30(2,:) r30(l,:); fori = l:length(R_RT30) if R_RT30(i)>3 R_RT30(i)=0; end end
R_RT30a = nonzeros(R_RT30); %Response time for Trigger 30 that is a repeat SRT30 = setdiff(S_RT3 0a,R_RT3 0a);
RRT30 = setdiff(R_RT3 0a, SRT3 0a);
RT30 = resp30 trig30;
% Trigger 40
trig40 = D(112).times; resp40 = D(43).times;
S40 = D(79).times; %Switch Trigger:40 s40 = [zeros(l,(length(trig40))); resp40]; forj = l:length(S40) for i = 1 :length(trig40)
if trig40(i)>S40(j) && trig40(i)<(S40G)+5.1) s40(l,i)=trig40(i); end end end
S_RT40 = s40(2,:) s40(l,:);
59


fori = l:length(S_RT40) if S_RT40(i)>3 S_RT40(i)=0; end end
S_RT40a = nonzeros(S_RT40); %Response time for trigger 40 that is a switch R40 = D(59).times; %Repeat Trigger:40; r40 = [zeros(l,(length(trig40))); resp40]; forj = l:length(R40) for i = 1 :length(trig40)
if trig40(i)>R40(j) && trig40(i)<(R40(j)+5.1) r40(l,i)=trig40(i); end end end
R RT40 = r40(2,:) r40(l,:); fori = l:length(R_RT40) if R_RT40(i)>3 R_RT40(i)=0; end end
R_RT40a = nonzeros(R_RT40); %Response time for Trigger 40 that is a repeat SRT40 = setdiff(S_RT40a,R_RT40a);
RRT40 = setdiff(R_RT40a,S_RT40a);
RT40 = resp40 trig40;
% Trigger 50
trig50 = D(114).times; resp50 = D(7).times;
S50 = D(80).times; %Switch Trigger:30 s50 = [zeros(l,(length(trig50))); resp50]; forj = l:length(S50) fori = l:length(trig50)
if trig50(i)>S50(j) && trig50(i)<(S50G)+5.1) s50(l,i)=trig50(i); end end end
S_RT50 = s50(2,:) s50(l,:); fori = l:length(S_RT50) if S_RT50(i)>3 S_RT50(i)=0; end end
S_RT50a = nonzeros(S_RT50); %Response time for trigger 30 that is a switch R50 = D(60).times; %Repeat Trigger:30; r50 = [zeros(l,(length(trig50))); resp50];
60


forj = l:length(R50) fori = l:length(trig50)
if trig50(i)>R50(j) && trig50(i)<(R50(j)+5.1) r50(l,i)=trig50(i); end end end
R RT50 = r50(2,:) r50(l,:); fori = l:length(R_RT50) if R_RT50(i)>3 R_RT50(i)=0; end end
R_RT50a = nonzeros(R_RT50); %Response time for Trigger 50 that is a repeat SRT50 = setdiff(S_RT50a,R_RT50a);
RRT50 = setdiff(R_RT 50a, S_RT 50a);
RT50 = resp50 trig50;
%Trigger 60
trig60 = D(116).times; resp60 = D(36).times;
S60 = D(81).times; %Switch Trigger:60 s60 = [zeros(l,(length(trig60))); resp60]; forj = l:length(S60) for i = 1 :length(trig60)
if trig60(i)>S60(j) && trig60(i)<(S60(j)+5.1) s60(l,i)=trig60(i); end end end
S_RT60 = s60(2,:)-s60(l,:); fori = l:length(S_RT60) if S_RT60(i)>3 S_RT60(i)=0; end end
S_RT60a = nonzeros(S_RT60); %Response time for trigger 60 that is a switch R60 = D(61).times; %Repeat Trigger:60; r60 = [zeros(l,(length(trig60))); resp60]; forj = l:length(R60) for i = 1 :length(trig60)
if trig60(i)>R60(j) && trig60(i)<(R60(j)+5.1) r60(l,i)=trig60(i); end end end
R RT60 = r60(2,:) r60(l,:);
61


fori = l:length(R_RT60) if R_RT60(i)>3 R_RT60(i)=0; end end
R_RT60a = nonzeros(R_RT60); %Response time for Trigger 60 that is a repeat SRT60 = setdiff(S_RT60a,R_RT60a);
RRT60 = setdiff(R_RT60a, S_RT 60a);
RT60 = resp60 trig60;
%Trigger 70
trig70 = D(118).times; resp70 = D(44).times;
S70 = D(82).times; %Switch Trigger:70 s70 = [zeros(l,(length(trig70))); resp70]; forj = l:length(S70) fori = l:length(trig70)
if trig70(i)>S70(j) && trig70(i)<(S70(j)+5.1) s70(l,i)=trig70(i); end end end
S_RT70 = s70(2,:)-s70(l,:); fori = l:length(S_RT70) if S_RT70(i)>3 S_RT70(i)=0; end end
S_RT70a = nonzeros(S_RT70); %Response time for trigger 70 that is a switch R70 = D(62).times; %Repeat Trigger:70; r70 = [zeros(l,(length(trig70))); resp70]; forj = l:length(R70) fori = l:length(trig70)
if trig70(i)>R70(j) && trig70(i)<(R70G)+5.1) r70(l ,i)=trig70(i); end end end
R RT70 = r70(2,:) r70(l,:); fori = l:length(R_RT70) if R_RT70(i)>3 R_RT70(i)=0; end end
R_RT70a = nonzeros(R_RT70); %Response time for Trigger 70 that is a repeat SRT70 = setdiff(S_RT70a,R_RT70a);
RRT70 = setdiff(R_RT70a,S_RT70a);
62


RT70 = resp70 trig70;
%Trigger 80
trig80 = D(120).times; resp80 = D(8).times;
S80 = D(83).times; %Switch Trigger:80 s80 = [zeros(l,(length(trig80))); resp80]; forj = l:length(S80) fori = l:length(trig80)
if trig80(i)>S80(j) && trig80(i)<(S80(j)+5.1) s80(l,i)=trig80(i); end end end
S_RT80 = s80(2,:)-s80(l,:); fori = l:length(S_RT80) if S_RT80(i)>3 S_RT80(i)=0; end end
S_RT80a = nonzeros(S_RT80); %Response time for trigger 80 that is a switch R80 = D(63).times; %Repeat Trigger:80; r80 = [zeros(l,(length(trig80))); resp80]; forj = l:length(R80) fori = l:length(trig80)
if trig80(i)>R80(j) && trig80(i)<(R80(j)+5.1) r80(l,i)=trig80(i); end end end
R RT80 = r80(2,:) r80(l,:); fori = l:length(R_RT80) if R_RT80(i)>3 R_RT80(i)=0; end end
R_RT80a = nonzeros(R_RT80); %Response time for Trigger 80 that is a repeat SRT80 = setdiff(S_RT80a,R_RT80a);
RRT80 = setdiff(R_RT80a,S_RT80a);
RT80 = resp80 trig80;
%Trigger 90
trig90 = D(122).times; resp90 = D(37).times;
S90 = D(84).times; %Switch Trigger:90 s90 = [zeros(l,(length(trig90))); resp90]; forj = l:length(S90) for i = 1 :length(trig90)
63


if trig90(i)>S90(j) && trig90(i)<(S90(j)+5.1) s90(l,i)=trig90(i); end end end
S_RT90 = s90(2,:)-s90(l,:); fori = l:length(S_RT90) if S_RT90(i)>3 S_RT90(i)=0; end end
S_RT90a = nonzeros(S_RT90); %Response time for trigger 90 that is a switch R90 = D(64).times; %Repeat Trigger:90; r90 = [zeros(l,(length(trig90))); resp90]; forj = l:length(R90) for i = 1 :length(trig90)
if trig90(i)>R90(j) && trig90(i)<(R90(j)+5.1) r90(l,i)=trig90(i); end end end
R RT90 = r90(2,:) r90(l,:); for i = 1 :length(R_RT90) if R_RT90(i)>3 R_RT90(i)=0; end end
R_RT90a = nonzeros(R_RT90); %Response time for Trigger 90 that is a repeat SRT90 = setdiff(S_RT90a,R_RT90a);
RRT90 = setdiff(R_RT90a,S_RT90a);
RT90 = resp90 trig90;
%Trigger 100
triglOO = D(87).times; resplOO = D(39).times;
SI00 = D(67).times; %Switch Trigger: 100 slOO = [zeros(l,(length(trigl00))); resplOO]; forj = l:length(S100) fori = l:length(trigl00)
if trig 100(i)>S 100(j) && trig 100(i)<(S 1000+5.1) s 100(1 ,i)=trig 100(i); end end end
S RT100 = si00(2,:) sl00(l,:); fori = l:length(S_RT100) if S_RT100(i)>3
64


Full Text

PAGE 1

! TASK SPECIFIC NEURAL CORRELATES OF THE CUED STROOP INTERFERENCE EFFEC T : AN MEG STUDY by COLLEEN BRIDGET MONAHAN B.S., Loyola University Maryland, 2011 M.Ed., University of Missouri St. Louis, 2013 A thesis submitted to the Faculty of the Graduate School of the University of Colorado in partial fulfillment of the requirements for the degree of Master of Science Bioengineering Program 2016

PAGE 2

"" This thesis for the Master of Science degree by Colleen Bridget Monahan has been approved for the Bioengineering Program by Richard Weir, Chair Benzi Kluger Advisor Robin Shandas July 30 2016

PAGE 3

""" Monahan, Colleen Bridget (M.S., Bioengineering) Task Specific Neural Correlates Of The Cued Stroop Interference Effect : An M EG Study Thesis directed by Associate Professor Benzi Kluger ABSTRACT The neural correlates associated with the Stroop interference effect remain a topic of debate despite being the subject of multiple neuroimaging studies over the past three decades This study aimed to spatially and temporally localize the neural networks behind the Stroop interference effect in a cued, task switching Stroop task where incongruent trials were compared to congruent or neutral trials in the context of the specific task perf ormed (color naming or word reading) This allowed a look at the influence that task type and conflict type had on the pattern of activation when interference was occurring and whether interference is resolved via a top down biasing method Comparisons of source estima tions revealed that certain occipital and parietal regions were recruited in all four variations of conflict while activation in the inferior temporal and pre frontal cortices was limited to the color naming trials. In comparing conflict ty pes (incongruent>neutral versus incongruent>congruent) we see distinct differences in activation within the pre frontal and occipital cortices but no distinct differences in the temporal or parietal regions The color naming, incongruent>neutral condition behaviorally displayed the strongest instance of conflict and had the most significant, sustained prefrontal activation of the four contrasts. The neural pattern of activation observed in this contrast provides evidence for a top down biasing model in ti mes of high conflict and attentionial demands. Each of the four contrasts examined in this study has its own distinct temporal and spatial pattern of neural activation.

PAGE 4

"# This supports a model of selective attentional and conflict resolution that recruits distinct, but parallel neural networks that are both task and conflict condition specific The form and content of this abstract are approved. I recommend its publication. Approved: Benzi Kluger

PAGE 5

# ACKNOWLEDGEMENTS This material is based upon work supp or ted by, or in part by, the U. S. Army Research Laboratory and the U. S. Army Research Office under contract number W911NF 10 1 0192 I would also like to acknowledge COMIRB for approval of this study under number 10 0539

PAGE 6

#" TABLE OF CONTENTS CHAPTER I. BACKGROUND AND SIGNIFICANCE.....1 II. REVIEW OF EXISITING LITERATURE ... .. ... .4 Introduction ...... ..4 Neuroimaging F indings ....... ..7 fMRI and PET 8 EEG 10 MEG ... ..11 Gaps in Existi ng Research...13 Conclusion.. .14 III. SPECIFIC AIMS AND HYPOTHESES .....15 IV. MATERIALS AND METHODS .17 Ethics Statement ...17 Participants... 17 Study Design: the Com puterize d Stroop Task.....17 Magnetoencephalography (MEG) Recording ..18 MEG Pre Processing ... .19 Source Localization Estimation...20 V. RESULTS ....22 Behavioral Results ... 22 Verification of the Stroop Interference E ffect with Reaction Time22 Verification of the Stroop Interference Effect with Resp onse A cc uracy..23

PAGE 7

#"" Source Space Analysis Results ....24 Color N aming: Incongruent>Neutral ..24 Color N aming: Incongruent> Congruent..25 Word Reading : Incongruent>Neutral ..25 Word Reading : Incongruent> Congruent.26 VI. DISCUSSION ......35 Behavioral Impli catio ns...35 Task Specific Prefrontal Activation ... 36 Task Specific Temporal Lobe Activation 39 Conflict Specific Activation. .... ...41 Common Posterior Activations42 Comparison to EE G Findings..43 Conclusi on...44 Future Work.45 VII. AUTHOR CONTRIBUTION S ...46 Affirmation o f Original Work .46 BIBLIOGRAPHY ....47 APPENDIX... ...54 A. Steps T ak en w ithin Brainstor m and Matlab.... ...54 A.1. Pre Processing (for Each S ubject) .54 A.2. Source Estimation (for E a ch Condition within Each S ubject) and Statistic al Tests...... .. 55 A.3. Statistical Tests W orkflow.......56

PAGE 8

#""" B. Custom Matlab Scripts.... ..57 B.1. Calculate Reaction Time... ...57 B.2. Calculate Center of Weight and Mean t Values. ... ...74

PAGE 9

$ CHAPTER I BACKGROUND AND SIGNIFICANCE Humans exhibit enormous flexibility in our ability to react and adapt to the changing environment of everyday life. We are able to take in vast amounts of visual information at any point in time and selectively place our atten tion on whatever information is relevant to what we are doing This is especially important when the object of focus has attributes that contribute conflicting information Our brains need to identify and process the conflict in order to make sense of it For the past three decades, researchers have been trying to localize the areas of our brains that work to detect and resolve visual conflict, as well as determine the chronological pattern of activation (Larson, Clayson, & Clawson, 2014; Nee, Wager, & Jonides, 2007) Research in this area is important because people with cognitive dysfunction from dem entia or brain injury can experience decline in these executive functions. In understanding the neural correlates behind these functions, researchers can begin to find solutions to delay or ameliorate the adverse effects associated with a decline in execu tive functioning. In order to better understand the neural processes associated with cognitive flexibility and con flict monitoring researchers use the Stroop task by contrasting conditions of high conflict (incongruent stimuli) with conditions of low c onflict (congruent or neutral stimuli) (MacLeod & Ma cDonald, 2000) However, inconsistencies exist among the research with regard to where these neural processes are localized and which neural processes are connected to which behaviors (Nee et al., 2007) These inconsistences most likely result from a number of methodological differences between studies that include how the neural signals were measured, what tasks were used, and which behaviors were examined.

PAGE 10

% The ability to detect and resolve the conflic t behind the Stroop effect is thought to occur via a top down biasing between the prefrontal cortex and task relevant posterior regions of the brain (Banich et al., 2000 b ; Cohen, Dunbar, & McClelland, 1990; Herd, Banich, & O'Reilly, 2006) Yet current measures of neural activity are either not spatially or not temporally resolved enough to provide meaningful results on the chrono logical patterns of activation following the onset of a conflicting, incongruent visual stimulus. The current measures of neural activity used to ident ify and localize the neural signals behind the Stroop effect are electroencephalograph (EEG), magnetoen cephalograph (MEG), and functional MRI (fMRI) Results from fMRI have high spatial resolution but low temporal resolution when compared to EEG and MEG. Both MEG and EEG have temporal resolution on the millisecond time scale but MEG has the advantage of being more spatially resolved than EEG (HŠmŠlŠinen, 1992) Thus, o f these three measures, MEG emerges as the greatest potential measure to provide localized, chronological data about brain activity associated with the Stroop eff ect. Multiple brain areas have been implicated with t he Stroop interference effect using fMRI results (Nee et al., 2007) However, these results tell us nothing about the order in which areas of the brain activate in the short time frame between the stimulus and response of the Stroop ta sk Thus fMRI studies regarding the Stroop effect of which there are dozens, can not definitively distinguish if conflict is being resolved in a t own down or bottom up manner. Contrary to fMRI, EEG studies provide information about time periods of activation but cannot localize the activation beyond general regions of the brain (i.e. centro frontal, temporo parietal). There is a gap in the current research on the Stroop interference

PAGE 11

& effect regarding whether the Stroop effect is actually resolved via top down biasing due to the lack of highly spatially, temporally resolved results. MEG studies can help fill that gap. The goal of this study is to contribute to the understanding of the neural correlates behind executive functio ns such as selective attention and conflict resolution by examining task specific neural processes measured using MEG in a cued, task switching computerized version of the Stroop task. The results of this research will add clarity to the complex processes that the brain undergoes while a person selectively attends to a visually conflicting stimulus By using MEG, this research will contribute temporal patterns of activation with high spatial localization in a way that has yet to be seen in the fMRI and EEG literature. If successful, the results of this research could incrementally alter how clinical treatment is applied to people who are experiencing cognitive decline associated aging, illness, or brain injury.

PAGE 12

' CHAPTER I I REVIEW OF EXISTING LITERATURE I nt roduction Over the past few decades many researchers in the field of cognitive neuroscience have focused their research on determining the underlying neural mechanisms behind the Stroop task (Cohen, Dunbar, & McClelland, 1990; MacLeod, 1991; Minzenberg, Laird, Thelen, Carter, & Glah n, 2009; Nee, Wager, & Jonides, 2007) To complete the Stroop task, typically, subjects are instructed to name the ink color of a list of color words, regardless of the meaning of the word. When the meaning of the word is incongruent with the ink color (e.g. "RED" written in blue ink) a phenomenon called the Stroop interference effect occurs (Stroop, 1935) The Stroop effect is the decrease in response time or increase in error rate when subjects are asked to name the ink color of incongruent stimuli as compared with congruent (e.g. "RED" written in red) or neutral stimuli (e.g. "CAR" written in red) (MacLeod & MacDonald, 2000) Over the years, t he Stroop task has been modified and used extensively as a means to test cognitive functioning including mechanisms of control, conflict monitoring, selective attention, and executi ve functioning (Cohen et al., 1990; MacLeod & MacDona ld, 2000) While t his research contributes substantially to the overall understanding of the mechanisms behind cognitive control and attention the use of different neuroimaging techniques and various versions of the Stroop task has led to mixed results on the localization and function ality of these neural processes (Nee et al., 2007) The goal of this study is to contribute to the understanding of the neural correlates behind the Stroop interference effect by examining tas k specific neural processes measured in a cued, task switching computerized version of the Stroop task. By using a cued, task

PAGE 13

( switching version of the Stroop task, a new task set is defined at the beginning of each trial (MacDonald, Cohen, Stenger, & Carter, 2000; West, 2003) For each trial, subjects are initially presented with a n instructional cue ("WORD" OR "COLOR") that indicates what attribute of the stimulus upon which to focus their selective attention. The cue is then followed by an incongruent congruent or neutral visual stimulus and subjects respond according to the cued instructions. In the original version of the Stroop task, where subjects perform the task of word reading or color naming for the whole block without switching, the interference effect is only observed in the color naming task (MacLeod & MacDonal d, 2000) Because word reading is a much more of an automatic practice when presented with the stimulus, the inter ference effect is not present during that task (Cohen et al., 1990) In the cued Stroop task, however, the interference effect is seen in both color naming and word reading tasks as measured by decrea sed reaction time during incongruent conditions as compared to the congruent or neutral conditions (MacDonald et al., 2000; Perlstein, Larson, Dotson, & Kelly, 2006; West, 2003) By having to update the task set on a trial by trial basis, the behavioral results associated with automaticity of word reading are somewhat suspended. In this way using the cued, task switching Stroop task allows the researcher to test the influence of the task type on the interference task. When presented with the stimulus in the Stroop task the subject must attend to one attribute of the stimulus over the other. This creates conflict when the different attributes of the stimuli are incongruent and when tasked with identifying the less automatic attribute (MacLeod & MacDonald, 2000) It has been posit ed that there are two distinct attentional systems within the brain that are recruited in the conflict processing and resolution of the Stroop task through top down processing (Banich et al., 2000 b ; Fan, Flombaum McCandliss,

PAGE 14

) Thomas, & Posner, 2003; Miller & Cohen, 2001) The executive attentional functioning system distinguishes the aspect of the stimuli that is task relevant and requiring attention This executive functioning is found in the anterior part of the brain but communicates closely with the regulatory system of attention located in the posterior of the brain. These anterior, prefrontal regions impose the attentional set by "alerting" or priming the pos terior brain area s associated with processing t he task relevant information about the stimulus (Banich et al., 2000 b ) Here, brain areas are activated based on the perceived stimulus attrib utes. In trials where color naming is the task, brain areas assoc iated with color processing are more active (Corbetta, Miezin, Dobmeyer, Shulman, & Petersen, 1991; MacLeod & MacDonald, 2000) In the Stroop effect attentional demands are higher during the incongruent conditions compared to the congruent trials due to higher conflict between stimulus attributes. Banich et al. (2001) proposes that the posterior system regulates attentional resources by processing info rmation associated with the tas k irrelevant attribute of the stimuli (the ink color in a word rea ding task). Indeed, the visual word form area (VFWA) was found to be more activated in a color naming interference effect (Wallentin, Gravholt, & Skakkebk, 2015) Interestingly, a decre a se in the interference effect vi a a daptation to conflict may be modulated by increasing activity in task relevant posterior brain regions (Purmann & Pollmann, 2015) Contrary to this, other research ha s shown an increase in activity in task relevant areas and decrease in task irrelevant areas (Harrison et al., 2005) Thus clarification needs to be added to the existing literature on how the anterior and posterior attentional systems modulate conflict in a task specific way. In trying to understand how activation in di fferent neural areas contributes to the interference effect, it is important to look at how the Stroop effect is measured within the

PAGE 15

* study design. In the literature, the Stroop effect has been achieved by comparing incongruent trials to either neutral or congruent conditions (Nee et al., 2007) Comparing incongruent and congruent conditions has the advantage that physical characteristics and cognitive processing (lexical and semantic) of the stimuli are the sam e, preventing unanticipated neural activation unrelated to the control and attentional processing of the interference effect (Milham & Banich, 2005; Peterson et al., 1999) This comparison, however, confounds the effect of stimulus encoding conflict with response related conflict (Nee et al., 2007) Conversely, comparing incongruent and neutral conditions gives the advantage of being able to see t he effect that response conflict has on the interference effect (Aarts, Roelofs, & van Turennout, 2009; Zysset, MŸller, Lohmann, & von Cramon, 2001) For the color naming task, the neutral stimulus is a non color word or non word (e .g. XXXX) presented in colored ink I n the word reading task, the neutral stimulus is a color word presented in black ink. In both of these conditions, the task relevant attribute is the only aspect of the stimulus that can inform a response decision. Th is allows researchers to investigate the neural correlates of the response conflict that exists in the incongruent condition. However, because physical and cognitive attributes of the stimuli are not the same, additional neural activation may be included in the analysis of the incongruent>neutral c omparison (Peterson et al., 1999) By making multiple comparisons between different experimental conditions with the Stroop task, the nuanc es associated with the conflict and attentional processing of the interference effect can be better understood. Neuroimaging F indings As previously mentioned, there have been numerous neuroimaging studies dedicated to the uncovering and understanding of t he neural activations behind the Stroop interference

PAGE 16

+ effect. There are several areas of brain activation during the Stroop effect that are widely accepted as contributing to attentional or control systems necessary to process the incongruent stimuli Howe ver, the manipulation of study design and the use of different imaging techniques has led to an ongoing debate on the exact spatial and temporal distribution of brain areas of interest (C. Carter & Van Veen, 2007; Larson et al., 2014; Nee et al., 2007) fMRI and PET Results from fMRI and a few positron emission topography (PET) studies have made use of their high spatial resolution to identify areas associated with various cognitive process es The anterior cingulate cortex (ACC), left pre frontal cortex (PFC), and various posterior regions ha ve been repeatedly f ound in comparing incongruent to congruent or neutral trials in the Stroop task (Nee et al., 2007) The anterior cingulate cortex (ACC) has been prop o sed to play a variety of roles in the cognitive processing of the interference effect. In several studies, the ACC has been shown to a hav e a executive role in detecting and reducing conflict in a top down manner (C. Carter & Van Veen, 2007; C. Carter et al., 2000; Nee et al., 2007; Silton et al., 2010; Van Veen & Carter, 2002) and contributing more generally to performance and conflict monitoring (Adleman et al., 2002; Barch et al., 2001; Botvinick, Cohen, & Carter, 2004; Braver, Barch, G ray, Molfese, & Snyder, 2001; Fan et al., 2003; Hanslmayr et al., 2008; Kerns, 2004; MacDonald et al., 2000; Nordahl et al., 2001; Van Veen & Carter, 2002) Aside from conflict monitoring, the ACC functions in executive attentional control with respect to response conflict alone (Milham et al., 2001; van Veen, Cohen, Botvinick, Stenger, & Carter, 2001) with respect to both task and response conflict (Aarts et al., 2009) and with respect to

PAGE 17

, coor dinating the activity of multiple attentional subsystems (Cieslik, Mueller, Eickhoff, Langne r, & Eickhoff, 2015; Peterson et al., 1999) Other areas of the prefrontal cortex (PFC) has also been implicated in task preparation and establishing the attentional set to the prese n t task during the Stroop effect (Banich et al., 2000 a ; Brass, Ullsperger, Knoesche, von Cramon, & Phillips, 2005; C. Carter & Van Veen, 2007; Harri son et al., 2005; MacDonald et al., 2000; Milham et al., 2001) In accordance with the idea of parallel, coordinated anterior and posterior attentional systems, evidence supports the role of the prefrontal cortex in integrating general task information and informing releva nt regions within the parietal cortex (Banich et al., 2001; Banich et al., 2000 b ; Brass et al., 2005) Particular areas of activation within the PFC during the Stroop effect include the l eft PFC (Bunge, Hazeltine, Scanlon, Rosen, & Gabrieli, 2002; Harrison et al., 2005; Milham et al., 2001; Van Veen & Carter, 2005; Zysset et al., 2001) left dorso lateral PFC (DLPFC) (Aarts et al., 2009; Banich et al., 2000 b ; C. Carter & Van Veen, 2007; MacLeod & MacDonald, 2000; Nee et al., 2007; Silton et al., 2010) ventro lateral PFC (VLPFC) (Aarts et al., 2009; Purmann & Pollmann, 2015) right lateral PFC (Milham et al., 2001) medial PFC (Oehrn et al., 2014; Purmann & Pol lmann, 2015) and pre motor cortex (Zysset, Schroeter, Neumann, & Yves von Cramon, 2007) The fMRI research on activation within the parietal lobe has been more varied in terms of spatial distribution as compared to ACC and PFC activ ation during the Stroop effect Specific areas of activation noted in the research include the p osterior parietal cortex (Adleman et al., 2002; Nee et al., 2007; Purmann & Pollmann, 2015; Rushworth, Johansen Berg, Gobel, & Devlin, 2003) inferior parietal cortex (C. Carter, Mintun, & Cohen, 1995; Rushworth, Paus, & Sipila, 2001) l eft parietal occipit al lobe (Adleman et al., 2002)

PAGE 18

$! intraparietal sulcus (Zysset et al., 2001) and general parietal cortex (Van Veen & Carter, 2005; Wallentin et al., 2015) The parietal cortex, particularly on the left side, is thought to play a role in processing the physical attributes of both task relevant and task irrelevant information in order to inform output res ponses (Banich et al., 2000 b ; Herd, Banich, & O'Reilly, 2006) Worth noting are studies that found additional activation in the occipital lobe (Adleman et al., 2002) occipito temporal cortex (Zysset et al., 2001) and VWFA within the temporal cortex (Wallentin et al., 2015) in the color naming Stroop effect. EEG Complementary to the spatially resolved fMRI research are scalp encephalography (EEG) studies that take advantage of their high temporal resolution to describe chronological activation during the Stroop effect. EEG non invasively measure s electrical activity of the brain using electrodes that are placed evenly over the subject's skull (Luck, 2005) Rooted within this electrical activity are neural responses called event related pote ntials (ERPs) that are associated with specific sensory, cognitive, and motor events (Barkley & Baumgartner, 2003; Malmivuo, 2 012) Two main ERPs associated with the interference effect have been consistently reported in the literature (Liotti, Woldorff, Perez, & Mayberg, 2000; West, 2003) The first is a medial dorsal or fronto central negativity, known as the medial frontal negativity (MFN), N450 or N INC that is characterized as occurring 350 to 500 ms post stim ulus (Donohue, Appelbaum, McKay, & Woldorff, 2016; Larson, Farrer, & Clayson, 2011; Liott i et al., 2000) This wave is thought to have its neural origins within the ACC (Badzakova Trajkov, Barnett, Waldie, & Kirk, 2009; Hanslmayr et al., 2008; Larson et al., 2014; Liotti et al., 2000; Szucs & Soltesz, 2010; Van Veen & Carter, 2002; W est & Alain, 2000) but a few sources have

PAGE 19

$$ found neural correlates elsewhere such as in the PFC (Markela Lerenc et al., 2004) and posterior cingulate cortex (West, Bailey, Tiernan, Boonsuk, & Gilbert, 2012; Xiao, Qiu, & Zhang, 2009) The cognitive functioning of the MFN is thought to be in conflict detection and monitoring (Appelbaum, Boehler, Davis, Won, & Woldorff, 2014; Coderre, Conklin, & Van Heuven, 2011; Larson et al., 2014; Szucs & Soltesz, 2010; Van Veen & Carter, 2002; West & Alain, 2000; West et al., 2012) The second ERP wave is described as a centro parietal positive slow wave known as the conflict slow potential (conflict SP) (Larson et al., 2014, 2011; Larson, Kaufman, & Perlstein, 2009) but is also referred to the late positive component (LPC) (Appelbaum et al., 2014; Donohue et al., 2016) The conflict SP develops approximately between 500ms and 800 ms post stimulus but has been shown to extend past 1000ms (West & Alain, 2000) Activation has been documented over centro parietal (Chen, Bailey, Tiernan, & West, 2011; Larson et al., 2011; West et al., 2012; West, 2003) as well as left temporo parietal scalp regions (Donohue Liotti, Perez, & Woldorff, 2012; Liotti et al., 2000; West & Alain, 2000) Source reconstruction techniques have identified its neural generator as the lateral frontal and posterior parietal cortex (Chen et al., 2011; West, 2003) It has been suggested that the conflict SP functions in conflict adjustment and resolution mechanisms (Appelbaum et al., 2014; Chen et al., 2011; Larson et al., 2014, 2011, 2009; W est & Alain, 2000) MEG The majority of research that focuses on explaining the neural mechanisms behind Stroop interference effect uses fMRI or EEG data. However, for the past two decades, magnetoencephalography (MEG) has been used as a method of nonin vasive functional imaging of the brain (Malmivuo, 2012; Wheless et al., 2004) Applying this measure of

PAGE 20

$% neural activity to the Stroop task paradigm has very rarely been seen in past research and would serve to add clarity to the pool of existing fMRI and EEG results on this subject (Galer et al., 2014; Ukai et al., 2002) Whenever an ERP is generated, a magnetic field, called an event related field (ERF) is also g enerated around the ERP dipole (Luck, 2005) MEG measure s the s e magnetic signal s generated by pyramidal cells that largely lie orthogonal to the cortical surface with m illisecond temporal resolution (HŠmŠlŠinen, 1992) The most important benefit of ERFs is that these magnetic fields are not blurred by contact with the skull (Barkley & Baumgartner, 2003; HŠmŠlŠinen, 1992) When EEG records electrical voltage at any given moment from electrodes on the skull, it reflects the combination of ERPs g enerated from several different sources throughout the brain. E lectric potentials tend to spread out laterally when they encounter the skull and thus a signal recorded by an EEG electrode reflects a skewed picture of the origin of those potentials (Luck, 2005) Therefore it is believed by many researchers that the spatial resolution of MEG is superior to that of EEG and can be expected to identify activated sources within the millimeter to centimeter range (Barkley & Baumgartner, 2003; Wheless et al., 2004) It is also worth noting that MEG and EEG are not spatially sensitive to the same neural sources (Ahlfors, Han, Belliveau, & HŠmŠlŠinen, 2010; Sharon, HŠmŠlŠinen, Tootell, Halgren, & Belliveau, 2007) MEG is highly sensiti ve to tangential sources close to the surface of the skull, while EEG can detect much deeper sources (Ahlfors et al., 2010; HŠmŠlŠinen, 1992) While several studies exist that measure Stroop effect related neural activity with EEG, there are only four studies that use the Stroop task with MEG (Galer et al., 2014; Kawaguchi et al., 2004, 2005; Ukai et al., 2002) In three of these studies, the Stroop task

PAGE 21

$& was used to locate the neural origins of the incongruent conditions without any comparison to congruen t or neutral conditions (Kawaguchi et al., 2004, 2005; Ukai et al., 2002) Therefore the researchers could only report on the brain activation associated with information processing during the Stroop task, rather than the Stroop effect. In the final study, Galer et al. (2015) used the Stroop test to look at interference related neural events by comparing incongruent with congruent trials. Using minimum norm source estimation, Galer et al. (2015) found higher acti vation in the left pre supplementary motor and left posterior parietal cortex in the incongruent trials from 480 to 700ms post stimulus (Galer et al., 2014) Gaps in Existing R esearch The Stroop interference effect h as been documented in many different conditions of conflict. Yet questions still remain about the extent to which the ACC, PFC, parietal and other posterior cortices chronologically contribute to various instances of the interference effect. Findings from fMRI studies provide evidence that areas within the PFC function in selective attention and conflict monitoring (Nee et al., 2007) Yet due to the poor temporal resolution of fMRI, the extent to which these structures function via top down biasing or sustained feedback loop s with posterior brain regions cannot be determined. On the other han d, the poor spatial resolution of EEG results only allows for general brain regions to be identified as active at any point in time post stimulus (Larson et al., 2014) There is a n eed for spatially resolved studies that show the patterns of neural activation over time. Studies using MEG can help fill that gap. The majority of studies that examine the Stroop interference effect only examine the neural correlates associated with one contrast (i.e. incongruent>neutral for color naming) (Nee et al., 2007) Yet the interference effect has been associated with both stimulus and

PAGE 22

$' response conflict within different types of tasks (Aarts et al., 2009; Milham & Banich, 2005; Nee et al., 2007) There is a need for interference related studies to examine multiple conflict contrasts within the same study design and subject population. This a llows for the critical, meaningful examination of differences in patterns of neural activation for different conflict types that result behaviorally in the interference effect. Conclusion The primary goal of the present study is to use an alternative me asure of electrophysi ologi ca l activity, MEG, to add spatial and temporal clarity to the roles that various neural areas play in different conditions of cognitive conflict. In comparing incongruent to congruent stimuli, neural networks contributing to conf lict conditions where both stimulus and response conflict are present can be explored The comparison of incongruent to neutral stimuli allows the role of response conflict to become salient within the interference task. Both of these interference conditions are examined in the context of their task type (color naming and word reading) to further determine if the task type influences the conflict and attentional pathways recruited during the different presentations of the Stroop effect. By using MEG neural generators of the Stroop effect can be explored with millisecond temporal resolution. In this way, the present study will report on the neural correlates of various conditions of conflict with temporal and spatial pre cision in a manner that is yet to be seen in the literature.

PAGE 23

$( CHAPTER III SPECIFIC AIMS AND HYPOTHESES Specific Aim 1: Localize the neural generators and describe the time course of activation behind the interference effect with respect to the task type (co lor naming versus word reading) using MEG as the measure of neural activity. Hypothesis 1: T h ere will be more significant activation in the prefrontal cortex in the color naming contrasts as compared to the word reading contrasts due to the automacity associated with word reading (Mac Leod & MacDonald, 2000) Posterior regions will have task specific patterns of activation (Corbetta et al., 1991; MacLeod & MacDonald, 2000) Activation in the prefronta l areas will initially precede activation in posterior regions in line with a top down biasing model but sustained activation in both regions will be seen from 500 1000ms post stimulus (Miller & Cohen, 2001) Specific Aim 2: Localize the neural generators and describe the time course of activation behind the interfere nce effect with respect to the type of conflict ( incongruent>neutral (I>N) versus incongruent>congruent (I>C)) using MEG as the measure of neural activity. Hypothesis 2: Both conflict types (incongruent>neutral and incongruent>congruent) will ha ve significant activation in the prefrontal cortex (Nee et al., 2007) I>N will have greater activation in the left DLPFC and I>C will have greater activation in the ACC (Nee et al., 2007) Posterior regions will have condition specific patterns of activation. Activation in the prefrontal areas will initially precede act ivation in posterior regions, but sustained activation in both regions will be seen from 500 1000ms post stimulus (Miller & Cohen, 2001)

PAGE 24

$) Specific Aim 3: Localize the neural generators and describe the time course of activation behind the interfere nce effect with respect to general conflict processing (i.e. task independent, conflict type independent) using MEG as the measure of neural activity. Hypothesis 3: There will be common areas of activation in all four contrasts in both the prefro ntal cortex (ACC, DLPFC) and posterior parietal cortex (Nee et al., 2007) The timing of the activation will reflect top down biasing. Significant activation in the ACC will occur first, between 350 and 500ms post stimulus. Sustained a ctivation in other regions of the PFC and posterior regions will occur between 500 and 1000ms post stimulus (Mil ler & Cohen, 2001)

PAGE 25

! CHAPTER I V MATERIALS AND METHODS Ethics Statement COMIRB approved this study protocol ( Number 10 0539 ) and all participants sig ned informed consent statements. Participants Brain activity of thirty participants was recorded using simul taneous MEG and EEG. Data from one participant was excluded from this study due to excessive motion artifact. Of t he twenty nine remai ning participant s, there were 18 females and 11 males ranging in age from 19 to 36 years old. All participants wer e right handed. Prior to the recordings, participants were instructed to get a full night of sleep and to refrain from consuming caffeine and nicotine in order to reduce cognitive changes not associated with the target testi ng. All recordings started at 10 :00 am. Study Design: Computerized Stroop Test Participants performed a computerized c ued version of the Stroop test originally developed by Cohen et al. (199 9 ), in an acoustically and magnetically sealed room. Each trial of the Stroop test consisted of three parts: the presentation of a cue (either "word" or "color") a 1 or 5 second cue stimulus interval, and a visual word stimulus ("green", "red", or "blue"). Initial cue s were presented as auditory instructions through a loudspeaker and subsequent stimuli were presented visually on an LCD screen. The stimulus features (word color and word meaning) were congruent incongruent, or neutral Congruent stimuli wer e color w ords printed in the same color (i.e. "BLUE" printed in blue) and incongruent stimuli were color words printed in a different color (i.e. "BLUE" printed in green). Neutral stimuli

PAGE 26

! $+ were non color words printed in red, blue, green for the color namin g task (i.e. "CAR" printed in red) and color words printed in black for the word reading tasks (i.e. "RED" printed in black). On trials where the cue was "word", participants stated the word meaning. On trials where the cue was "color", participa nts stat ed the ink color of the word The trials were evenly divided based on task type (word reading or color naming) transition type (switch or repeat) cue stimulus interval (1 or 5 seconds) and congruency of stimuli (congruent, incongruent, or neutral) Participants responded to each trial stimulus by pressing one of three buttons on a keypad After five minutes of practice trials, participants performed this experiment for a single three hour session. This analysis only looks at the first 360 trials Magnetoe nchephalography (MEG) R ecording Magnetic evok ed fields were recorded using a Magnes 3600 WH whole head MEG device (4 D Neuroimaging, San Diego, CA, USA), which comprises 248 first order axial Figure 1 The Stroop task paradigm consisting of three parts: task specific auditory cue, followed by the visual stimulus to which subjects responded with a keypress. The time period of interest in this study was 200ms to 1000 with 0ms indicating s timulus onset. Reaction time was calculated between the stimulus onset and time of appropriate keypress.

PAGE 27

! $, gradiometer sensors in a helmet shaped array. Data was collected at a sampling rate of 291 Hz Each participant's head position was determined with respect to the sensor array using f ive head position indicator coils attached to the subject's scalp The locations of the coils with respect to three anatomical landmarks (nasion and pre auricular points) and two extra non fiducial points, as well as th e scalp surface were determined with a 3D digitizer (Polhemus, Colchester, VT, USA). MEG Pre P rocessing Pre processing of the data was performed wi th Brainstorm (Tadel, Baillet, Mosher, Pantazis, & Leahy, 2011) which is documented and freely available for download online under the GNU general public license ( http://neuroimage.usc.edu/bra instorm ). Custom scripts written in Matlab r2104a ( The MathWorks, Inc., Natick, M A USA) were also employed. The continuous data was initially visually inspected in order to reject time segments with poo r signal quality and excessive noise. The raw data set was band pass filtered between 0.1 Hz and 30 Hz. Unwanted artifacts including eye blinks and cardiac signal were removed using SSP (Signal Space Projection) functions within Brainstorm. Epochs were set up spanning 200ms to 1000 ms relative to cue on set (0ms) and o nly those trials with correct responses were included in the subsequent analysis. Epochs were baseline corrected to remove any DC offset and tria ls with excessive amplitude (>45 00fT) were rejected. Baseline was defined as 200 to 5ms relative to stimulus onset For each subject, trials within each experimental condition were then averaged. Pseudo individual brain anatomies were created for each subject by warping the default anatomy provided by Brainstorm to fit the digitized head po ints collected with a magnetic tracking system prior to MEG data acquisition.

PAGE 28

! %! Source L ocalization Estimation In order to spatially estimate the neural activity measured by the MEG sensors, two separate modeling steps were taken. First the head model, or forward model, was constructed within Brainstorm using overlapping spheres for each sensor to model the tissue layers of the brain in relation to the sensor instrumentation. This allowed the magnetic fields measured at the sensors to be co nnected to their electric generators within the brain. Second, the ill posed inverse model was addressed. This step of modeling involves estimating the activity of thousands of cortical dipoles modeled with the head model with the activation from 248 MEG channels. The challenge lies with the possibility of an infinite number of combinations causing the pattern of activatio n measured at the sensor level. The inverse model was applied via established minimum norm source estimation techniques (HŠmŠlŠinen, 2010; Liu, Dale, & Belliveau, 2002) Within Brainstorm, a whitened and depth weighted linear L2 minimum norm estimation algorithm was implemented (HŠmŠlŠinen, 2010) This estimation was calculated assuming one dipole at each vertex of the cortical surface, oriented normal to the surface. These estimates produce a dens ity map of flowing current at the surface of the cortex After sources were reconstructed separately for each experimental condition within each subject, source images we re spatially smoothed using a G a u ssian kernel of 12mm FWHM. Paired t tests were cond ucted on the following sets of contrasts : in congruent, color>congruent, color (IC>CC) ; incongruent, color>incongruent, neutral (IC>IN) ; incongruent, word>congruent, word (IW>CW) ; incongruent, word>neutral word (IW>NW) Each paired t test resulted in a map of voxel values consisting of t values. For each of the four contrast s t maps were generated for 100ms averaged time windows, ranging from 0 to

PAGE 29

! %$ 1000 ms post stimulus. Areas of significant difference were reported at p <0.01 cor rected for false discovery rate (FDR). On those maps displaying significant difference, regions of interest were defined based on local maximum t values in areas of significance of at least 40 voxels For each regions of interest the center o f weight and mean t value were calcula ted using custom Matlab scripts Spatial lo cations of ROIs were defined by Brodman's Area using MRIcron software (Rorden, Karnath, & Bonilha, 2007)

PAGE 30

! %% CHAPTER V RESULTS Beh avioral R esults Verification of the St roop Interference E ffect with Reaction Time (RT) In the color naming condition, there was a significant difference in mean reactions time between incongruent (M= 1.18, SD = 0.33 ) and neutral (M= .85 SD = 0. 29 ) conditions ; t(28) = 13.4, p< 0.001 and the incongruent and congruent (M= .91, SD = 0.29 ) condit ions; t(28) = 8.2 p< 0.001 (see Table 1) In the word reading condition, there was also a significant difference in mean reactions time between incongruent (M= 1.18, SD = 0.34) and neutral (M= .87, SD = 0.26) conditions; t(28) = 14.7, p< 0.001 and the inc ongruent and congruent (M= .97, SD = 0.32) conditions; t(28) = 10.0 p< 0.001 This verifies that during both the color naming and word reading tasks, an interference effect was experienced in the form of decreased reaction time during the incongruent sti mulus. Mean reaction time was measured on trials answered correctly. Aside from measuring the traditional interference effect, it is worth noting that in both the color naming and word reading conditions, neutral tasks were performed significantly faster than congruent tasks; t(28) = 7.6, p< 0.001 for word Table 1 Mean Percent Accuracy Task Type Congruency Condition Color N aming Word R eading Incongruent 86.5% 91.3% Congruent 95.3% 95.7% Neutral 95.3% 94.6%

PAGE 31

! %& reading; t(28) = 4.4, p<0. 001 for color naming. This indicates that more conflict was experienced during the congruent tasks than the neutral tasks, regardless of the stimulus attribute to which attentio n was given (color or word). Verification of the Stroop I nterferenc e E ffect with Response A ccuracy In the color naming condition, a significant difference in response accuracy was found between incongruent (M= 0.87 SD = 0.14) and neutral conditions (M= .95, SD = 0.06 ) ; t(28) = 4.41 p< 0 .00 1 and the incongruent and congruent (M= .95 SD = 0.05 ) conditions; t(28) = 4.14 p< 0.00 1 (see Table 2) In the word reading condition, a paired t test between incongruent (M= 0.91, SD = 0.08) and neutral (M= 0.95, SD = 0.07) conditions had a t(28) = 5.11 p< 0.00 1 and the incongruent and congruent (M= 0.96, SD = 0.07) conditions had t(28) = 5.83 p< 0.001 This shows that in all four contrasts, the interference effect was seen in how accurate subjects were in completing the task at hand It is also worth noting that there is a significant difference in response accuracy between incongruent conditions of the color naming task and incongruent conditions of the word reading task; t(28) =2.6, p< 0 .05 This demonstrates that on a behavioral level, there is a higher level of accuracy in responding to the word reading task s over color naming in the incongruent Table 2 Mean Reaction Times and Standard Deviations (in seconds) Task Type Congruency Condition Color N aming Word R eading Incongruent 1.18 (0.33) 1.18 (0.34) Congruent 0.91 (0.29) 0.97 (0.32) Neutral 0.85 (0.29) 0.87 (0.26)

PAGE 32

! %' conditions. This provides evidence for the argument that, even within the cued version of the Stroop task, word reading is more automatic than color naming (MacLeod & MacDonald, 2000) Source Space Analysis Results For each of the four contrasts ( IC>CC IC>NC IW>C W, IW>NW), ten paired t tests were run using the minimum norm estimate source images on 100ms time averaged window s ranging from 0 to 1000ms post stimulus. Areas of significant difference were reported at p <0.01 corrected for false discovery rate (FDR). For all four contrasts, no significance was found from 0 to 500ms. For IC>NC and IW>CW, significant difference was detected on the five t value source maps generated from 500ms to 1000ms. For IC>CC and IW>NW, significant difference was detected on the four t value source maps generated from 600ms to 1000ms. On each t value map displaying significant difference, regions of interest (ROIs) were defined based on local maximum t values in areas of significance of at least 40 voxels. Color N aming: Incongruent>N eutral The two contrasts where color naming was performed resulted in more significantly activated regions of interests than the two contrasts of the word reading task. For the IC>NC contrast, ten ROIs were identified over the 500 to 1000ms windows. These areas include left frontopolar, dorso medial prefrontal (DMPFC) and premotor regions, left and right occipital cortices, and left occipito temporal, temporal, and parietal regions. Activation in the fronta l regions was strongest on the front end of the time series (500 700ms) and activation in the posterior areas of the brain (parietal, occipital, temporal) was strongest between 700 and 900ms post stimulus. In T able 3, each significant ROIs for the IC>NC c ontrast are listed by anatomical and functional region. For each 100ms time averaged window, the center of

PAGE 33

! %( weight and mean t value are recorded for those areas where significant difference was detected within that time window Figure 2 gives a visual rep resentation of the data described in Table 3 C olor N aming: Incongruent>Congruent I n the IC>C C Stroop effect, activation was seen in eight distinct regions over the 600 to 1000ms time windows These ROIs are focused in the left hemisphere in the dorsolateral prefrontal cortex ( DLPFC ) premotor cortex, inferior temporal gyrus, occipito temporal cortex tempo ral/ parietal junction and various occipital regions. Activation is also seen in the right posterior parietal cortex. Nota bly, throughout the t ime series (600 1000ms), strong activation was recorded in the i nferior temporal gyrus and tempo ral/parietal junction, areas highly associated with language (Bonner & Price, 2013; R. Carter & Huettel, 2013) Activation in the frontal cortex is more transient with significant activation limited to 100ms time blocks for the DLPFC and premotor areas. Occipital regions are most active later in the time series (700 1000ms). In Table 4, the significant ROIs for the IC>CC contrast are listed by anatomical and functional region. For each 100ms time averaged window, the center of weight and mean t value are recorded for those areas where significant difference was detected within that time window. Figure 3 gives a visual representation of the data described in Table 4 Word R eading: Incongruent>Neutral In the IW>NW Stroop effect, activation w as seen in five areas over the 6 00ms to 1000ms time windows. ROIs were restricted to the occipital and parietal lobes in both the right and left hemispheres. Unlike the color naming contrasts, no activation was seen in the temporal or frontal lobes. Activation across the tim e series was strongest in the left occipital

PAGE 34

! %) and parietal cortices. In Table 5 the significant ROIs for the IW>NW contrast are listed by anatomical and functional region. For each 100ms time averaged window, the center of weight and mean t value are rec orded for those areas where significant difference was detected wi thin that time window. Figure 4 gives a visual representation of the data described in Table 5 Word R eading: Incongruent>Congruent In the IW>CW Stroop eff ect, activation was seen in fo ur areas over the 500ms to 1000ms time windows. ROIs were identified in the left primary and association visual cortices and right secondary visual cortex. The strongest and most sustained activ ation was recorded in the tempor al/parietal junction. In Tabl e 6 t he significant ROIs for the IW>C W contrast are listed by anatomical and functional region. For each 100ms time averaged window, the center of weight and mean t value are recorded for those areas where significant difference was detected wi thin that time window. Figure 5 gives a visual representation of the data described in Table 6

PAGE 35

! %* Figure 2 On the left, v isual maps of posterior and left sagittal views of significant t values with p<0.01 FDR for IC>NC contrast in 100ms time averaged wind ows from 500ms to 1000ms post stimulus. On the right, t he current density, time series activation for three regions of i nterest are displayed. The x axis is time in ms, spanning the 200 to 1000ms time window, with 0ms as the stimulus onset. The y axis i s the current density amplitude in pA.m.

PAGE 36

! %+ Table 3 Color Naming: Incongruent>Neutral Regions of Interest (ROIs) 500 600ms 600 700ms 700 800ms 800 900ms 900 1000ms Functional Region (Brodman's Area) MNI t MNI t MNI t MNI t MNI t L Premotor (6) [ 29, 3, 53] 4.5 [ 28, 3, 53] 3.6 [ 28, 3, 54] 3.5 L DMPFC (8/9) [ 8, 39, 49] 5.2 [ 9, 38, 49] 5.1 [ 10, 38, 48] 3.3 [ 10, 39, 48] 3.5 L Wernicke's Area (21/22) [ 60, 30, 3] 3.6 L Frontopolar (10) [ 5, 58, 4] 3.8 [ 5, 58, 4] 3.2 [ 5, 58, 3] 3.6 L TPJ (21/39) [ 50, 60, 23] 4.4 [ 50, 60, 23] 4.9 [ 50, 59, 23] 4.9 [ 50, 60, 23] 4.2 L V1 (17) [ 7, 88, 2] 4.3 [ 6, 88, 1] 5.5 [ 7, 88, 3] 5.4 [ 7, 88, 3] 5.4 L Vis Assoc (19) [ 42, 84, 9] 5.1 [ 41, 84, 9] 6.2 [ 41, 85, 9] 5.3 [ 41, 85, 9] 4.7 R V2 (18) [17, 95, 2] 3.3 [17, 94, 1] 3.4 L Inf Temp (20) [ 39, 18, 34] 3.4 [ 40, 20, 34] 3.8 L Occ/Temp (37) [ 54, 66, 13] 4.8 Note The ROis identified in the IC>NC contrast with significant activation; p<0.01 with FDR correction. Each ROI was identified by local maximum t values and is made up of at least 40 voxels. The MNI [x,y,z] coordinates for the center of weight and mean t value of ea ch ROI are reported for each 100ms time averaged window (ie. 500 600ms) where significant activation was found. Functional Region Abbreviations: DMPFC = dorsomedial prefrontal cortex; TPJ = temporal/parietal junction; V1 = primary visual cortex; Vis Asso c = visual association cortex; V2 = secondary visual cortex; Inf Temp = inferior temporal cortex; Occ/Temp = occipital temporal cortex

PAGE 37

! %, Figure 3. On the left, v isual maps of posterior and left sagittal views of significant t values with p<0.01 FDR for IC>CC contrast in 100ms time averaged windows from 600ms to 1000ms post stimulus. On the right, t he current density, time series activation for three regions of i nterest are displayed. The x axis is time in ms, spanning the 200 to 1000ms time window with 0ms as the stimulus onset. The y axis is the current density amplitude in pA.m.

PAGE 38

! &! Table 4 Color Naming: Incongruent>Congruent Regions of Interest (ROIs) 600 700ms 700 800ms 800 900ms 900 1000ms Functional Region (Brodman's Area) MNI t MNI t MNI t MNI t R Pos t Parietal (7) [22, 74, 43] 4.1 [21, 73, 43] 4.2 [21, 73, 43] 3.6 L TP J (21/39) [ 53, 57, 19] 4 [ 53, 57, 19] 4.7 [ 53, 57, 18] 5.0 [ 53, 57, 18] 4.5 L Inf Temp (20) [ 40, 18, 34] 4. 1 [ 39, 18, 34] 4.1 [ 39, 18, 34] 3.7 [ 39, 18, 35] 3.6 L Occ/Temp (37) [ 54, 67, 12] 3.3 [ 54, 66, 13] 4.4 [ 54, 65, 13] 4.0 [ 54, 65, 13] 3.6 L DLPFC (46) [ 36, 31, 37] 3.5 L V1 (17) [ 13, 98, 2] 3.7 [ 13, 98, 2] 3.6 [ 13, 98, 2] 3.4 L Visual Assoc (19) [ 35, 85, 31] 4.2 [ 35, 85, 31] 4.6 [ 35, 85, 31] 4.1 L Premotor (6) [ 49, 6, 41] 4.3 Note The ROis identified in the IC>CC contrast with significant activation; p< 0.01 with FDR correction. Each ROI was identified by local maximum t values and is made up of at least 40 voxels. The MNI [x,y,z] coordinates for the center of weight and mean t value of each ROI are reported for each 100ms time averaged window (i.e. 500 600ms) where significant activation was found. Functional Region Abbreviations: Post Parietal = posterior parietal cortex; TPJ = temporal/parietal junction; Inf Temp = inferior temporal cortex; Occ/Temp = occipital temporal cortex; DLPFC = dorsolateral p refrontal cortex; V1 = primary visual cortex; Vis Assoc = visual association cortex;

PAGE 39

! &$ Figure 4. On the left, v isual maps of posterior and left sagittal views of significant t values with p<0.01 FDR for IW>NW contrast in 100ms time averaged windows from 600ms to 1000ms post stimulus. On the right, t he current density, time series activation for three regions of i nterest are displayed. The x axis is time in ms, spanning the 200 to 1000ms time window, with 0ms as the stimulus onset. The y axis is the current density amplitude in pA.m. onset.

PAGE 40

! &% Table 5 Word Reading: Incongruent>Neutral Regions of Interest (ROIs) 600 700ms 700 800ms 800 900ms 900 1000ms Functional Region (Brodman's Area) MNI t MNI t MNI t MNI t L V1 (17) [ 12, 98, 1] 5.1 [ 12, 98, 2] 6.0 [ 11, 98, 1] 6. 1 [ 11, 98, 1] 5.4 R V2 (18) [17, 95, 2] 4.0 [17, 94, 2] 3.7 [17, 94, 2] 4.2 [17, 95, 2] 4.2 L Visual Assoc (19) [ 25, 85, 32] 5.2 R Pos t Parietal (7) [22, 74, 43] 4.2 L TP J (7/21/39) [ 37, 71, 38] 4.4 [ 37, 71, 38] 5.0 [ 37, 71, 38] 5.1 Note The ROis identified in the IW>NW contrast with significant activation; p< 0.01 with FDR correction. Each ROI was identified by local maximum t values and is made up of at least 40 voxels. The MNI [x,y,z] coordinates for the center of weight and mean t value of each ROI are reported for each 100ms time averaged window (i.e. 600 700ms) where significant activation was found. Functional Region Abbreviations: V1 = primary visual cortex; V2 = secondary visual cortex; Vis Assoc = visual association cortex; Post Parietal = posterior parietal cortex; TPJ = temporal/parietal junction;

PAGE 41

! && Figure 5 On the left, v isual maps of posterior and left sagittal views of significant t values with p<0.01 FDR for IW>CW contrast in 100ms time averaged windows from 500ms to 1000ms post stimulus. On the right, t he current density, time series activation for three regions of i nterest are displayed. The x axis is time in ms, spanning the 200 to 1000ms time window, with 0ms as the stimulus onset. The y axis is the current density amplitude in pA.m.

PAGE 42

! &' Table 6 Word Reading: Incongruent>Congr uent Regions of Interest (ROIs) 500 600ms 600 700ms 700 800ms 800 900ms 900 1000ms Functional Region (Brodman's Area) MNI t MNI t MNI t MNI t MNI t L TPJ (21/39) [ 48, 64, 25] 5.0 [ 48, 64, 25] 4.7 [ 48, 64, 25] 6.7 [ 48, 64, 25] 6.0 [ 48, 64, 25] 5.7 L Visual Assoc (19) [ 19, 88, 33] 4.1 [ 19, 88, 33] 5.3 [ 35, 89, 0] 4. 6 [ 35, 89, 0] 4. 5 R V2 (18) [19, 94, 16] 3.7 L VI (17) [ 14, 98, 13] 5.2 [ 14, 98, 13] 4.4 Note The ROis identified in the IW> NW contrast with significant activation; p<0.01 with FDR correction. Each ROI was identified by local maximum t values and is made up of at least 40 voxels. The MNI [x,y,z] coordinates for the center of weight and mean t value of each ROI are reported fo r each 100ms time averaged window (i.e. 500 600ms) where significant activation was found. Functional Region Abbreviations: TPJ = temporal/parietal junction; Vis Assoc = visual association cortex; V2 = secondary visual cortex; V1 = primary visual cortex;

PAGE 43

! &( CHAPTER V I DISCUSSION In the present study, we used the spatial and temporal precision of MEG to investigate the neural correlates of the Stroop interference effect in a cued, task switching Stroop task where incongruent trials were compared to congruent and neutral trials separately in the context of the specific task performed (color naming or word reading) These different c ontrast s represent four variations of neural conflict and attentional processing that could be present in an interference effect. The behavioral results reveal that the interference effect was present in all four contrasts in the form of significantly reduced accuracy and reaction time in conditions of incongruent stimuli as compared to either neutral or congrue nt stimuli. Comparisons of source estimation maps reveal that certain posterior areas of the brain are recruited in all four variations of conflict while activation in the temporal and frontal lobes was limited to the color naming trials However, the overall spatial and temporal patterns of activations were distinct for all four conflict effects examined in this study. Behavioral Implications In measuring the mean reaction time and accuracy for all subject s across all trials, we found that behaviorally, the Stroop interference effect was present when comparing incongruent trial s to either neutral or congruent trials in both the color naming and word reading conditions. However, in both tasks, neutral trials were performed significantly faster than congruent t rials This indicates that more conflict was experienced during the congruent tasks than the neutral tasks, regardless of the stimulus attribute to which attention was given (color or word). In exam in ing the nuances in response accu racy, subjects were significantly more accurate in responding to incongruent conditions of the word reading task than

PAGE 44

! &) incongruent conditions of the color naming task. This larger interference effect for the color naming trials compared to word reading trials is consistent with task specific interference effects seen previously in cued Stroop tasks (MacDonald et al., 2000; West, 2003) It provides supporting evidence for the argument that word reading is more automatic than color naming (MacLeod & MacDonald, 2000) From these behavioral results, it follows that the color naming, incongruent>neu tral trials would produce the most robust neural interference effect. Looking at the source estimation results we indeed see more areas of activation in IC>NC than any other contrast particularly in the prefrontal cortex, an area of the brain very well documented as playing a prominent role in conflict processing (C. Carter & Van Veen, 2007; Larson et al., 2014; MacLeod & MacDonald, 2000; Nee et al., 2 007) Task Specific P refrontal A ctivation In the present study, significant activation in the prefrontal cortex was observed in the color naming trials, but not the word reading trials. Activation in the IC> NC contrast was seen in the left premotor region (500 800ms), left DM PFC (500 900ms) and left frontopolar cortex (600 900ms). In the IC>CC contrast, we see significant activation in the left premotor cortex (900 1000ms); and left DLPFC (700 800ms). In the context of the Stroop interference effect, these regions have been implicated in selectively attending to and integrating general task information, especially when the attentional needs of the task are the greatest (Banich et al., 2000 a ; Bunge et al., 2002; Harrison et al., 2005; MacDonald et al., 2000; Milham et al., 2001) This directly reinforces the behavioral results that there is more of an interference effect with incongruent color na ming trials than word reading trials. Because word reading is more of an automatic process than color naming, it follows that selectively attending to the

PAGE 45

! &* color of the stimulus is more difficult and requires higher activation within the left prefrontal regions associated with directing attention to the task relevant attributes In looking at the specific activated regions with in the left PFC in the context of the two different interference effects in the color naming task we see different patterns of activation The premotor cortex is act ivated in both IC>NC and IC>CC and plays a crucial role in the selection of movements (Rushworth, Kenner ley, & Walton, 2005) In the IC>NC contrast, we also see activation in the DMPFC and frontopolar cortex seen previously in studies on the interference effect (Aarts et al., 2009; Purmann & Pollmann, 2015; Zysset et al., 2001) Notably, activation in these regions begins prior to the activated p osterior areas of the brain. This provides supporting evidence to the top down model of control, specifically that areas in the PFC function in selective attention and activate associated regions in the posterior regions of the brain (Banich et al., 2000 b ) Significant activation in the left prefrontal regions extended for hundreds of milliseconds, eventually occurring simultaneously with posterior regions. This extended activation time range points to a mechanism of sustained crosstalk between anterior and p osterior attentional systems in resolvin g high conflict interference. In the IC>CC contrast, besides activation in the premotor cortex, t he DLPFC was activated This supports the role of the DLPFC in executive aspects of attentional selection (Banich et al., 2009; Banich, Milham, Atchley, Cohen, Webb, Wszalek, Kramer, Liang, Wright, et al., 2000; Manso uri, Tanaka, & Buckley, 2009; Silton et al., 2010) and conflict resolution between competing stimuli attributes (C. Carter & Van Veen, 2007) While activation in the DLPFC was more temporally transient in this condition than prefrontal activation in the IC>NC contrast, examination of the stimulus locked time series waveforms

PAGE 46

! &+ comparing IC a nd CC activation within the DLPF C region of interest, we can see that differences between the incongruent and congruent waves begins around 500ms and continues until 1000ms (see Figure 3) Significant differences in activation, however, are only observed during the 700 to 800ms tim e window. Based on prior research and the behavioral results of this study, it is not surprising that the color naming conditions saw more activation in the PFC than word reading conditions. However, the pattern of activation within the color naming tri als differed when incongruent trials were compared to congruent versus neutral trials. The IC>NC contrast saw much stronger medial PFC activation, whereas the IC>CC contrast had lateral activation within the PFC. From the analysis for fMRI studies, activa tion in both lateral and medial prefrontal regions was observed in both IC>NC and IC>CC contrasts across a number of studies (Nee et al., 2007) The strong early, sustained medial prefrontal activation that we see in the IC>NC contrast in this study provides evidence that this region pla ys a role in the top down biasing in times when conflict, particularly response conflict, is high. Noticeably missing from these color naming specific prefrontal activations is any significant activation within the anterior cingulate cortex ( ACC ) This is surprising as the ACC is one of the brain region s most implicated as active during resolution of the Stroop interference effect (C. Carter & Van Veen, 2007; Nee et al., 2007) Thi s absence of activation could be related to methodological or physiological reasons. Methodologicall y, one of the limitations of using MEG is its hypersensitivity to superficial cortical regions and relative insensitivity to deeper cortical sources (Goldenholz et al., 2009) Another limitation to the MEG approach is that we are not directly measuring the magnetic activity within a specific region but rather estimating where we think the neural generator is most likely to be.

PAGE 47

! &, Taking this into account, it is possible that the minimum norm estimation was not able to distinguish distinct sources between the DMPFC and the ACC. The one other study that examined the Stroop effect using M EG was also unable to localize any conflict related activity within the ACC (Galer et al., 2014) Physiologically, the ACC has been linked, not only to conflict monitoring, but also to the anticipation of high conflict stimuli (Wang, Ding, & Kluger, 2015) If we consider the experimental paradigm used in this study, each stimulus was preceded with an instructional cue that t old the subject which aspect of the stimuli to attend to. When presented with the cue "Color," I propose that subjects anticipated a high conflict situation, knowing that attention had to be placed on the less automatic stimulus at tribute. Thus in all color naming trials, regardless of congruency condition, the ACC could be highly active in anticipation of a potential incongruent stimulus. In analyzing the source estimation differences between incongruent and congruent or neutral trials, no significant difference was found in the ACC because the ACC had been primed in all of these potentially high conflict trials. Task Specific Temporal Lobe Activation Task specific significant activation was also observed in the temporal lobe. In color naming trials, the left anterior inferior temporal lobe and left occipito temporal cortex were activated in both IC>NC and IC>CC conditions. Both of these regions fall along the ventral visual stream, a network of neural regions within the inferior temporal lobe associated with object and visual identification and recognition (Goodale & Milner, 1992) Looking first at the posterior regions, we see activation in the left occipito temporal cortex an area involved in lexic al and semantic processing (Simons, Koutstaal, Prince, Wagner, & Schacter, 2003) A recent study examining the colo r naming St roop effect found activation in left occipito

PAGE 48

! '! temporal cortex and particularly in the visual word form area (VWFA), a region associated with reading words (Wallentin et al., 2015) As you move anterior along the ventral visual stream, we move away from lower level visual processing into high level conceptual associations in the anterior ventral temporal regions (Bonner & Price, 2013) Here was also see activation the color naming incongruent trials as compared to either neutral or congruent trials. The activation in the lef t inferior temporal lobe in the color naming instances of the interference effect is in line with the theory of two attentional systems working by top down biasing After PFC placed the attentional set on the color form of the stimuli, posterior systems w ere activated accordingly. Our results that the left occipito temporal cortex and anterior temporal lobe were active in both color naming interference effects but not in the wor d reading tasks supports the idea that additional neural resources are recruited in conditions of higher conflict (i e color naming over word reading). While there is an ongoing argument as to whether the conflict is modulated by increasing activation in task relevant or task irrelevant posterior regions, this study does not provide definitive evidence one way or the other. In this study, areas along the ventral visual stream associated with lexical and sematic processing were most active during the color naming interference effect. Yet that lexical and semantic processing c annot be distinguished as either task relevant or task irrelevant However, the increased activation in the ventral visual stream in the color naming contrasts provides evidence that in the word reading task, the interference effect was not as attentional ly demanding and thus did not require the recruitment of additional brain resources to resolve the conflict.

PAGE 49

! '$ Conflict Specific Activation Even though the interfere nce effect ha s long been a targ eted area of research, distinguishing between the different types of conflict has been researched more often in recent years (Aarts et al., 2009; Milham & Banich, 2005; Szucs & Soltesz, 2010) The interference effect has been seen with both response conflict and stimulus conflict (Nee et al., 2007) In the present study, we can examine the neu ral correlates behi nd two different combinations of conflict Incongruent>c ongruent is a measure of both stimulus and response conflict T he compari son of incongruent and neutral trials however, is an examination of the contributions of response conflict to the interferenc e effect This is because, in the neutral conditions, the task irrelevant attribute does not contribute information for an appropriate response and thus (Chen et al., 20 11; Milham & Banich, 2005) In examining the patterns of activation between the incongruent>neutral (I>N) and incongruent>congruent (I>C) contrasts, we see similar general patterns of activation within the various brain areas with potentially conflic t specific variations within each area In the color naming tasks, both interference contrasts had prefrontal activation but the stimulus conflict (I>C) conditions saw more lateral prefrontal activation and the response conflict (I>N) condition had a much stronger medial prefrontal activation. In the occipital lobe all interference contrasts had significant activation, but both the left V1 and right V2 cortices had much stronger activation in both of the response conflict ( I>N ) conditions as compared to the response conflict ( I>C ) conditions. One possible explanation is that the incongruent and congruent conditions were similar in terms of sema n tics and form Because of this, there was less of a difference in lower level visual processin g of I>C than I>N Of note, there was

PAGE 50

! '% no difference in the general pattern of temporal or parietal activation between the stimulus and response conflict comparisons. We argue then that while incongruent>neutral and incongruent>congruent both exhibit the Stroop interference effect, they respectively follow slightly different neural paths to successfully resolve the conflict. Within the frontal and occipital lobes, we see distinct differences in strength or region of activation, but no distinct difference s are observed in the temporal or parietal lobe. This po ints to a model of selective attentional and conflict resolution with both conflict dependent and conflict independent areas of activity. Common Posterior Activations In all four instances of the interference effect examined in this study, we see common activations in the occipital lobe, parietal, and temporal/parietal junction. Of particular interest is the high activation in the left temporal/parietal junction for all four contrasts. This area of the brain is an association cortex, incorporating information from a variety of subcortical and sensory modalities (R. Carter & Huettel, 2013) The left temporal/parietal junction in particular has been associated with language processing and attentional reorienting (R. Carter & Huettel, 2013) Consistently across all four contrasts, we see strong, sustained activation in the temporal/pariet al junction from 500ms to 1000ms post stimulus. This provides evidence that the left temporal/parietal junction is strongly activated to help resolve the interferenc e caused by both color naming and word reading incongruent stimuli. In addition, activatio n in the left anterior inferior parietal lobule and in the posterior superior parietal lobule, both areas activated in this study, has also been linked contributing to motor attention (Behrmann, Geng, & Shomstein, 2004; Rushworth et al., 2003)

PAGE 51

! '& When studying the interference effect elicited using a single word Stroop stimulus, these posterior areas are activated in a general, task independent way. While the distinct regions of interest varied slightly between the four contr asts, this provides evidence that there are posterior regions of the brain that are recruited in general conflict processing in the Stroop interference effect. Comparison to EEG findings In examining the temporal aspect of activation observed within the f our contrast s we see differences in our activation time series as compared to those seen in previous EEG papers. Very notably, the N450, which is the medial frontal negativity that usually occurs between 350 and 500 ms post stimulus, is not found in the current study. This makes sense in light of our source estimation results, since the neural generator of the N450 is usually attributed to the ACC (Badzakova Trajkov et al., 2009; Hanslmayr et al., 2008; L iotti et al., 2000; West, 2003) Galer et al. (2015) the only other MEG study to have examined the Stroop effect, was also lacking a distinct N450 wave that is so present in previous EEG Stroop effect studies (Donohue et al., 2016; Ergen et al., 2014; Larson et al., 2011; Szucs & Soltesz, 2010 ; West & Alain, 2000; West, 2003) However, in the present study, t he conflict SP, which is the centro parietal or left temporo parietal slow wave that lasts from 500 to up to1000ms post stimulus is very salient. What is lacking from current EEG resu lts concerning the conflict SP is an ability to distinguish between what specific areas of the brain are contributing to this sustained positivity. In the present study, we are able to show, using MEG, that the both prefrontal and posterior regions contri bute to the conflict SP in co nditions of high conflict, namely in the color naming contrasts. Particularly, in the color naming, incongruent>neutral contrast, we

PAGE 52

! '' see a pattern of activation within the conflict SP that supports a method of top down biasing employed to resolve the conflict. Each of the four contrasts, however, show distinct patterns of spatial and temporal activation within the conflict SP that can explain why the timing and localization of this waveform varies slightly from study to study within the EEG literature (Larson et al., 2014) Concerning the lack of significant N450, o ne p otential contributing factor could be the differential sensitivities that MEG and EEG have toward various regions of the brain (Goldenholz et al., 2009) MEG is most sensitive to superficia l and tangentially oriented sources while EEG is more sensitive to deep, radial sources (Ahlfors et al., 2010; Goldenholz et al., 2009) Many researchers have argued that in order to optimize on the temporal and spatial resolution of EEG and MEG, they should be used in conjunction with each other (Goldenholz et al., 2009; Liu et al., 2002; Sharon et al., 2007) Conclusion In this study, we see the interference effect present in different task types (color naming and word reading) and conditions of conflict (stimulus and response). Comparisons of source estimations reveal that certain posterior areas of the brain are recrui ted in all four variations of conflict while activation in the temporal and frontal lobes was limited to the less automatic, color naming trials. In comparing conflict types we see distinct differences in activation within the frontal and occipital lobes but no distinct differences are observed in the temporal or parietal lobe. Each of the four contrasts examined in this study has its own distinct temporal and spatial pattern of neural activation. This supports a model of selective attentional and confl ict resolution that recruits distinct but parallel neural networks that are both task and conflict condition specific (Fan et al., 2003; Van Veen & Carter, 2005)

PAGE 53

! '( The col or naming, incongruent>neutral condition behaviorally displayed the strongest instance of conflict and had the most significant, sustained prefrontal activation of the four contrasts. The neural pattern of activation observed in this contrast provides evi dence for a top down biasing model in times of high conflict and attention al demands. Future Work In order to strengthen the spatial an d temporal understanding of the neural networks behind the Stroop interference effect, a combination of EEG and MEG s hould be used to localize sources A different source estimation method such as beamforming, should also be used to examine the neural estimated behind these four contrasts Also, by examining the cue locked activations in the cued Stroop task, perhaps we can get more insight into the lack of activation in the ACC in the stimulus locked

PAGE 54

! ') CHAPTER V I I AUTHOR CONTRIBUTIONS Conceived and designed the experiments: B enzi K luger Performed the experiments: Jo Shattuck. Analyzed the data: Colleen Monahan Co ntributed materials/analysis tools: Colleen Monahan Wrote the paper: Colleen Monahan Affirmation of Original Work I hereby affirm that the work presented in this thesis is original except where due references are made. It does not contain any work for w hich a degree or diploma has been awarded by any University/Institution.

PAGE 55

! '* BIBLIOGRAPHY Aarts, E., Roelofs, A., & van Turennout, M. (2009). Attentional control of task and response in lateral and medial frontal cortex: Brain activity and reaction time distributions. Neuropsychologia 47 (10), 2089 2099. http://doi.org/10.1016/j.neuropsychologia.2009.03.019 Adleman, N. E., Menon, V., Blasey, C. M., White, C. D., Warsofsky, I. S., Glover, G. H., & Reiss, A. L. (2002). A developmental fMRI study of the Stroop color word task. Neuroimage 16 (1), 61 75. http://doi.org/10.1006/nimg.2001.1046 Ahlfors, S. P., Han, J., Bellivea u, J. W., & HŠmŠlŠinen, M. S. (2010). Sensitivity of MEG and EEG to source orientation. Brain Topography 23 (3), 227 232. http://doi.org/10.1007/s10548 010 0154 x Appelbaum, L., Boehler, C., Davis, L., Won, R., & Woldorff, M. (2014). The dynamics of proact ive and reactive cognitive control processes in the human brain. Journal of Cognitive Neuroscience 25 (5), 1021 1038. Badzakova Trajkov, G., Barnett, K. J., Waldie, K. E., & Kirk, I. J. (2009). An ERP investigation of the Stroop task: The role of the cingu late in attentional allocation and conflict resolution. Brain Research 1253 139 148. http://doi.org/10.1016/j.brainres.2008.11.069 Banich, M., Burgess, G., Depue, B., Ruzic, L., Bidwell, L., Hitt Laustsen, S., Willcutt, E. (2009). The neural basis of s ustained and transient attentional control in young adults with ADHD. Neuropsychologia 47 (14), 3095 3104. http://doi.org/10.1016/j.neuropsychologia.2009.07.005 Banich, M., Milham, M., Atchley, R., Cohen, N., Webb, A., Wszalek, T., Brown, C. (2000). Pref rontal regions play a predominant role in imposing an attentional "set": Evidence from fMRI. Cognitive Brain Research 10 (1 2), 1 9. http://doi.org/10.1016/S0926 6410(00)00015 X Banich, M., Milham, M., Atchley, R., Cohen, N., Webb, A., Wszalek, T., Magin R. (2000). fMri studies of Stroop tasks reveal unique roles of anterior and posterior brain systems in attentional selection. Journal of Cognitive Neuroscience 12 (6), 988 1000. http://doi.org/10.1162/08989290051137521 Banich, M., Milham, M., Jacobson, B ., Webb, A., Wszalek, T., Cohen, N., & Kramer, A. (2001). Attentional selection and the processing of task irrelevant information: Insights from fMRI examinations of the stroop task. Progress in Brain Research 134 459 470. http://doi.org/10.1016/S0079 61 23(01)34030 X Barch, D. M., Braver, T. S., Akbudak, E., Conturo, T., Ollinger, J., & Snyder, A. (2001). Anterior cingulate cortex and response conflict: Effects of response modality and processing domain. Cerebral Cortex 11 (9), 837 848. http://doi.org/10. 1093/cercor/11.9.837 Barkley, G. L., & Baumgartner, C. (2003). MEG and EEG in epilepsy. Journal of Clinical Neurophysiology!: Official Publication of the American Electroencephalographic Society 20 (3), 163 178. http://doi.org/10.1097/00004691 200305000 00002 Behrmann, M., Geng, J. J., & Shomstein, S. (2004). Parietal cortex and attention. Current

PAGE 56

! '+ Opinion in Neurobiology http://doi.org/10.1016/j.conb.2004.03.012 Bonner, M. F., & Price, A. R. (2013). Where Is the Anterior Temporal Lobe and What Does It Do? Journal of Neuroscience 33 (10), 4213 4215. http://doi.org/10.1523/JNEUROSCI.0041 13.2013 Botvinick, M. M., Cohen, J. D., & Carter, C. S. (2004). Conflict monitoring and anterior cingulate cortex: An update. Tren ds in Cognitive Sciences http://doi.org/10.1016/j.tics.2004.10.003 Brass, M., Ullsperger, M., Knoesche, T. R., von Cramon, D. Y., & Phillips, N. (2005). Who comes first? The role of the prefrontal and parietal cortex in cognitive control. Journal of Cogni tive Neuroscience 17 (9), 1367 1375. http://doi.org/10.1162/0898929054985400 Braver, T. S., Barch, D. M., Gray, J. R., Molfese, D. L., & Snyder, a. (2001). Anterior cingulate cortex and response conflict: effects of frequency, inhibition and errors. Cereb ral Cortex (New York, N.Y.!: 1991) 11 (9), 825 36. http://doi.org/10.1093/cercor/11.9.825 Bunge, S. A., Hazeltine, E., Scanlon, M. D., Rosen, A. C., & Gabrieli, J. D. E. (2002). Dissociable contributions of prefrontal and parietal cortices to response sele ction. Neuroimage 17 (3), 1562 1571. http://doi.org/10.1006/nimg.2002.1252 Carter, C., Macdonald, A., Botvinick, M., Ross, L. L., Stenger, V., Noll, D., & Cohen, J. D. (2000). Parsing executive processes: strategic vs. evaluative functions of the anterior cingulate cortex. Proceedings of the National Academy of Sciences of the United States of America 97 (4), 1944 1948. http://doi.org/10.1073/pnas.97.4.1944 Carter, C., Mintun, M., & Cohen, J. (1995). Interference and facilitation effects during selective at tention: an H215O PET study of Stroop task performance. Neuroimage http://doi.org/10.1006/nimg.1995.1034 Carter, C., & Van Veen, V. (2007). Anterior cingulate cortex and conflict detection: An update of theory and data. Cognitive, Affective, & Behavioral Neuroscience 7 (4), 367 379. http://doi.org/10.3758/CABN.7.4.367 Carter, R., & Huettel, S. (2013). A nexus model of the temporal parietal junction. Trends in Cognitive Sciences http://doi.org/10.1016/j.tics.2013.05.007 Chen, A., Bailey, K., Tiernan, B. N. & West, R. (2011). Neural correlates of stimulus and response interference in a 2 1 mapping stroop task. International Journal of Psychophysiology 80 (2), 129 138. http://doi.org/10.1016/j.ijpsycho.2011.02.012 Cieslik, E. C., Mueller, V. I., Eickhoff, C. R., Langner, R., & Eickhoff, S. B. (2015). Three key regions for supervisory attentional control: Evidence from neuroimaging meta analyses. Neuroscience and Biobehavioral Reviews http://doi.org/10.1016/j.neubiorev.2014.11.003 Coderre, E., Conklin, K., & Van Heuven, W. J. B. (2011). Electrophysiological measures of conflict detection and resolution in the Stroop task. Brain Research 1413 51 59. http://doi.org/10.1016/j.brainres.2011.07.017 Cohen, J. D., Dunbar, K., & McClelland, J. L. (1990). On the cont rol of automatic processes:

PAGE 57

! ', a parallel distributed processing account of the Stroop effect. Psychological Review 97 (3), 332 61. http://doi.org/10.1037/0033 295X.97.3.332 Corbetta, M., Miezin, F. M., Dobmeyer, S., Shulman, G. L., & Petersen, S. E. (1991). Selective and divided attention during visual discriminations of shape, color, and speed: functional anatomy by positron emission tomography. The Journal of Neuroscience!: The Official Journal of the Society for Neuroscience 11 (8), 2383 2402. Donohue, S. E., Appelbaum, L. G., McKay, C. C., & Woldorff, M. G. (2016). The neural dynamics of stimulus and response conflict processing as a function of response complexity and task demands. Neuropsychologia 84 14 28. http://doi.org/10.1016/j.neuropsy chologia.2016.01.035 Donohue, S. E., Liotti, M., Perez, R., & Woldorff, M. G. (2012). Is conflict monitoring supramodal? Spatiotemporal dynamics of cognitive control processes in an auditory Stroop task. Cognitive, Affective, & Behavioral Neuroscience 12 1 15. http://doi.org/10.3758/s13415 011 0060 z Ergen, M., Saban, S., Kirmizi Alsan, E., Uslu, A., Keskin Ergen, Y., & Demiralp, T. (2014). Time frequency analysis of the event related potentials associated with the Stroop test. International Journal of Ps ychophysiology!: Official Journal of the International Organization of Psychophysiology 94 (3), 463 472. http://doi.org/10.1016/j.ijpsycho.2014.08.177 Fan, J., Flombaum, J. I., McCandliss, B. D., Thomas, K. M., & Posner, M. I. (2003). Cognitive and brain c onsequences of conflict. NeuroImage 18 (1), 42 57. http://doi.org/10.1006/nimg.2002.1319 Galer, S., Op De Beeck, M., Urbain, C., Bourguignon, M., Ligot, N., Wens, V., De Tiege, X. (2014). Investigating the Neural Correlates of the Stroop Effect with Magn etoencephalography. Brain Topography 28 (1), 95 103. http://doi.org/10.1007/s10548 014 0367 5 Goldenholz, D. M., Ahlfors, S. P., HŠmŠlŠinen, M. S., Sharon, D., Ishitobi, M., Vaina, L. M., & Stufflebeam, S. M. (2009). Mapping the Signal To Noise Ratios of C ortical Sources in Magnetoencephalography and Electroencephalography. Human Brain Mapping 30 (4), 1077 1086. http://doi.org/10.1002/hbm.20571.Mapping Goodale, M. A., & Milner, A. D. (1992). Separate visual pathways for perception and action. Trends in Neur osciences 15 (1), 20 25. http://doi.org/10.1016/0166 2236(92)90344 8 HŠmŠlŠinen, M. (1992). Magnetoencephalography: A tool for functional brain imaging. Brain Topography 5 (2), 95 102. http://doi.org/10.1007/BF01129036 HŠmŠlŠinen, M. (2010). MNE Software U ser's guide. MGH/HMS/MIT Athinoula A. Martinos Center for Biomedical Imaging 16 (December). Retrieved from http://nmr.mgh.harvard.edu/meg/manuals/MNE manual 2.5.pdf Hanslmayr, S., Pastštter, B., BŠuml, K. H., Gruber, S., Wimber, M., & Klimesch, W. (2008). The electrophysiological dynamics of interference during the Stroop task. Journal of Cognitive Neuroscience 20 (2), 215 225. http://doi.org/10.1162/jocn.2008.20020 Harrison, B., Shaw, M., YŸcel, M., Purcell, R., Brewer, W., Strother, S., Pantelis, C. (20 05). Functional connectivity during Stroop task performance. NeuroImage 24 (1),

PAGE 58

! (! 181 191. http://doi.org/10.1016/j.neuroimage.2004.08.033 Herd, S. A., Banich, M. T., & O'Reilly, R. C. (2006). Neural mechanisms of cognitive control: an integrative model of s troop task performance and FMRI data. Journal of Cognitive Neuroscience 18 (1), 22 32. http://doi.org/10.1162/089892906775250012 Kawaguchi, S., Ukai, S., Shinosaki, K., Ishii, R., Yamamoto, M., Ogawa, A., Takeda, M. (2005). Information Processing Flow an d Neural Activations in the Dorsolateral Prefrontal Cortex in the Stroop Task in Schizophrenic Patients. Neuropsychobiology 51 (4), 191 203. http://doi.org/http://dx.doi.org/10.1159/000085594 Kawaguchi, S., Ukai, S., Shinosaki, K., Yamamoto, M., Ishii, R., Ogawa, A., Takeda, M. (2004). Neuroimaging of the information processing flow in schizophrenia during the Stroop task using a spatially filtered MEG analysis. In Frontiers in Human Brain To pography (Vol. 1270, pp. 361 364). http://doi.org/10.1016/j.ics.2004.05.050 Kerns, J. G. (2004). Anterior Cingulate Conflict Monitoring and Adjustments in Control. Science 303 (5660), 1023 1026. http://doi.org/10.1126/science.1089910 Larson, M. J., Clayson P. E., & Clawson, A. (2014). Making sense of all the conflict: A theoretical review and critique of conflict related ERPs. International Journal of Psychophysiology http://doi.org/10.1016/j.ijpsycho.2014.06.007 Larson, M. J., Farrer, T. J., & Clayson, P E. (2011). Cognitive control in mild traumatic brain injury: Conflict monitoring and conflict adaptation. International Journal of Psychophysiology 82 (1), 69 78. http://doi.org/10.1016/j.ijpsycho.2011.02.018 Larson, M. J., Kaufman, D. a S., & Perlstein, W. M. (2009). Conflict adaptation and cognitive control adjustments following traumatic brain injury. Journal of the International Neuropsychological Society!: JINS 15 (6), 927 37. http://doi.org/10.1017/S1355617709990701 Liotti, M., Woldorff, M. G., Pere z, R., & Mayberg, H. S. (2000). An ERP study of the temporal course of the Stroop color word interference effect. Neuropsychologia 38 (5), 701 711. http://doi.org/10.1016/S0028 3932(99)00106 2 Liu, A. K., Dale, A. M., & Belliveau, J. W. (2002). Monte Carlo simulation studies of EEG and MEG localization accuracy. Human Brain Mapping 16 (1), 47 62. http://doi.org/10.1002/hbm.10024 Luck, S. (2005). An Introduction to the Event Related Potential Technique Cambridge, MA: The MIT Press. MacDonald, A., Cohen, J. D., Stenger, V. a, & Carter, C. S. (2000). Dissociating the role of the dorsolateral prefrontal and anterior cingulate cortex in cognitive control. Science (New York, N.Y.) 288 (5472), 1835 1838. http://doi.org/10.1126/science.288.5472.1835 MacLeod, C. M. (1991). Half a century of research on the Stroop effect: an integrative review. Psychological Bulletin 109 (2), 163 203. http://doi.org/Doi 10.1037//0033 2909.109.2.163 MacLeod, C. M., & MacDonald, P. A. (2000). Interdimensional interference in the Stroop effect: Uncovering the cognitive and neural anatomy of attention. Trends in Cognitive Sciences 4 (10), 383 391. http://doi.org/10.1016/S1364 6613(00)01530 8

PAGE 59

! ($ Malmivuo, J. (2012). Comparison of the properties of EEG and MEG in detecting the electric activity of the brain. Brain Topography 25 (1), 1 19. http://doi.org/10.1007/s10548 011 0202 1 Mansouri, F. a, Tanaka, K., & Buckley, M. J. (2009). Conflict induced behavioural adjustment: a clue to the executive functions of the prefrontal cortex. Nature Reviews. Neuroscience 10 (2), 141 52. http://doi.org/10.1038/nrn2538 Markela Lerenc, J., Ille, N., Kaiser, S., Fiedler, P., Mundt, C., & Weisbrod, M. (2004). Prefrontal cingulate activation during executive control: Which comes first? Cognitive Brain Research 18 ( 3), 278 287. http://doi.org/10.1016/j.cogbrainres.2003.10.013 Milham, M., & Banich, M. (2005). Anterior cingulate cortex: An fMRI analysis of conflict specificity and functional differentiation. Human Brain Mapping 25 (3), 328 335. http://doi.org/10.1002/hbm.20110 Milham, M., Banich, M., Webb, A., Barad, V., Cohen, N., Wszalek, T., & Kramer, A. (2001). The relative involvement of anterior cingulate and prefrontal cortex in attentional control depends on nature of conflict. Cognitive Brain Research 12 (3), 467 473. http://doi.org/10.1016/S0926 6410(01)00076 3 Miller, E. K., & Cohen, J. D. (2001). An integrative theory of prefrontal cortex function. Annual Review of Neuroscience 24 (FEBRUARY 2001), 167 202. http://doi.org/10.1146/annure v.neuro.24.1.167 Minzenberg, M. J., Laird, A. R., Thelen, S., Carter, C. S., & Glahn, D. C. (2009). Meta analysis of 41 functional neuroimaging studies of executive function in schizophrenia. Archives of General Psychiatry 66 (8), 811 822. http://doi.org/1 0.1001/archgenpsychiatry.2009.91 Nee, D. E., Wager, T. D., & Jonides, J. (2007). Interference resolution: insights from a meta analysis of neuroimaging tasks. Cognitive, Affective & Behavioral Neuroscience 7 (1), 1 17. http://doi.org/10.3758/CABN.7.1.1 Nor dahl, T. E., Carter, C. S., Salo, R. E., Kraft, L., Baldo, J., Salamat, S., Kusubov, N. (2001). Anterior cingulate metabolism correlates with Stroop errors in paranoid schizophrenia patients. Neuropsychopharmacology 25 (1), 139 148. http://doi.org/10.101 6/S0893 133X(00)00239 6 Oehrn, C. R., Hanslmayr, S., Fell, J., Deuker, L., Kremers, N. a, Do Lam, A. T., Axmacher, N. (2014). Neural communication patterns underlying conflict detection, resolution, and adaptation. The Journal of Neuroscience!: The Offic ial Journal of the Society for Neuroscience 34 (31), 10438 52. http://doi.org/10.1523/JNEUROSCI.3099 13.2014 Perlstein, W. M., Larson, M. J., Dotson, V. M., & Kelly, K. G. (2006). Temporal dissociation of components of cognitive control dysfunction in seve re TBI: ERPs and the cued Stroop task. Neuropsychologia 44 (2), 260 274. http://doi.org/10.1016/j.neuropsychologia.2005.05.009 Peterson, B. S., Skudlarski, P., Gatenby, J. C., Zhang, H., Anderson, A. W., & Gore, J. C. (1999). An fMRI study of stroop word c olor interference: Evidence for cingulate subregions subserving multiple distributed attentional systems. Biological Psychiatry

PAGE 60

! (% 45 (10), 1237 1258. http://doi.org/10.1016/S0006 3223(99)00056 6 Purmann, S., & Pollmann, S. (2015). Adaptation to recent confli ct in the classical color word Stroop task mainly involves facilitation of processing of task relevant information. Frontiers in Human Neuroscience 9 (March), 1 11. http://doi.org/10.3389/fnhum.2015.00088 Rorden, C., Karnath, H. O., & Bonilha, L. (2007). I mproving lesion symptom mapping. Journal of Cognitive Neuroscience 19 (7), 1081 1088. http://doi.org/10.1162/jocn.2007.19.7.1081 Rushworth, M., Johansen Berg, H., Gobel, S., & Devlin, J. (2003). The left parietal and premotor cortices: Motor attention and selection. In NeuroImage (Vol. 20). http://doi.org/10.1016/j.neuroimage.2003.09.011 Rushworth, M., Kennerley, S., & Walton, M. (2005). Cognitive neuroscience: resolving conflict in and over the medial frontal cortex. Current Biology!: CB 15 (2), R54 6. htt p://doi.org/10.1016/j.cub.2004.12.054 Rushworth, M., Paus, T., & Sipila, P. (2001). Attention systems and the organization of the human parietal cortex. The Journal of Neuroscience!: The Official Journal of the Society for Neuroscience 21 (14), 5262 71. ht tp://doi.org/21/14/5262 [pii] Sharon, D., HŠmŠlŠinen, M. S., Tootell, R. B. H., Halgren, E., & Belliveau, J. W. (2007). The advantage of combining MEG and EEG: Comparison to fMRI in focally stimulated visual cortex. NeuroImage 36 (4), 1225 1235. http://doi.org/10.1016/j.neuroimage.2007.03.066 Silton, R. L., Heller, W., Towers, D. N., Engels, A. S., Spielberg, J. M., Edgar, J. C., Miller, G. A. (2010). The time course of activity in dorsolateral prefrontal cortex and anterior cingulate cortex dur ing top down attentional control. NeuroImage 50 (3), 1292 1302. http://doi.org/10.1016/j.neuroimage.2009.12.061 Simons, J. S., Koutstaal, W., Prince, S., Wagner, A. D., & Schacter, D. L. (2003). Neural mechanisms of visual object priming: Evidence for perc eptual and semantic distinctions in fusiform cortex. NeuroImage 19 (3), 613 626. http://doi.org/10.1016/S1053 8119(03)00096 X Stroop, J. R. (1935). Studies of interference in serial verbal reactions. Journal of Experimental Psychology 18 (6), 643 662. http ://doi.org/10.1037/h0054651 Szucs, D., & Soltesz, F. (2010). Stimulus and response conflict in the color word Stroop task: A combined electro myography and event related potential study. Brain Research 1325 63 76. http://doi.org/10.1016/j.brainres.2010.0 2.011 Tadel, F., Baillet, S., Mosher, J. C., Pantazis, D., & Leahy, R. M. (2011). Brainstorm: A user friendly application for MEG/EEG analysis. Computational Intelligence and Neuroscience 2011 http://doi.org/10.1155/2011/879716 Ukai, S., Shinosaki, K., I shii, R., Ogawa, A., Mizuno Matsumoto, Y., Inouye, T., Takeda, M. (2002). Parallel distributed processing neuroimaging in the Stroop task using spatially filtered magnetoencephalography analysis. Neuroscience Letters 334 (1), 9 12. http://doi.org/10.1016 /S0304 3940(02)01002 9

PAGE 61

! (& Van Veen, V., & Carter, C. (2002). The anterior cingulate as a conflict monitor: FMRI and ERP studies. Physiology and Behavior 77 (4 5), 477 482. http://doi.org/10.1016/S0031 9384(02)00930 7 Van Veen, V., & Carter, C. (2005). Separat ing semantic conflict and response conflict in the Stroop task: A functional MRI study. NeuroImage 27 (3), 497 504. http://doi.org/10.1016/j.neuroimage.2005.04.042 van Veen, V., Cohen, J. D., Botvinick, M. M., Stenger, V. a, & Carter, C. S. (2001). Anterio r cingulate cortex, conflict monitoring, and levels of processing. NeuroImage 14 (6), 1302 8. http://doi.org/10.1006/nimg.2001.0923 Wallentin, M., Gravholt, C. H., & Skakkebk, A. (2015). Broca's region and Visual Word Form Area activation differ during a predictive Stroop task. Cortex 73 257 270. http://doi.org/10.1016/j.cortex.2015.08.023 Wang, C., Ding, M., & Kluger, B. M. (2015). Functional roles of neural preparatory processes in a cued Stroop task revealed by linking electrophysiology with behaviora l performance. PLoS ONE 10 (7), 1 16. http://doi.org/10.1371/journal.pone.0134686 West, R. (2003). Neural correlates of cognitive control and conflict detection in the Stroop and digit localisation tasks. Neuropsychologia 41 1122 1135. West, R., & Alain, C. (2000). Effects of task context and fluctuations of attention on neural activity supporting performance of the Stroop task. Brain Research 873 (1), 102 111. http://doi.org/10.1016/S0006 8993(00)02530 0 West, R., Bailey, K., Tiernan, B. N., Boonsuk, W., & Gilbert, S. (2012). The temporal dynamics of medial and lateral frontal neural activity related to proactive cognitive control. Neuropsychologia 50 (14), 3450 3460. http://doi.org/10.1016/j.neuropsycholo gia.2012.10.011 Wheless, J. W., Castillo, E., Maggio, V., Kim, H. L., Breier, J. I., Simos, P. G., & Papanicolaou, A. C. (2004). Magnetoencephalography (MEG) and magnetic source imaging (MSI). The Neurologist 10 (3), 138 153. http://doi.org/10.1097/01.nrl. 0000126589.21840.a1 Xiao, X., Qiu, J., & Zhang, Q. (2009). The dissociation of neural circuits in a Stroop task. Neuroreport 20 (7), 674 8. http://doi.org/10.1097/WNR.0b013e32832a0a10 Zysset, S., MŸller, K., Lohmann, G., & von Cramon, D. Y. (2001). Color W ord Matching Stroop Task: Separating Interference and Response Conflict. NeuroImage 13 (1), 29 36. http://doi.org/10.1006/nimg.2000.0665 Zysset, S., Schroeter, M. L., Neumann, J., & Yves von Cramon, D. (2007). Stroop interference, hemodynamic response and aging: An event related fMRI study. Neurobiology of Aging 28 (6), 937 946. http://doi.org/10.1016/j.neurobiolaging.2006.05.008

PAGE 62

! (' APPENDIX A Steps T ake n within Brainstorm and Matlab A.1. Pre Processing (for Each S ubject) Import raw, continuous data files into Brainstorm Warp default anatomy within Brainstorm using digitized head and fiducial points to create pseudo individual anatomy Band pass filter raw data from 0.1 to 30 Hz Detect and remove unwanted artifact using SSP Visually inspect continuous data to reject time segments with poor quality and excessive noise Reject trials with incorrect responses Run custom Matlab code to calculate reaction time and accuracy (See Appendix B.1. ) Set up epochs for each condition spanning 200 to 1000 ms relative to stimulus onset (0ms) R emove DC offset and reject trials with excessive amplitude (>4500fT)

PAGE 63

! (( A.2. Source Estimation (for Each C ondition within Each S ubject) and Statistical Tests Average epoch files for each condition Create a head model using overlapping spheres Calculate inverse model using a whitened and depth weighted linear L2 minimum norm estimation S patially smooth source images using a Gaussian kernel of 12mm FWHM Run paired t tests on the following contrasts: IC>NC IC>CC IW>NW IW>CW (See Appendix A.3.) Define regions of activation masks based on local maximum t values For each region of activation (ROA), calculate: center of weight mean t value (See Appendix B.2.) Define spatial locations of ROAs by Brodman's Area using MRIcron software

PAGE 64

! () A.3. Statistical Tests Workflow Note : Paired t tests were conducted on the following sets of contrasts: incongruent, color>congruent, color (IC>CC); incongruent, color>incongruent, neutral (IC>IN); incongruent, word>congruent, word (IW>CW); incongrue nt, word>neutral word (IW>NW). Each paired t test resulted in a map of voxel values consisting of t values. For each of the four contrasts, t maps were generated for 100ms averaged time windows, ranging from 0 to 1000 ms post stimulus. Areas of signific ant difference were reported at p<0.01 corrected for false discovery rate (FDR). On those maps displaying significant difference, regions of interest were defined based on local maximum t values in areas of significance of at least 40 voxels. For each re gions of interest, the center of weight and mean t value were calculated using custom Matlab scripts. Incongruent>Neutral Incongruent>Congruent Incongruent>Neutral Incongruent>Congruent Color Word !"#$%&'()(%*(*'' !"#$#%&'()& 0ms 100ms 200ms 300ms 400ms 500ms 600ms 700ms 800ms 900ms 1000 Color Word Interference Contrasts Define regions of activation (ROA) masks based on local maximum t values

PAGE 65

! (* APPENDIX B Custom Matlab Scripts B.1. Calculate Reaction Time % C alculate reaction time and accuracy for Transition T ype, Task T ype and Congruency as % well as the combination of those three categories for each subject % RTs: Differences between the stimuli and the responses D = SUB.F.events; % Trigger 10 trig10 = D(88).times; resp10 = D(39).times; S10 = D(68).times; %Switch Trigger:10 s10 = [zeros(1,(length(trig10))); resp10]; for j = 1:length(S10) for i = 1:length(trig10) if trig10(i)>S10(j) && trig10(i)<(S10(j)+5.1) s10(1,i)=trig10(i); end end end S_RT10 = s10(2,:) s10(1,:); for i = 1:length(S_RT10) if S_RT10(i)>3 S_RT10(i)=0; end end S_RT10a = nonzeros(S_RT10); %Response time for trigger 10 that is a switch R10 = D(48).times; %Repeat Trigger:10; r10 = [zeros(1,(length(trig10))); resp10]; for j = 1:length(R10) for i = 1:length(trig10 ) if trig10(i)>R10(j) && trig10(i)<(R10(j)+5.1) r10(1,i)=trig10(i); end end end R_RT10 = r10(2,:) r10(1,:); for i = 1:length(R_RT10) if R_RT10(i)>3 R_RT10(i)=0; end end R_RT10a = nonzeros (R_RT10); %Response time for Trigger 10 that is a repeat

PAGE 66

! (+ S_RT10 = setdiff(S_RT10a,R_RT10a); R_RT10 = setdiff(R_RT10a,S_RT10a); RT10 = resp10 trig10; %Response times of all Trigger 10 % Trigger 20 trig20 = D(108).times; resp20 = D(6).times; S20 = D(77) .times; %Switch Trigger:20 s20 = [zeros(1,(length(trig20))); resp20]; for j = 1:length(S20) for i = 1:length(trig20) if trig20(i)>S20(j) && trig20(i)<(S20(j)+5.1) s20(1,i)=trig20(i); end end end S_RT20 = s20(2,:) s20(1 ,:); for i = 1:length(S_RT20) if S_RT20(i)>3 S_RT20(i)=0; end end S_RT20a = nonzeros(S_RT20); %Response time for trigger 20 that is a switch R20 = D(57).times; %Repeat Trigger:20; r20 = [zeros(1,(length(trig20))); resp20]; for j = 1:length(R20) for i = 1:length(trig20) if trig20(i)>R20(j) && trig20(i)<(R20(j)+5.1) r20(1,i)=trig20(i); end end end R_RT20 = r20(2,:) r20(1,:); for i = 1:length(R_RT20) if R_RT20(i)>3 R_RT20(i)=0; e nd end R_RT20a = nonzeros(R_RT20); %Response time for Trigger 20 that is a repeat S_RT20 = setdiff(S_RT20a,R_RT20a); R_RT20 = setdiff(R_RT20a,S_RT20a); RT20 = resp20 trig20; % Trigger 30 trig30 = D(110).times; resp30 = D(35).times; S30 = D(78).times; %S witch Trigger:30 s30 = [zeros(1,(length(trig30))); resp30];

PAGE 67

! (, for j = 1:length(S30) for i = 1:length(trig30) if trig30(i)>S30(j) && trig30(i)<(S30(j)+5.1) s30(1,i)=trig30(i); end end end S_RT30 = s30(2,:) s30(1,:); for i = 1:length(S_RT30) if S_RT30(i)>3 S_RT30(i)=0; end end S_RT30a = nonzeros(S_RT30); %Response time for trigger 30 that is a switch R30 = D(58).times; %Repeat Trigger:30; r30 = [zeros(1,(length(trig30))); resp30]; for j = 1:length(R30) for i = 1:length(trig30) if trig30(i)>R30(j) && trig30(i)<(R30(j)+5.1) r30(1,i)=trig30(i); end end end R_RT30 = r30(2,:) r30(1,:); for i = 1:length(R_RT30) if R_RT30(i)>3 R_RT30(i)=0; end end R_RT30a = nonzeros( R_RT30); %Response time for Trigger 30 that is a repeat S_RT30 = setdiff(S_RT30a,R_RT30a); R_RT30 = setdiff(R_RT30a,S_RT30a); RT30 = resp30 trig30; % Trigger 40 trig40 = D(112).times; resp40 = D(43).times; S40 = D(79).times; %Switch Trigger:40 s40 = [zeros(1,(length(trig40))); resp40]; for j = 1:length(S40) for i = 1:length(trig40) if trig40(i)>S40(j) && trig40(i)<(S40(j)+5.1) s40(1,i)=trig40(i); end end end S_RT40 = s40(2,:) s40(1,:);

PAGE 68

! )! for i = 1:length(S_RT40) if S_RT40(i)>3 S_RT40(i)=0; end end S_RT40a = nonzeros(S_RT40); %Response time for trigger 40 that is a switch R40 = D(59).times; %Repeat Trigger:40; r40 = [zeros(1,(length(trig40))); resp40]; for j = 1:length(R40) for i = 1:length(trig40 ) if trig40(i)>R40(j) && trig40(i)<(R40(j)+5.1) r40(1,i)=trig40(i); end end end R_RT40 = r40(2,:) r40(1,:); for i = 1:length(R_RT40) if R_RT40(i)>3 R_RT40(i)=0; end end R_RT40a = nonzeros(R_RT40); %Response time for Trigger 40 that is a repeat S_RT40 = setdiff(S_RT40a,R_RT40a); R_RT40 = setdiff(R_RT40a,S_RT40a); RT40 = resp40 trig40; % Trigger 50 trig50 = D(114).times; resp50 = D(7).times; S50 = D(80).times; %Switch Trigger:30 s50 = [zeros(1,(length(trig50 ))); resp50]; for j = 1:length(S50) for i = 1:length(trig50) if trig50(i)>S50(j) && trig50(i)<(S50(j)+5.1) s50(1,i)=trig50(i); end end end S_RT50 = s50(2,:) s50(1,:); for i = 1:length(S_RT50) if S_RT50(i)>3 S _RT50(i)=0; end end S_RT50a = nonzeros(S_RT50); %Response time for trigger 30 that is a switch R50 = D(60).times; %Repeat Trigger:30; r50 = [zeros(1,(length(trig50))); resp50];

PAGE 69

! )$ for j = 1:length(R50) for i = 1:length(trig50) if trig50(i)> R50(j) && trig50(i)<(R50(j)+5.1) r50(1,i)=trig50(i); end end end R_RT50 = r50(2,:) r50(1,:); for i = 1:length(R_RT50) if R_RT50(i)>3 R_RT50(i)=0; end end R_RT50a = nonzeros(R_RT50); %Response time for Trigger 50 that is a repeat S_RT50 = setdiff(S_RT50a,R_RT50a); R_RT50 = setdiff(R_RT50a,S_RT50a); RT50 = resp50 trig50; %Trigger 60 trig60 = D(116).times; resp60 = D(36).times; S60 = D(81).times; %Switch Trigger:60 s60 = [zeros(1,(length(trig60))); resp60]; for j = 1: length(S60) for i = 1:length(trig60) if trig60(i)>S60(j) && trig60(i)<(S60(j)+5.1) s60(1,i)=trig60(i); end end end S_RT60 = s60(2,:) s60(1,:); for i = 1:length(S_RT60) if S_RT60(i)>3 S_RT60(i)=0; end end S_RT60a = nonzeros(S_RT60); %Response time for trigger 60 that is a switch R60 = D(61).times; %Repeat Trigger:60; r60 = [zeros(1,(length(trig60))); resp60]; for j = 1:length(R60) for i = 1:length(trig60) if trig60(i)>R60(j) && trig60(i)<(R60(j) +5.1) r60(1,i)=trig60(i); end end end R_RT60 = r60(2,:) r60(1,:);

PAGE 70

! )% for i = 1:length(R_RT60) if R_RT60(i)>3 R_RT60(i)=0; end end R_RT60a = nonzeros(R_RT60); %Response time for Trigger 60 that is a repeat S_RT60 = setdi ff(S_RT60a,R_RT60a); R_RT60 = setdiff(R_RT60a,S_RT60a); RT60 = resp60 trig60; %Trigger 70 trig70 = D(118).times; resp70 = D(44).times; S70 = D(82).times; %Switch Trigger:70 s70 = [zeros(1,(length(trig70))); resp70]; for j = 1:length(S70) for i = 1:length(trig70) if trig70(i)>S70(j) && trig70(i)<(S70(j)+5.1) s70(1,i)=trig70(i); end end end S_RT70 = s70(2,:) s70(1,:); for i = 1:length(S_RT70) if S_RT70(i)>3 S_RT70(i)=0; end end S_RT70a = nonzeros(S_ RT70); %Response time for trigger 70 that is a switch R70 = D(62).times; %Repeat Trigger:70; r70 = [zeros(1,(length(trig70))); resp70]; for j = 1:length(R70) for i = 1:length(trig70) if trig70(i)>R70(j) && trig70(i)<(R70(j)+5.1) r7 0(1,i)=trig70(i); end end end R_RT70 = r70(2,:) r70(1,:); for i = 1:length(R_RT70) if R_RT70(i)>3 R_RT70(i)=0; end end R_RT70a = nonzeros(R_RT70); %Response time for Trigger 70 that is a repeat S_RT70 = setdiff(S_RT70a,R_RT70a); R_RT70 = setdiff(R_RT70a,S_RT70a);

PAGE 71

! )& RT70 = resp70 trig70; %Trigger 80 trig80 = D(120).times; resp80 = D(8).times; S80 = D(83).times; %Switch Trigger:80 s80 = [zeros(1,(length(trig80))); resp80]; for j = 1:length(S80) for i = 1:length(trig80) if trig80(i)>S80(j) && trig80(i)<(S80(j)+5.1) s80(1,i)=trig80(i); end end end S_RT80 = s80(2,:) s80(1,:); for i = 1:length(S_RT80) if S_RT80(i)>3 S_RT80(i)=0; end end S_RT80a = nonzeros(S_RT80); %Response time for tr igger 80 that is a switch R80 = D(63).times; %Repeat Trigger:80; r80 = [zeros(1,(length(trig80))); resp80]; for j = 1:length(R80) for i = 1:length(trig80) if trig80(i)>R80(j) && trig80(i)<(R80(j)+5.1) r80(1,i)=trig80(i); end end end R_RT80 = r80(2,:) r80(1,:); for i = 1:length(R_RT80) if R_RT80(i)>3 R_RT80(i)=0; end end R_RT80a = nonzeros(R_RT80); %Response time for Trigger 80 that is a repeat S_RT80 = setdiff(S_RT80a,R_RT80a); R_RT80 = setdiff(R_RT80a,S _RT80a); RT80 = resp80 trig80; %Trigger 90 trig90 = D(122).times; resp90 = D(37).times; S90 = D(84).times; %Switch Trigger:90 s90 = [zeros(1,(length(trig90))); resp90]; for j = 1:length(S90) for i = 1:length(trig90)

PAGE 72

! )' if trig90(i)>S90(j) && tr ig90(i)<(S90(j)+5.1) s90(1,i)=trig90(i); end end end S_RT90 = s90(2,:) s90(1,:); for i = 1:length(S_RT90) if S_RT90(i)>3 S_RT90(i)=0; end end S_RT90a = nonzeros(S_RT90); %Response time for trigger 90 that is a switch R90 = D(64).times; %Repeat Trigger:90; r90 = [zeros(1,(length(trig90))); resp90]; for j = 1:length(R90) for i = 1:length(trig90) if trig90(i)>R90(j) && trig90(i)<(R90(j)+5.1) r90(1,i)=trig90(i); end end end R_RT90 = r9 0(2,:) r90(1,:); for i = 1:length(R_RT90) if R_RT90(i)>3 R_RT90(i)=0; end end R_RT90a = nonzeros(R_RT90); %Response time for Trigger 90 that is a repeat S_RT90 = setdiff(S_RT90a,R_RT90a); R_RT90 = setdiff(R_RT90a,S_RT90a); RT90 = resp90 tr ig90; %Trigger 100 trig100 = D(87).times; resp100 = D(39).times; S100 = D(67).times; %Switch Trigger:100 s100 = [zeros(1,(length(trig100))); resp100]; for j = 1:length(S100) for i = 1:length(trig100) if trig100(i)>S100(j) && trig100(i)< (S100(j)+5.1) s100(1,i)=trig100(i); end end end S_RT100 = s100(2,:) s100(1,:); for i = 1:length(S_RT100) if S_RT100(i)>3

PAGE 73

! )( S_RT100(i)=0; end end S_RT100a = nonzeros(S_RT100); %Response time for trigger 100 that is a sw itch R100 = D(47).times; %Repeat Trigger:100; r100 = [zeros(1,(length(trig100))); resp100]; for j = 1:length(R100) for i = 1:length(trig100) if trig100(i)>R100(j) && trig100(i)<(R100(j)+5.1) r100(1,i)=trig100(i); end en d end R_RT100 = r100(2,:) r100(1,:); for i = 1:length(R_RT100) if R_RT100(i)>3 R_RT100(i)=0; end end R_RT100a = nonzeros(R_RT100); %Response time for Trigger 100 that is a repeat S_RT100 = setdiff(S_RT100a,R_RT100a); R_RT100 = setdiff (R_RT100a,S_RT100a); RT100 = resp100 trig100; %Trigger 110 trig110 = D(90).times; resp110 = D(3).times; S110 = D(69).times; %Switch Trigger:110 s110 = [zeros(1,(length(trig110))); resp110]; for j = 1:length(S110) for i = 1:length(trig110) if trig110(i)>S110(j) && trig110(i)<(S110(j)+5.1) s110(1,i)=trig110(i); end end end S_RT110 = s110(2,:) s110(1,:); for i = 1:length(S_RT110) if S_RT110(i)>3 S_RT110(i)=0; end end S_RT110a = nonzeros(S_RT110); %Response time for trigger 110 that is a switch R110 = D(49).times; %Repeat Trigger:110; r110 = [zeros(1,(length(trig110))); resp110]; for j = 1:length(R110) for i = 1:length(trig110)

PAGE 74

! )) if trig110(i)>R110(j) && trig110(i)<(R110(j)+5.1) r110(1 ,i)=trig110(i); end end end R_RT110 = r110(2,:) r110(1,:); for i = 1:length(R_RT110) if R_RT110(i)>3 R_RT110(i)=0; end end R_RT110a = nonzeros(R_RT110); %Response time for Trigger 110 that is a repeat S_RT110 = setdiff( S_RT110a,R_RT110a); R_RT110 = setdiff(R_RT110a,S_RT110a); RT110 = resp110 trig110; %Trigger 120 trig120 = D(92).times; resp120 = D(32).times; S120 = D(70).times; %Switch Trigger:120 s120 = [zeros(1,(length(trig120))); resp120]; for j = 1:length(S120) for i = 1:length(trig120) if trig120(i)>S120(j) && trig120(i)<(S120(j)+5.1) s120(1,i)=trig120(i); end end end S_RT120 = s120(2,:) s120(1,:); for i = 1:length(S_RT120) if S_RT120(i)>3 S_RT120(i)=0; end end S_RT120a = nonzeros(S_RT120); %Response time for trigger 120 that is a switch R120 = D(50).times; %Repeat Trigger:120; r120 = [zeros(1,(length(trig120))); resp120]; for j = 1:length(R120) for i = 1:length(trig120) if trig120(i)>R120(j) && trig1 20(i)<(R120(j)+5.1) r120(1,i)=trig120(i); end end end R_RT120 = r120(2,:) r120(1,:); for i = 1:length(R_RT120) if R_RT120(i)>3

PAGE 75

! )* R_RT120(i)=0; end end R_RT120a = nonzeros(R_RT120); %Response time for Trigger 120 that i s a repeat S_RT120 = setdiff(S_RT120a,R_RT120a); R_RT120 = setdiff(R_RT120a,S_RT120a); RT120 = resp120 trig120; %Trigger 130 trig130 = D(94).times; resp130 = D(41).times; S130 = D(71).times; %Switch Trigger:130 s130 = [zeros(1,(length(trig130))); resp13 0]; for j = 1:length(S130) for i = 1:length(trig130) if trig130(i)>S130(j) && trig130(i)<(S130(j)+5.1) s130(1,i)=trig130(i); end end end S_RT130 = s130(2,:) s130(1,:); for i = 1:length(S_RT130) if S_RT130(i)>3 S_RT130(i)=0; end end S_RT130a = nonzeros(S_RT130); %Response time for trigger 130 that is a switch R130 = D(51).times; %Repeat Trigger:130; r130 = [zeros(1,(length(trig130))); resp130]; for j = 1:length(R130) for i = 1:length(trig130) if trig130(i)>R130(j) && trig130(i)<(R130(j)+5.1) r130(1,i)=trig130(i); end end end R_RT130 = r130(2,:) r130(1,:); for i = 1:length(R_RT130) if R_RT130(i)>3 R_RT130(i)=0; end end R_RT130a = nonzeros(R_RT130); %Respo nse time for Trigger 130 that is a repeat S_RT130 = setdiff(S_RT130a,R_RT130a); R_RT130 = setdiff(R_RT130a,S_RT130a); RT130 = resp130 trig130;

PAGE 76

! )+ %Trigger 140 trig140 = D(96).times; resp140 = D(4).times; S140 = D(72).times; %Switch Trigger:140 s140 = [zeros(1,(length(trig140))); resp140]; for j = 1:length(S140) for i = 1:length(trig140) if trig140(i)>S140(j) && trig140(i)<(S140(j)+5.1) s140(1,i)=trig140(i); end end end S_RT140 = s140(2,:) s140(1,:); for i = 1:l ength(S_RT140) if S_RT140(i)>3 S_RT140(i)=0; end end S_RT140a = nonzeros(S_RT140); %Response time for trigger 140 that is a switch R140 = D(52).times; %Repeat Trigger:140; r140 = [zeros(1,(length(trig140))); resp140]; for j = 1:length(R140) for i = 1:length(trig140) if trig140(i)>R140(j) && trig140(i)<(R140(j)+5.1) r140(1,i)=trig140(i); end end end R_RT140 = r140(2,:) r140(1,:); for i = 1:length(R_RT140) if R_RT140(i)>3 R_RT140(i)=0; end end R_RT140a = nonzeros(R_RT140); %Response time for Trigger 140 that is a repeat S_RT140 = setdiff(S_RT140a,R_RT140a); R_RT140 = setdiff(R_RT140a,S_RT140a); RT140 = resp140 trig140; %Trigger 150 trig150 = D(98).times; resp150 = D(33).times; S150 = D(73).ti mes; %Switch Trigger:150 s150 = [zeros(1,(length(trig150))); resp150]; for j = 1:length(S150) for i = 1:length(trig150) if trig150(i)>S150(j) && trig150(i)<(S150(j)+5.1) s150(1,i)=trig150(i);

PAGE 77

! ), end end end S_RT150 = s150( 2,:) s150(1,:); for i = 1:length(S_RT150) if S_RT150(i)>3 S_RT150(i)=0; end end S_RT150a = nonzeros(S_RT150); %Response time for trigger 150 that is a switch R150 = D(53).times; %Repeat Trigger:150; r150 = [zeros(1,(length(trig150))); resp150]; for j = 1:length(R150) for i = 1:length(trig150) if trig150(i)>R150(j) && trig150(i)<(R150(j)+5.1) r150(1,i)=trig150(i); end end end R_RT150 = r150(2,:) r150(1,:); for i = 1:l ength(R_RT150) if R_RT150(i)>3 R_RT150(i)=0; end end R_RT150a = nonzeros(R_RT150); %Response time for Trigger 150 that is a repeat S_RT150 = setdiff(S_RT150a,R_RT150a); R_RT150 = setdiff(R_RT150a,S_RT150a); RT150 = resp150 trig150; %Trigger 160 trig160 = D(100).times; resp160 = D(42).times; S160 = D(74).times; %Switch Trigger:160 s160 = [zeros(1,(length(trig160))); resp160]; for j = 1:length(S160) for i = 1:length(trig160) if trig160(i)>S160(j) && trig160(i)<(S160(j)+5.1) s160(1,i)=trig160(i); end end end S_RT160 = s160(2,:) s160(1,:); for i = 1:length(S_RT160) if S_RT160(i)>3 S_RT160(i)=0; end

PAGE 78

! *! end S_RT160a = nonzeros(S_RT160); %Response time for trigger 160 that is a switch R160 = D( 54).times; %Repeat Trigger:160; r160 = [zeros(1,(length(trig160))); resp160]; for j = 1:length(R160) for i = 1:length(trig160) if trig160(i)>R160(j) && trig160(i)<(R160(j)+5.1) r160(1,i)=trig160(i); end end end R_RT160 = r160(2,:) r160(1,:); for i = 1:length(R_RT160) if R_RT160(i)>3 R_RT160(i)=0; end end R_RT160a = nonzeros(R_RT160); %Response time for Trigger 160 that is a repeat S_RT160 = setdiff(S_RT160a,R_RT160a); R_RT160 = setdiff(R_RT160a,S_RT160a); RT160 = resp160 trig160; %Trigger 170 trig170 = D(102).times; resp170 = D(5).times; S170 = D(75).times; %Switch Trigger:170 s170 = [zeros(1,(length(trig170))); resp170]; for j = 1:length(S170) for i = 1:length(trig170) if trig170(i)>S170(j) && trig170(i)<(S170(j)+5.1) s170(1,i)=trig170(i); end end end S_RT170 = s170(2,:) s170(1,:); for i = 1:length(S_RT170) if S_RT170(i)>3 S_RT170(i)=0; end end S_RT170a = nonzeros(S_RT170); %Response time for trigger 17 0 that is a switch R170 = D(55).times; %Repeat Trigger:170; r170 = [zeros(1,(length(trig170))); resp170]; for j = 1:length(R170) for i = 1:length(trig170) if trig170(i)>R170(j) && trig170(i)<(R170(j)+5.1) r170(1,i)=trig170(i);

PAGE 79

! *$ end end end R_RT170 = r170(2,:) r170(1,:); for i = 1:length(R_RT170) if R_RT170(i)>3 R_RT170(i)=0; end end R_RT170a = nonzeros(R_RT170); %Response time for Trigger 170 that is a repeat S_RT170 = setdiff(S_RT170a,R_RT170a); R_RT170 = setdiff(R_RT170a,S_RT170a); RT170 = resp170 trig170; %Trigger 180 trig180 = D(104).times; resp180 = D(34).times; S180 = D(76).times; %Switch Trigger:180 s180 = [zeros(1,(length(trig180))); resp180]; for j = 1:length(S180) for i = 1:length(trig180) if trig180(i)>S180(j) && trig180(i)<(S180(j)+5.1) s180(1,i)=trig180(i); end end end S_RT180 = s180(2,:) s180(1,:); for i = 1:length(S_RT180) if S_RT180(i)>3 S_RT180(i)=0; end end S_RT180a = nonzeros(S_RT180); %Response time for trigger 180 that is a switch R180 = D(56).times; %Repeat Trigger:180; r180 = [zeros(1,(length(trig180))); resp180]; for j = 1:length(R180) for i = 1:length(trig180) if trig180(i)>R180(j) && trig180(i)<(R180(j)+5.1) r180(1,i)=trig180(i); end end end R_RT180 = r180(2,:) r180(1,:); for i = 1:length(R_RT180) if R_RT180(i)>3 R_RT180(i)=0; end

PAGE 80

! *% end R_RT180a = nonzeros(R_RT180); %Response time for Trigger 180 that is a repeat S_RT180 = setdiff(S_RT180a,R_RT180a); R_RT180 = setdiff(R_RT180a,S_RT180a); RT180 = resp180 trig180; %% Final Reaction Times %REPEAT/COLOR/CONGRUENT RE/ 100,110,120 a1 = vertcat(R_RT100,R_RT110,R_RT120); rcc = sum(a1); F_RCC = rcc/numel(a1); F_RCCb = std(a1); %REPEAT/COLOR/INCONGRUENT RE/ 40,50,60 a2 = vertcat(R_RT40,R_RT50,R_RT60); rci = sum(a2); F_RCI = rci/numel(a2); F_RCIb = std(a2); %REPEAT/WORD/CONGRUENT RE/ 70,80,90 a3 = vertcat(R_RT70,R_RT80,R_RT90); rwc = sum(a3); F_RWC = rwc/numel(a3); F_RWCb = std(a3); %REPEAT/WORD/INCONGRUENT RE/ 10,20,30 a4 = vertcat(R_RT10,R_RT20,R_RT30); rwi = sum(a4); F_RWI = rwi/numel(a4); F_RWIb = std(a4); %SWITCH/COLOR/CONGRUENT SW/ 100,110,120 a5 = vertcat(S_RT100,S_RT110,S_RT120); scc = sum(a5); F_SCC = scc/numel( a5); F_SCCb = std(a5); %SWITCH/COLOR/INCONGRUENT SW/ 40,50,60 a6 = vertcat(S_RT40,S_RT50,S_RT60); sci = sum(a6); F_SCI = sci/numel(a6); F_SCIb = std(a6); %SWITCH/WORD/CONGRUENT SW/ 70,80,90 a7 = vertcat(S_RT70,S_RT80,S_RT90); swc = sum(a7); F_SWC = s wc/numel(a7); F_SWCb = std(a7); %SWITCH/WORD/INCONGRUENT SW/ 10,20,30 a8 = vertcat(S_RT10,S_RT20,S_RT30); swi = sum(a8); F_SWI = swi/numel(a8);

PAGE 81

! *& F_SWIb = std(a8); %COLOR b1=horzcat(RT40,RT50,RT60,RT100,RT110,RT120,RT130,RT140,RT150); col=sum(b1); F_CueACol=col/numel(b1); F_CueAColb = std(b1); %WORD b2=horzcat(RT10,RT20,RT30,RT70,RT80,RT90,RT160,RT170,RT180); word=sum(b2); F_CueAWord=word/numel(b2); F_CueAWordb=std(b2); %REPEAT b3=vertcat(R_RT10,R_RT20,R_RT30,R_RT40,R_RT50 ,R_RT60,R_RT70,R_RT80,R_RT 9 0, R_RT100,R_RT110,R_RT120,R_RT130,R_RT140,R_RT150,R_RT160,R_RT170,R_RT1 80); c2=sum(b3); F_CueBRepeat=c2/numel(b3); F_CueBRepeatb=std(b3); %SWITCH b4=vertcat(S_RT10,S_RT20,S_RT30,S_RT40,S_RT50,S_RT60,S_ RT70,S_RT80,S_RT90, S_RT100,S_RT110,S_RT120,S_RT130 ,S_RT140,S_RT150,S_RT160,S_RT170,S_RT180); c1=sum(b4); F_CueBSwitch=c1/numel(b4); F_CueBSwitchb=std(b4); %CONG b5=horzcat(RT70,RT80,RT90,RT100,RT110,RT120); cong=sum(b5); F_CueCCong=cong/numel(b5); F_CueCCongb=std(b5); %INCONG b6=horzcat (RT10,RT20,RT30,RT40,RT50,RT60); incong=sum(b6); F_CueCIncong=incong/numel(b6); F_CueCIncongb=std(b6);

PAGE 82

! *' B.2. Calculate Center of Weight and Mean t value % Code to calculate the mean T values and center of gravity MNI coordinate for each ROI %Make one c olumn of t values for source map tvalues = map.tmap(:,1); %Call to the vertices for i=1:5 verts = cortex.Atlas(15).Scouts(i).Vertices; %Indices for dipoles within scout i scoutT = tvalues(verts); %corresponding t values to ROI vertices s coutverts = cortex.Vertices(verts,:); %corresponding vertices to ROI vertices cg_xyz(i,:) = (sum(repmat(scoutT,1,3).*scoutverts, 1)/sum(scoutT))*1000; meanT(i) = mean(scoutT); maxT(i) = max(scoutT); r = find(scoutT==max(scoutT)); max_xyz(i,:) = scoutverts(r,:)*1000; end MeanandMaxT = vertcat(meanT,maxT);