UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Modulation of rostrolateral prefrontal cortex during cognitive introspection using real-time fmri Keramatian, Kamyar 2009

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
24-ubc_2009_fall_keramatian_kamyar.pdf [ 3.62MB ]
Metadata
JSON: 24-1.0070823.json
JSON-LD: 24-1.0070823-ld.json
RDF/XML (Pretty): 24-1.0070823-rdf.xml
RDF/JSON: 24-1.0070823-rdf.json
Turtle: 24-1.0070823-turtle.txt
N-Triples: 24-1.0070823-rdf-ntriples.txt
Original Record: 24-1.0070823-source.json
Full Text
24-1.0070823-fulltext.txt
Citation
24-1.0070823.ris

Full Text

MODULATION OF ROSTROLATERAL PREFRONTAL CORTEX DURING COGNITIVE INTROSPECTION USING REAL-TIME FMRI  by KAMYAR KERAMATIAN MD, Isfahan University of Medical Sciences, 2003  A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE in  THE FACULTY OF GRADUATE STUDIES (Neuroscience) THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver)  June, 2009 © Kamyar Keramatian, 2009  Abstract The main purpose of this study was to examine the ability of healthy individuals to gain control over RLPFC activation by making use of feedback provided to them in real-time about the level of activation in their RLPFC. It was hypothesized that RLPFC is involved in introspective evaluation of thought processes. It was also hypothesized that healthy volunteers could achieve improved modulation of their RLPFC activation by using real-time fMRI feedback from that region. Seven healthy volunteers completed a pre-training session, four to six training sessions, and a post-training session. Subjects were instructed to turn their attention toward their own thoughts in order to up-regulate, and turn their attention toward external sensations to down-regulate the target brain region. During the training sessions, subjects received feedback about their level of activation in bilateral RLPFC, while no feedback was provided during the pre- and posttraining sessions. Group analysis of individual sessions revealed enhanced left RLPFC activation throughout the feedback training. In addition, direct comparison of posttraining versus pre-training sessions resulted in a significant cluster of activation in left RLPFC. These findings are consistent with the hypothesized role of RLPFC in introspective evaluation of thought processes. They also demonstrate the feasibility of using real-time fMRI feedback training to achieve enhanced modulation of higher cognitive regions such as RLPFC. Finally, the findings underscore important limitations of real-time fMRI studies including global signal modulation and potential undesirable effects of feedback on task performance. Future studies will need to address these limitations.  ii  Table of Contents Abstract ............................................................................................................................... ii Table of Contents............................................................................................................... iii List of Tables ...................................................................................................................... v List of Figures .................................................................................................................... vi Acknowledgments............................................................................................................. vii 1 Introduction................................................................................................................. 1 1.1 Real-time functional Magnetic Resonance Imaging Feedback Training: a Brief Review ............................................................................................................................ 5 1.2 Research Approach and Objectives .................................................................... 8 1.3 Hypotheses.......................................................................................................... 8 2 Methods..................................................................................................................... 10 2.1 Subjects ............................................................................................................. 10 2.2 Questionnaires and Personality Measures ........................................................ 10 2.3 fMRI Data Acquisition ..................................................................................... 10 2.4 Experimental Protocol ...................................................................................... 11 2.4.1 Region of Interest Specification................................................................ 11 2.4.2 Real-time Feedback Display ..................................................................... 12 2.4.3 Task Instructions....................................................................................... 14 2.5 Subject Debriefing ............................................................................................ 15 2.6 Data Analysis .................................................................................................... 15 2.6.1 Real-time Analysis.................................................................................... 15 2.6.2 Off-line Data Analysis .............................................................................. 15 2.6.3 Region of Interest Analysis....................................................................... 17 2.6.4 Group Analyses......................................................................................... 17 2.6.5 Individual Subject Activation Maps ......................................................... 17 2.6.6 Pre- and Post-training Comparison........................................................... 18 3 Results....................................................................................................................... 19 3.1 Questionnaires and Personality Measures ........................................................ 19 3.2 Group Analyses Results.................................................................................... 20 3.3 Pre- and Post-training Comparison................................................................... 24 3.4 Individual Subject Results ................................................................................ 25 3.4.1 Subject 1.................................................................................................... 25 3.4.2 Subject 2.................................................................................................... 28 3.4.3 Subject 3.................................................................................................... 31 3.4.4 Subject 4.................................................................................................... 34 3.4.5 Subject 5.................................................................................................... 37 3.4.6 Subject 6.................................................................................................... 40 3.4.7 Subject 7.................................................................................................... 43 3.5 Subject Debriefing ............................................................................................ 46 3.5.1 Subjective Ratings .................................................................................... 46 3.5.2 Subjective Experience............................................................................... 47 4 Discussion ................................................................................................................. 50 4.1 ROI Definition .................................................................................................. 53 4.2 Feedback Display.............................................................................................. 56 iii  4.3 Sample Size....................................................................................................... 58 4.4 Global Signal Change ....................................................................................... 59 4.5 Future Direction ................................................................................................ 60 4.6 Conclusion ........................................................................................................ 61 5 Bibliography ............................................................................................................. 62 6 Appendices................................................................................................................ 68 6.1 Appendix A: Debriefing Questions .................................................................. 68 6.2 Appendix B: Questionnaires and Personality Measures................................... 70 6.3 Appendix C: Ethics Approval Certificate ......................................................... 72  iv  List of Tables Table 3.1: Means and standard deviations of RRQ scores ............................................... 19 Table 3.2: Means and standard deviations of KIMS scores.............................................. 19 Table 3.3: Means and standard deviations of BFI-44 scores. ........................................... 20 Table 3.4: Mean subjective rating scores from debriefing sessions ................................. 47 Table 3.5: Thought contents and labels used by subjects during up-regulation. .............. 48 Table 6.1: Mean scores for RRQ ...................................................................................... 70 Table 6.2: Mean scores for KIMS..................................................................................... 70 Table 6.3: Mean scores for BFI-44 ................................................................................... 71  v  List of Figures Figure 2.1: Feedback display ............................................................................................ 13 Figure 3.1: Pre-training session group analysis: up-regulation versus down-regulation contrast.............................................................................................................................. 21 Figure 3.2: First training session group analysis: up-regulation versus down-regulation contrast.............................................................................................................................. 21 Figure 3.3: Second training session group analysis: up-regulation versus down-regulation contrast.............................................................................................................................. 22 Figure 3.4: Third training session group analysis: up-regulation versus down-regulation contrast.............................................................................................................................. 22 Figure 3.5: Fourth training session group analysis: up-regulation versus down-regulation contrast.............................................................................................................................. 23 Figure 3.6: Post training session group analysis: up-regulation versus down-regulation contrast.............................................................................................................................. 23 Figure 3.7: Post-training versus pre-training contrast....................................................... 24 Figure 3.8: Subject 1 - activation maps ............................................................................ 26 Figure 3.9: Subject 1 - correlations between task function and activation signal............. 27 Figure 3.10: Subject 2 – activation maps.......................................................................... 29 Figure 3.11: Subject 2 - correlations between task function and activation signal........... 30 Figure 3.12: Subject 3 – activation maps.......................................................................... 32 Figure 3.13: Subject 3 - correlations between task function and activation signal........... 33 Figure 3.14: Subject 4 – activation maps.......................................................................... 35 Figure 3.15: Subject 4 - correlations between task function and activation signal........... 36 Figure 3.16: Subject 5 – activation maps.......................................................................... 38 Figure 3.17: Subject 5 - correlations between task function and activation signal........... 39 Figure 3.18: Subject 6 – activation maps.......................................................................... 41 Figure 3.19: Subject 6 - correlations between task function and activation signal........... 42 Figure 3.20: Subject 7 – activation maps.......................................................................... 44 Figure 3.21: Subject 7 - correlations between task function and activation signal........... 45  vi  Acknowledgments My utmost gratitude goes to my thesis advisor, Dr. Kalina Christoff for allowing me to join her lab, for her expertise, kindness, and patience, and for her perpetual encouragement, guidance and support.  I would like to thank Dr. Adele Diamond, Dr. Elton Ngan, and Dr. Stan Floresco for agreeing to serve on my supervisory committee, and for their guidance and support. This thesis would not have been possible without the help of Graeme McCaig who developed the real-time fMRI software and assisted with data collection.  All my lab-mates at Cognitive Neuroscience of Thought Laboratory created a pleasant place to work. I would like to specifically thank Rachelle Smith for our countless discussions, her valuable comments and suggestions, and her friendship and encouragement.  vii  1 Introduction The prefrontal cortex is one of the most highly evolved areas of the human brain comprising nearly 30 % of human neocortex (Fuster, 1980). Previous patient and functional neuroimaging (including positron emission tomography; PET, and functional magnetic resonance imaging; fMRI) studies have implicated this region in complex cognitive processes (Duncan, Burgess, & Emslie, 1995; Miller & Cohen, 2001), although functional specialization within the prefrontal cortex is a matter of debate (Duncan & Owen, 2000). A number of functional subregions within the prefrontal cortex, however, have been proposed including ventromedial (Damasio, 1996), orbitofrontal (Rolls, 1996), and dorsolateral (Goldman-Rakic, 1987; Petrides, 1991) prefrontal cortices. Recently, increasing attention has been directed toward the most anterior part of the lateral prefrontal cortex, also known as rostrolateral prefrontal cortex (RLPFC), or lateral Brodmann area 10 (BA 10) (Christoff & Gabrieli, 2000; Ramnani & Owen, 2004). Several lines of evidence from anatomical studies suggest the important role of this region in higher-order human cognitive functions. First of all, BA 10 has been shown to be the largest subregion within the human prefrontal cortex (Ongur, Ferry, & Price, 2003). Furthermore, a comprehensive anatomical study on BA 10 has shown that this region (relative to the rest of the brain) is significantly larger in the human brain compared to the brains of other primates (Semendeferi, Armstrong, Schleicher, Zilles, & Van Hoesen, 2001). The same study reported that BA 10 has a lower cell density in humans than it has in other primates, suggesting that it has more space available for its supragranular layers to make connections with other higher-order association areas  1  (Semendeferi, Armstrong, Schleicher, Zilles, & Van Hoesen, 2001). Finally, in a recent anatomical study on corticocortical connections of rostral prefrontal cortex in the macaque monkey, Petrides and Pandya reported that among prefrontal subregions, rostral prefrontal cortex (including RLPFC) is the only one that is not directly connected to lowlevel perceptual or motor areas (Petrides & Pandya, 2007). The authors suggest that “the rostral prefrontal cortex, by not interacting with these posterior cortical areas, does not regulate attention to events occurring in external space” (Petrides & Pandya, 2007). This latter finding may provide the anatomical basis for the proposed role of RLPFC in processing of internally-generated information (see below). Altogether, the findings of the above anatomical experiments are consistent with the results of functional neuroimaging studies that implicate the RLPFC in a variety of complex cognitive processes. RLPFC activation has been reported during reasoning tasks such as the Tower of London (Baker et al., 1996), the Wisconsin Card Sorting Test (WCST) (Berman et al., 1995; Goldberg et al., 1998; Nagahama et al., 1996), and Raven’s Progressive Matrices Test (Christoff et al., 2001). Activation of RLPFC has also been reported in tasks requiring episodic memory retrieval (Tulving, Markowitsch, Craik, Habib, & Houle, 1996; Velanova et al., 2003), cognitive branching (Koechlin, Basso, Pietrini, Panzer, & Grafman, 1999; Koechlin, Ody, & Kouneiher, 2003), working memory (Jonides et al., 1997; Rypma, Prabhakaran, Desmond, Glover, & Gabrieli, 1999), and prospective memory (Burgess, Quayle, & Frith, 2001; den Ouden, Frith, Frith, & Blakemore, 2005; Okuda et al., 1998). By reviewing and re-examining previous functional neuroimaging studies of reasoning and episodic memory, Christoff and Gabrieli (2000) proposed a unifying hypothesis for the functional characterization of  2  RLPFC. According to this hypothesis, RLPFC is specifically involved in the processing of internal mental states or in introspective evaluation of internally-generated information i.e. information that cannot be readily perceived from the external environment but has to be inferred or self-generated (Christoff & Gabrieli, 2000). To test this hypothesis Christoff and colleagues designed an event-related fMRI experiment during which subjects performed a simple matching task under two conditions (Christoff, Ream, Geddes, & Gabrieli, 2003). In the internal condition, subjects had to infer the dimension of change (shape or texture) between a pair of objects and determine whether another pair of objects changed along the same dimension. In the external condition, subjects had to determine whether an object matched the other two objects along the specified dimension (texture or shape) by encoding the objects in terms of their perceptual features. The results showed increased RLPFC activation in the internal versus external task condition. Importantly, the differential recruitment of RLPFC activation was observed only during the period when subjects were evaluating the self-generated information, and not during the generation period itself. In a more recent fMRI study, Smith et al. (2007) employed a simplified version of the above task and showed that RLPFC was reliably recruited across individual subjects. More recently, Christoff and colleagues (in press) suggested a functional organization within the lateral prefrontal cortex based on different levels of abstraction. Using a verbal reasoning task, the authors showed that abstract concepts are represented in RPLFC, whereas more concrete concepts are represented in more posterior prefrontal subregions (Christoff, Keramatian, Gordon, Smith, & Mädler, in press). Abstract concepts by definition are not understood via direct sensory experience; rather, it has  3  been suggested that our understanding of such concepts relies on introspective mental processes (Barsalou, 1999; Christoff & Keramatian, 2008). Thus, the above findings can be interpreted as further evidence for the involvement of RLPFC in introspective thought processes. A number of recent functional neuroimaging studies (Burgess, Quayle, & Frith, 2001; den Ouden, Frith, Frith, & Blakemore, 2005; Okuda et al., 1998; Simons, Scholvinck, Gilbert, Frith, & Burgess, 2006) have implicated RLPFC activation in prospective memory, which is defined as remembering to carry out an intention in the future. Prospective memory experiments require subjects to bear in mind an intention while performing an ongoing (unrelated) task. It has been suggested that actively maintaining an intention while performing a different task involves internally-generated thought processes, a mechanism that may explain the recruitment of RLPFC during prospective memory tasks (Simons, Scholvinck, Gilbert, Frith, & Burgess, 2006). In sum, findings from anatomical as well as functional neuroimaging studies suggest the critical role of RLPFC in high-order cognition in humans. A review of these studies leads to a hypothesis that RLPFC is specifically involved in introspective evaluation of internally-generated thought processes. However, due to the wide range of tasks that are shown to activate this region along with the high level of complexity associated with those tasks, considerable debate over the precise functional role of this region continues (Bunge, Helskog, & Wendelken, 2009; Burgess, Dumontheil, & Gilbert, 2007; Gilbert et al., 2006; Ramnani & Owen, 2004). (see Ramnani & Owen, 2004 for review of other hypotheses). In the following sections I propose a novel neuroimaging technique for studying the relation between RLPFC activation and thought processes.  4  1.1 Real-time functional Magnetic Resonance Imaging Feedback Training: a Brief Review Recent technological and computational advances have resulted in the emergence of real-time fMRI, a methodology that allows simultaneous data acquisition and processing. Using this novel technology, information regarding localized brain activation can be immediately presented to the subject as a form of neurofeedback. This information can be used by the subject to guide cognitive processes that in turn can alter brain activation. In this way, subjects could learn to gain explicit control over localized brain activation (region of interest or ROI) (deCharms, 2008; Weiskopf, Scharnowski et al., 2004). Therefore, while conventional fMRI experiments have been used almost exclusively to identify associations between mental processes and regional cortical activities, real-time fMRI provides a non-invasive means to directly investigate the dynamic interaction between brain and cognition. Earlier real-time fMRI experiments targeted brain regions with well-established functions that also seemed to be easier to control such as somatosensory areas (Yoo & Jolesz, 2002), amygdala (Posse et al., 2003), and primary motor cortex (deCharms et al., 2004). A 2002 study by Yoo and colleagues provided the first evidence for successful real-time fMRI feedback training. Five subjects were presented with functional brain activation maps following 60-second periods consisting of dummy scans acquisition, hand movement task interleaved with rest, and data processing. All subjects showed an increase in the extent of activation in somatosensory and motor areas following the fMRI feedback training (Yoo & Jolesz, 2002). In another study by Posse and colleagues, six subjects received feedback from amygdala activation at the end of 60-second trials of 5  baseline and negative mood induction (Posse et al, 2003). Feedback was given over the headphones on a five point scale rating based on the intensity and spatial extent of activation. Amygdala activation (mostly left sided) was seen in 78.3% of negative mood trials. Since the authors used feedback solely to reinforce sadness induction, the effect of feedback training in this study was not explored. The above studies investigated the feasibility of providing subjects with discrete (and inevitably delayed) feedback of fMRI signal. The first study which used continuous feedback of fMRI signal with minimal delay of less than 2 seconds was reported by Weiskopf and colleagues (Weiskopf et al., 2003). The participating subject was presented with two continuously updating curves from rostral-ventral and dorsal subdivisions of the anterior cingulate cortex (ACC). The subject was instructed to attempt to modulate activation of both regions during alternating 60-second blocks of upregulation and down-regulation. No strategy was provided and the subject was instructed to develop his own strategies. The authors reported increased signal change in the rostral-ventral subdivision of ACC across the feedback sessions indicating a possible learning effect (Weiskopf et al., 2003). In another study by the same research group, four subjects were trained to modulate activations in two separate brain regions; supplementary motor area (SMA), and parahippocampal place area (PPA) (Weiskopf, Mathiak et al., 2004). The difference of mean fMRI signal from SMA and PPA was presented to subjects in a continuously updating curve display. No specific strategy was suggested to the subjects. The results showed that all subjects achieved differential fMRI signal intensity between SMA and PPA, with two subjects showing improvement across training sessions (Weiskopf, Mathiak et al., 2004).  6  Other studies employed alternative forms of continuous feedback of activation from other brain regions. deCharms and colleagues used virtual reality videos (e.g. a weightlifter raising a lift above his head as the activation goes up) in addition to continuously updating curves to determine whether subjects can learn enhanced control over primary motor cortex activation (deCharms et al., 2004). Six subjects underwent 3 sessions of feedback training which comprised alternating 30-second blocks of imagined hand movement task (up-regulation) and rest (down-regulation). Three other subjects were trained using the same paradigm, but received sham fMRI information as feedback in order to control for any training effect caused by practice alone. Increased activation in their primary motor cortex following the training was observed in the training group but not in the control group (deCharms et al., 2004). Another research group chose right anterior insular cortex as a target of modulation (Caria et al., 2007). Nine subjects underwent four feedback training sessions consisting of alternating blocks of regulation and rest conditions (training group). They were instructed to focus on emotion induction during regulation, and to count back from 100 during baseline. Feedback was displayed as a form of a graduated thermometer indicating changes of fMRI signal with increasing or decreasing number of bars. Using the same paradigm, three subjects received sham feedback and three others received no feedback at all (control groups). A linear increase of activation in the right anterior insular cortex (ROI) was observed while left insular cortex showed no increase in activation. Also, neither of the control groups showed increased fMRI signal in the right anterior insular cortex.  7  1.2 Research Approach and Objectives As discussed above, previous studies have shown that real-time fMRI information can be used to improve self-regulation of localized brain regions such as the primary motor cortex (deCharms et al., 2004), anterior cingulate cortex (Weiskopf et al., 2003), and insular cortex (Caria et al., 2007). It is not known, however, whether subjects can achieve enhanced control over activation in high level cognitive regions such as RLPFC. The primary purpose of this study was to examine the ability of healthy individuals to gain control over RLPFC activation by making use of feedback provided to them in real-time on the level of activation in their RLPFC. The second purpose was to further investigate the connection between cognitive self-awareness and RLPFC activity. Subjects were instructed about the purpose of the study, the hypothesized function of RLPFC, and corresponding strategies they could use to modulate the activation in that region. They were instructed to turn their attention towards their own thoughts to up-regulate, and to turn their attention away from their own thoughts, towards external sensations, to downregulate their RLPFC.  1.3 Hypotheses This study aimed to investigate two hypotheses regarding the function of RLPFC in the human brain: First, it was hypothesized that activation in RLPFC is linked to evaluation of internally-generated thought processes. More specifically, it was hypothesized that activation of RLPFC increases when attention is directed toward internal thoughts and decreases when it is directed to external perceptions and bodily sensations. Second, it was hypothesized that healthy volunteers can achieve improved 8  modulation of their RLPFC activation by using real-time fMRI feedback from that region. Subjects were predicted to be able to maintain the improved modulation even in the absence of feedback.  9  2 Methods 2.1 Subjects Ten subjects, (8 female; mean age, 24 years; range, 19–31) gave informed written consent to participate in this experiment. Subjects were recruited through the University of British Columbia community, were right-handed, had normal or corrected vision, and no psychiatric history. Formal interviews were conducted by the experimenters before the scanning day to ensure that subjects met all the inclusion criteria. Two subjects were excluded from the analysis due to excessive task-correlated global signal resulting from task-related breath holding. One subject was excluded due to artifact in the anterior rim of the brain. All protocols were approved by the UBC Clinical Research Ethics Board and by the UBC High Field Magnetic Imaging Centre.  2.2 Questionnaires and Personality Measures Prior to the scanning session, subjects completed a series of standardized questionnaires including the Kentucky Inventory of Mindfulness Skills (KIMS) (Baer, Smith, & Allen, 2004), the Big Five Inventory-44 (BFI-44) (Benet-Martinez & John, 1998), and the Rumination-Reflection Questionnaire (RRQ) (Trapnell & Campbell, 1999).  2.3 fMRI Data Acquisition The images were acquired at a 3.0 Tesla Philips Intera MRI scanner with an eight element, six-channel phased-array head coil with parallel imaging capability (SENSE) 10  (Pruessmann, Weiger, Scheidegger, & Boesiger, 1999). A memory foam pillow was used to limit head movement. Functional images were acquired using a T2*-weighted single shot echo-planar imaging (EPI) gradient echo sequence sensitive to blood oxygen level-dependent (BOLD) contrast (time of repetition (TR) 1 s; echo time (TE) 30 ms; flip angle (FA) 90°; field of view (FOV) 224 × 244 × 67 mm; matrix size 64 × 64; SENSE factor 2.0). Each functional volume consisted of 17 3-mm axial slices with gaps of 1 mm. Each session lasted 6 minutes and consisted of 365 functional volumes. Prior to functional imaging, an inversion recovery prepared T1- weighted fast spin-echo anatomic volume was acquired (TR 2 s; TE 10 ms; spin echo turbo factor 5, FA 90°; FOV 224 × 224 × 67 mm; acquisition matrix 240 × 235; reconstructed matrix 480 × 480, inversion delay IR 800 ms), containing 17 3-mm axial slices with gaps of 1 mm acquired in the same slice locations and used for functional images and was used for ROI specification. A high resolution 3DT1 anatomical volume (TE 3.5 ms; TR 7.7 ms; FOV 256 × 200 ×170; acquisition matrix 256 × 256; 1 × 1 × 1 isotropic voxels) was then obtained for the purpose of normalization and spatial localization of activations.  2.4 Experimental Protocol 2.4.1 Region of Interest Specification 2.4.1.1 Motor Task A total of 5 dynamics containing the most dorsal section of the brain were obtained to define the ROI for the motor task. Two rectangular areas were hand-drawn on the left and right motor cortex on 2 to 3 upper slices.  11  2.4.1.2 RLPFC Modulation Task A total number of 5 dynamics consisting of 17 slices and the IR T1 image obtained in the same alignment were used to define the ROI for the RLPFC modulation task. For subjects 1, 2, 3 and 5, the ROI was first defined in Talairach space as the region of intersection between Brodmann area 10 and middle frontal gyrus (Christoff et al., 2001). The predefined ROI was then transferred onto the subject’s anatomical space, by inverting the normalization parameters derived from normalizing the subject’s brain to the MNI template, and applying them to the ROI. For subject 4, 6 and 7 ROIs were selected visually using anatomical landmarks identified on 3D high-resolution structural images of each individual subject’s brain (see discussion). The ROI was defined bilaterally, by identifying intersection of medial / intermediate frontomarginal sulci, proceeding vertically up the intermediate frontal sulcus (Petrides & Pandya, 2004). The ROI was then manually drawn on an in-plane structural image, collected in the same slice locations as those to be used for functional acquisition. Finally, the ROI was transferred onto the corresponding slices of the functional image.  2.4.2 Real-time Feedback Display While in the scanner, subjects were presented with 3 different feedback panels and a vertical arrow (Fig. 2.1). The top right panel showed a continuously updated, scrolling graph of the BOLD time course extracted from the individually-defined ROI. This panel was removed for subject five to seven due to distraction reported by previous subjects. The top left panel, modeled after Caria et al (2007), displayed a fluctuating thermometer indicating the current level of activation. The thermometer turned red as the  12  activation went above the average and turned blue as it fell below the average. The average level of activation was measured using the first 60-second calibration period during which the subject performed the task while no feedback was provided. The bottom panel contained a history bar graph with each white bar indicating the optimal average level of activation during each block. Black bars represented the average level of activation in the preceding blocks during the whole session. A new black bar appeared following the completion of each block. The vertical arrow served to indicate the current task regulation: an upwards red arrow represented up-regulation whereas a downwards blue arrow indicated down-regulation.  Figure 2.1: Feedback display Display viewed by subjects in scanner during real-time feedback training sessions. Direction and color of the vertical arrow indicates current task (up- or down-regulation). Top right panel displays a scrolling time-course of ROI activation (removed for subject five to seven). Top left thermometer indicates current level of activation. Black bars in the bottom indicate average level of activation in preceding blocks during the whole session, updated at the end of each block.  13  2.4.3 Task Instructions 2.4.3.1 Motor Task To become familiarized with the feedback display and the delay in the activation signal due to hemodynamic response and data reconstruction, all subjects performed a motor task prior to the RLPFC regulation task. This task comprised alternating 30second blocks of finger tapping (up-regulation) and rest (down-regulation). Real-time feedback was provided from individually defined bilateral motor cortex. 2.4.3.2 RLPFC Modulation Task RLPFC modulation task consisted of 1 pre-training, 4 feedback training and 1 post-training session. Each session was 6 minutes long and comprised 12 blocks (30 seconds each) of alternating up-regulation and down-regulation. In the up-regulation condition, subjects were instructed to turn their attention inwards by trying to observe their thoughts as they happened or through labelling their thoughts as they occur in order to increase the RLPFC activation. In the down-regulation condition, subjects were instructed to turn their attention away from their thoughts, towards external sensations such as bodily sensations or scanner noise to decrease the RLPFC activation. They were also encouraged to use the feedback display to gain additional control over the activation signal during the feedback training sessions. No feedback was provided during the preand post-training sessions, however, subjects were instructed to perform the exact same task using the central arrow cue as an indication of the direction of modulation for the current block.  14  2.5 Subject Debriefing Following each RLPFC modulation session, subjects were asked to verbally answer visually displayed questions regarding their thought process, and what labels they used (if any) during the preceding up-regulation blocks. They were also asked if they experienced any intruding thoughts during the down-regulation blocks. Finally they were requested to rate their level of mental effort and how much they use the feedback display during up- and down-regulation blocks on a scale of 1 to 7 ( the latter question was asked only for the feedback training sessions; see Appendix A for list of questions).  2.6 Data Analysis 2.6.1 Real-time Analysis Imaging data were analyzed in real-time using custom software programmed in C++ based on a DLL (dynamically linked library) generated by Philips for the purpose of real-time data acquisition. Data processing comprised motion correction and continuous measurement of ROI activation as indexed by percent signal change from the average that was calculated based on the first 60-second calibration period. Data were then temporally filtered using a second-order butterworth bandpass filter (0.0125 - 0.08Hz) to eliminate high and low frequency noise. Motion estimates were computed based on the first volume for each session to correct for between-session motions.  2.6.2 Off-line Data Analysis The fMRI images were preprocessed and analyzed using statistical parametric mapping (SPM5) (Wellcome Department of Imaging Neuroscience, London). Preprocessing started with slice timing correction to correct for the different sampling  15  times of the slices. Functional volumes were then realigned to the first volume of the 5 dynamic functional scans to correct for slight head movements. Finally, data were spatially smoothed with an 8-mm full-width at half-maximum (FWHM) Gaussian kernel. ROI data analysis was performed on non-normalized data to avoid any additional preprocessing that may introduce additional differences between the observed time courses by subjects in the scanner, and the ones obtained off-line. For group fixed effects analysis, all functional volumes were realigned to the first one in the time series. The structural T1-weighted volume was segmented to extract a gray matter image for each subject, which was spatially normalized to a gray matter image of the MNI template. The derived spatial transformations for each subject were applied to the realigned functional volumes to bring them into standardized MNI space. After normalization, all volumes were resampled in 2 × 2 × 4 mm voxels using sinc interpolation in space. Statistical analysis was performed at each voxel to assess the magnitude of differences between conditions. An anatomically-defined gray matter mask was created and explicitly specified to ensure that statistical analysis was performed in all brain regions, including those where signal may be low due to susceptibility artifacts. To remove low-frequency drift in the BOLD signal, data were high-pass filtered using an upper cut-off period of 128 s. No global scaling was performed. Condition effects at each voxel were estimated according to the general linear model (Friston et al., 1995). The model included: i) the observed time-series of intensity values, representing the dependent variable; and ii) regressor functions constructed by convolving conditionspecific box-car functions with a synthetic hemodynamic response function (HRF),  16  modeling the up-regulation and down-regulation conditions. Regionally specific effects for up-regulation were estimated by positively weighting the parameter estimate for the up-regulation regressor and negatively weighting the parameter estimate for the downregulation regressor in a linear contrast. The first three functional images from each session were excluded from all analyses to account for the reconstruction delay during scanning (i.e. the lag before subjects were able to observe changes in the feedback display).  2.6.3 Region of Interest Analysis Activation time-courses were extracted from the non-normalized images from every session for each subject using the SPM5 volumes toolbox. The extraction volume was specified using the ROI image file from the scanning session. The signal was high and low-pass filtered at 0.0078 and 0.15 respectively and linear detrended. Finally, the correlation between the mean intensity value for each time point and an HRF-convolved boxcar task function was computed to indicate performance for individual subjects during each run. Higher positive r-values indicated better modulation of the ROI activation.  2.6.4 Group Analyses Group fixed effects analyses from all 7 subjects were performed on each individual session separately. This included pre and post-training sessions and four training sessions (for those subjects who completed more than 4 training sessions, only the first four training sessions were included in the group analysis).  2.6.5 Individual Subject Activation Maps To examine the pattern of activation for each session separately for each subject, individual sessions were specified within separate models using the non-normalized data.  17  2.6.6 Pre- and Post-training Comparison A paired t-test of the contrast images from the up-regulation versus downregulation contrast was used in order to compare the activations in pre- and post-training sessions.  18  3 Results 3.1 Questionnaires and Personality Measures Scores on RRQ, KIMS, and BFI-44 measures were compared with previously published scores for healthy adults using two-tailed t-tests; no significant difference was observed for any of these measures ( see Table 3.1 to 3.3): Rumination [t(1142) = -0.62, ns]; Reflection [t(1142) = 1.62, ns]; KIMS (Observe) [t(220) = 1.70, ns]; KIMS (Describe) [t(220) = -0.23, ns]; KIMS (Act with awareness) [t(220) = 0.24, ns]; KIMS (Accept without judgment) [t(220) = -0.13, ns]; BFI-44 (Extraversion) [t(776) = 0.63, ns]; BFI-44 (Agreeableness) [t(776) = -1.11, ns]; BFI-44 (Conscientiousness) [t(776) = 0.07, ns]; BFI-44 (Neuroticism) [t(776) = -0.70, ns]; and BFI-44 (Openness) [t(776) = 1.76, ns]. See Appendix B for individual subject scores.  Table 3.1: Means and standard deviations of RRQ scores  Current study N=7 RRQ scale M SD Rumination 3.27 0.89 Reflection 3.61 0.54 * (Trapnell & Campbell, 1999)  Previously published* N = 1137 M SD 3.46 0.71 3.14 0.76  Table 3.2: Means and standard deviations of KIMS scores  KIMS scale Observe Describe Act with awareness Accept without judgment *(Baer, Smith, & Allen, 2004)  Current study N=7 M SD 43.72 7.48 27.72 6.04 29.71 5.94 29.29 8.62  Previously published* N = 215 M SD 38.63 7.80 28.21 5.48 29.22 5.37 29.61 6.50  19  Table 3.3: Means and standard deviations of BFI-44 scores.  Personality traits Extraversion Agreeableness Conscientiousness Neuroticism Openness *(Benet-Martinez & John, 1998)  Current study N=7 M SD 3.39 0.97 3.59 0.65 3.62 0.74 2.79 1.13 4.1 0.41  Previously published* N = 711 M SD 3.2 0.8 3.8 0.5 3.6 0.7 3.0 0.8 3.7 0.6  3.2 Group Analyses Results The group fixed effects analyses performed separately for each session consistently revealed clusters of activation in the left RLPFC. All activations were significant at p< 0.05, corrected for multiple comparisons, with the exception of the first training session which was significant at a more lenient threshold of p< 0.001 uncorrected (Fig 3.1-6). No activation was observed in the right RLPFC. The results show a gradual increase in the extent of activation across training sessions and also from pre- to posttraining session.  20  Figure 3.1: Pre-training session group analysis: up-regulation versus down-regulation contrast. MNI coordinates of most significant voxel within the left RLPFC (encircled): -26, 66, 8 (t = 7.46, p < 0.05 corrected). Activation map is overlaid on single subject T1 template.  Figure 3.2: First training session group analysis: up-regulation versus down-regulation contrast. MNI coordinates of most significant voxel within the left RLPFC (encircled): -36, 54, 20 (t = 4.43, p = 0.065 corrected, p < 0.001 uncorrected). Activation map is overlaid on single subject T1 template.  21  Figure 3.3: Second training session group analysis: up-regulation versus down-regulation contrast. MNI coordinates of most significant voxel within the left RLPFC (encircled): -42, 52, 12 (t = 6.43, p < 0.05 corrected). Activation map is overlaid on single subject T1 template.  Figure 3.4: Third training session group analysis: up-regulation versus down-regulation contrast. MNI coordinates of most significant voxel within the left RLPFC (encircled): -36, 56, 12 (t = 6.79, p < 0.05 corrected). Activation map is overlaid on single subject T1 template.  22  Figure 3.5: Fourth training session group analysis: up-regulation versus down-regulation contrast. MNI coordinates of most significant voxel within the left RLPFC (encircled): -32, 50, 28 (t = 7.50, p < 0.05 corrected). Activation map is overlaid on single subject T1 template.  Figure 3.6: Post training session group analysis: up-regulation versus down-regulation contrast. MNI coordinates of most significant voxel within the left RLPFC (encircled): -20, 54, 28 (t = 11.34, p < 0.05 corrected). Activation map is overlaid on single subject T1 template.  23  3.3 Pre- and Post-training Comparison A paired t-test of the up-regulation versus down-regulation contrast images from post- and pre-training resulted in a significant cluster of activation in left RLPFC in the post versus pre-training contrast (Fig. 3.7). No significant activation was observed in the pre- versus post-training contrast.  Figure 3.7: Post-training versus pre-training contrast. Paired t-test of up-regulation vs. down-regulation contrast showing areas more active during the post-training than pre-training session MNI coordinates of 10 voxel cluster within the left RLPFC 34, 56,16 (t = 3.14, p < 0.01 uncorrected)  24  3.4 Individual Subject Results 3.4.1 Subject 1 Subject one completed a pre-training and 6 training sessions followed by a posttraining session. The activations maps for this subject revealed significant clusters of activations in the right RLPFC in pre-training session and most training sessions (Fig. 3.8). However, no significant positive correlation was observed between the task function and ROI as defined by the scanning session (Fig. 3.9). A significant negative correlation was observed between the task function and ROI in pre- and post-training and also first training sessions (Fig 3.9). Significant positive correlation was seen between the task function and global signal in the first training session. Global signal was negatively correlated with the task function in the third and fifth training sessions and also post-training session.  25  Region of Interest  Pre-training Session  First Training Session  Second Training Session  Third Training Session  Forth Training Session  Fifth Training Session  Sixth Training Session  Post-training Session  Figure 3.8: Subject 1 - activation maps Activation maps overlaid on subject’s non-normalized anatomical image. Voxels more active during up-regulation blocks are shown as red/yellow; voxels more active during down-regulation blocks are shown as blue/green. (p < 0.005, uncorrected).  26  0.15 0.11* 0.1  0.10  0.05 0.00 Correlation (r)  0 -0.01 -0.01 -0.05 -0.1  -0.09  Global -0.10  -0.12*  -0.15 -0.12* -0.2  ROI  -0.05  -0.17**  -0.17** -0.20**  -0.20** -0.21**  -0.25  -0.27**  -0.3 Pretraining  One  Two  Three  Four  Five  Six  Posttraining  Session  Figure 3.9: Subject 1 - correlations between task function and activation signal. Correlations between the task function convolved with the HRF, and the time courses extracted from the ROI as defined from the scanning session and the global signal. *p<0.05, **p<0.001  27  3.4.2 Subject 2 Subject 2 completed a pre-training and 5 training sessions followed by a posttraining session. The activation maps for this subject revealed significant clusters of activation in both right and left RLPFC in the third to fifth training session, and posttraining sessions (Fig. 3.10). A significant cluster of activation was observed in the left RLPFC in the second training session. Significant clusters of deactivation were seen in the right RLPFC in pre-training and first training sessions. Significant positive correlations were observed between the task function and ROI in the second to fifth training, and post-training sessions (Fig 3.11). Significant negative correlation was observed between the task function and ROI in pre-training and first training sessions. Global correlations followed the same pattern as ROI correlation in all sessions (Fig. 3.11).  28  Region of Interest  Pre-training Session  First Training Session  Second Training Session  Third Training Session  Forth Training Session  Fifth Training Session  Post-training Session  Figure 3.10: Subject 2 – activation maps Activation maps overlaid on subject’s non-normalized anatomical image. Voxels more active during up-regulation blocks are shown as red/yellow; voxels more active during down-regulation blocks are shown as blue/green. (p < 0.005, uncorrected).  29  0.8 0.68** 0.6  0.54**  0.35** 0.30**  Correlation (r)  0.4  0.34**  0.34** 0.26**  0.23**  0.34** 0.30** ROI  0.2  Global  0  -0.2 -0.23** -0.4  -0.14* -0.18**  -0.32** Pretraining  One  Two  Three  Four  Five  Posttraining  Session  Figure 3.11: Subject 2 - correlations between task function and activation signal. Correlations between the task function convolved with the HRF, and the time courses extracted from the ROI as defined from the scanning session and the global signal. *p<0.05, **p<0.001  30  3.4.3 Subject 3 Subject 3 completed a pre-training and 5 training sessions followed by a posttraining session. The activation maps for this subject revealed areas of deactivation within left and right RLPFC in pre-training, and all training sessions (Fig. 3.12). Significant negative correlations were observed between the task function and ROI in the pre-training and all training sessions, but not in the post-training session. Similarly, global signal was negatively correlated with the task function in the pre-training and all training sessions. (Fig. 3.13).  31  Region of Interest  Pre-training Session  First Training Session  Second Training Session  Third Training Session  Forth Training Session  Fifth Training Session  Post-training Session  Figure 3.12: Subject 3 – activation maps Activation maps overlaid on subject’s non-normalized anatomical image. Voxels more active during up-regulation blocks are shown as red/yellow; voxels more active during down-regulation blocks are shown as blue/green. (p < 0.005, uncorrected).  32  0.00  0  -0.07  -0.1 -0.15*  Correlation (r)  -0.2  -0.14* -0.21**  -0.24** -0.3  -0.22**  -0.23**  ROI  -0.30**  Global -0.33**  -0.4 -0.41** -0.5  -0.45**  -0.48** -0.53**  -0.6 Pretraining  One  Two  Three  Four  Five  Posttraining  Session  Figure 3.13: Subject 3 - correlations between task function and activation signal. Correlations between the task function convolved with the HRF, and the time courses extracted from the ROI as defined from the scanning session and the global signal. *p<0.05, **p<0.001  33  3.4.4 Subject 4 Subject 4 completed a pre-training and 4 training sessions followed by a posttraining session. The activation maps for this subject revealed areas of activation within right RLPFC in the pre-training session. Significant clusters of activation were observed in the left RLPFC during the 4 training sessions and post-training session (Fig. 3.14). Significant positive correlations were observed between the task function and ROI in the pre-training, third and fourth training and post-training sessions. Global signal was negatively correlated with the task function in the first and third training sessions, while there was a positive global correlation in the fourth training session. (Fig. 3.15).  34  Region of Interest  Pre-training Session  First Training Session  Second Training Session  Third Training Session  Forth Training Session  Post-training Session  Figure 3.14: Subject 4 – activation maps Activation maps overlaid on subject’s non-normalized anatomical image. Voxels more active during up-regulation blocks are shown as red/yellow; voxels more active during down-regulation blocks are shown as blue/green. (p < 0.005, uncorrected).  35  0.5 0.38**  0.4  0.33**  0.32**  Correlation (r)  0.3  0.26**  0.26**  0.18**  0.2  ROI  0.1  0.07  0.05  Global  0 -0.01  -0.02  -0.1 -0.14*  -0.14*  -0.2 Pre-training  One  Two  Three  Four  Post-training  Session  Figure 3.15: Subject 4 - correlations between task function and activation signal. Correlations between the task function convolved with the HRF, and the time courses extracted from the ROI as defined from the scanning session and the global signal. *p<0.05, **p<0.001  36  3.4.5 Subject 5 Subject 5 completed a pre-training and 4 training sessions followed by a posttraining session. The activation maps for this subject revealed areas of deactivation within right and left RLPFC in pre-training, all training, and post-training sessions. In addition, significant clusters of activation were observed in the left RLPFC during the pre-training and first training sessions (Fig. 3.16). A significant overall negative correlation was observed between the task function and ROI in the pre-training session, all training sessions and post-training session. Global signal was negatively correlated with the task function in the pre-training and all training sessions. (Fig. 3.17).  37  Region of Interest  Pre-training Session  First Training Session  Second Training Session  Third Training Session  Forth Training Session  Post-training Session  Figure 3.16: Subject 5 – activation maps Activation maps overlaid on subject’s non-normalized anatomical image. Voxels more active during up-regulation blocks are shown as red/yellow; voxels more active during down-regulation blocks are shown as blue/green. (p < 0.005, uncorrected).  38  0 -0.05 -0.1  -0.09 -0.11**  -0.11* Correlation (r)  -0.15 -0.2 -0.25  -0.17*  -0.18**  Global  -0.24** -0.26**  -0.3  ROI  -0.20** -0.24** -0.26**  -0.28**  -0.35 -0.4  -0.39**  -0.45 Pre-training  One  Two  Three  Four  Post-training  Session  Figure 3.17: Subject 5 - correlations between task function and activation signal. Correlations between the task function convolved with the HRF, and the time courses extracted from the ROI as defined from the scanning session and the global signal. *p<0.05, **p<0.001  39  3.4.6 Subject 6 Subject 6 completed a pre-training and 4 training sessions followed by a posttraining session. The activation maps for this subject revealed areas of activation within right and left RLPFC in the first, third and forth training sessions. In addition, significant clusters of deactivation were observed in the right and left RLPFC during the second, third and fourth training sessions, as well as the post-training session (Fig. 3.18). A significant positive correlation was observed between the task function and ROI in the first and third training sessions. ROI was negatively correlated with the task function in the post-training session. Global signal was negatively correlated with the task function in the pre- and post-training, while there was a positive global correlation in all training sessions. (Fig. 3.19).  40  Region of Interest  Pre-training Session  First Training Session  Second Training Session  Third Training Session  Forth Training Session  Post-training Session  Figure 3.18: Subject 6 – activation maps Activation maps overlaid on subject’s non-normalized anatomical image. Voxels more active during up-regulation blocks are shown as red/yellow; voxels more active during down-regulation blocks are shown as blue/green. (p < 0.005, uncorrected).  41  0.25 0.21**  0.21** 0.2  0.17**  0.15  0.12* 0.11*  0.10*  0.09  Correlation (r)  0.1 0.03  0.05  ROI Global  0 -0.05  -0.01  -0.1 -0.15 -0.2  -0.14* -0.15*  -0.17** Pre-training  One  Two  Three  Four  Post-training  Session  Figure 3.19: Subject 6 - correlations between task function and activation signal. Correlations between the task function convolved with the HRF, and the time courses extracted from the ROI as defined from the scanning session and the global signal. *p<0.05, **p<0.001  42  3.4.7 Subject 7 Subject 7 completed a pre-training and 4 training sessions followed by a posttraining session. The activation maps for this subject revealed areas of activation in the left RLPFC in the pre-training and third and fourth training sessions. In addition, significant clusters of deactivation were observed in the right RLPFC during the second, training session (Fig. 3.20). Significant positive correlations were observed between the task function and ROI in the pre-training, and third and fourth training sessions. Global signal was positively correlated with the task function in the pre- and post-training and second training sessions (Fig. 3.21).  43  Region of Interest  Pre-training Session  First Training Session  Second Training Session  Third Training Session  Forth Training Session  Post-training Session  Figure 3.20: Subject 7 – activation maps Activation maps overlaid on subject’s non-normalized anatomical image. Voxels more active during up-regulation blocks are shown as red/yellow; voxels more active during down-regulation blocks are shown as blue/green. (p < 0.005, uncorrected).  44  0.25  0.23**  0.2 0.16*  0.16*  Correlation (r)  0.15  0.17**  0.13* 0.11*  0.10  0.1 ROI 0.05 0.04  0.05  0.05  Global  0.03  0  -0.05  -0.05  -0.1 Pre-training  One  Two  Three  Four  Post-training  Session  Figure 3.21: Subject 7 - correlations between task function and activation signal. Correlations between the task function convolved with the HRF, and the time courses extracted from the ROI as defined from the scanning session and the global signal. *p<0.05, **p<0.001  45  3.5 Subject Debriefing 3.5.1 Subjective Ratings In order to test for the possibility of differences among subjects with regards to their perception of task difficulty, and feedback helpfulness both in general and also for up-regulation and down-regulation tasks separately, analyses of variance was used with subject and regulation condition used as independent and task difficulty, and feedback helpfulness used as dependent factors. The result showed a significant difference among subjects in terms of the degree to which they found the feedback helpful in performing the tasks, F(5,40) = 6.53, p < 0.001. Also subjects in general found the feedback more helpful for up-regulation, compared to down-regulation, F(1,40) = 7.46, p < 0.05. Moreover, there was a significant difference amongst subjects in their perception of task difficulty, F(5,64) = 7.55, p < 0.001. Also all subjects found the up-regulation task more difficult than the down-regulation task (see table 3.1), F(1,64) = 68.76, p < 0.001. There was also a significant subject by regulation interaction for the task difficulty, F(5,64) = 3.25, p < 0.05. This significant interaction was followed up by paired t-tests on the individual subject data. For subject two, no significant difference was observed in task difficulty between up- and down-regulation tasks, t(6) = 0.19, p > 0.05. Subject three, in contrast, found up-regulation significantly more difficult than down-regulation, t(6) = 3.87, p < 0.05. Similarly, for subject four, up-regulation was reported as significantly more difficult than down-regulation, t(5) = 4.04, p <0.05. The result for subject five also showed significantly more difficulty with up-regulation as compared to down-regulation, t(5) = 8.17, p < 0.001. The result was similar for subject six and subject seven, t(5) = 5.97, p < 0.05, and t(5) = 9.22, p < 0.001, respectively.  46  Table 3.4: Mean subjective rating scores from debriefing sessions  Subject*  Task Difficulty Up-Regulate Down-Regulate  Feedback Helpfulness Up-Regulate Down-Regulate  Two  4.57(0.98)  4.43(1.51)  2.80(1.30)  3.40(1.14)  Three  6.14(0.90)  4.71(0.95)  5.80(0.84)  5.60(0.89)  Four  6.33(0.52)  3.83(1.17)  3.00(1.41)  2.75(1.50)  Five  5.50(1.05)  3.00(1.10)  6.00(2.00)  4.25(1.50)  Six  4.17(1.17)  2.33(0.52)  4.50(1.73)  2.50(1.00)  Seven  5.67(0.52)  2.83(0.75)  6.25(1.5)  3.50(1.73)  Mean  5.39(1.15)  3.58(1.33)  4.69(1.96)  3.73(1.59)  Note: Rating scales ranged from 1 – 7 with larger values indicating higher levels. * No data is available for subject one.  3.5.2 Subjective Experience 3.5.2.1 Thought Content and Labelling When subjects were asked about the content of their thoughts and the words they used to label them during up-regulation, most of them mentioned words and phrases such as “future plans”, “memories”. Subject 2 and 3 also reported “emotional thoughts” and “feelings” respectively, whereas subject 6 engaged in “logic puzzles” and “math equations”, and subject 7 reported used “curious and abstract thoughts” among others. For the complete list of individual thought contents see Table 3.2.  47  Table 3.5: Thought contents and labels used by subjects during up-regulation.  Subjects Subject two Subject three Subject four  Subject five  Subject six  Subject seven  Thought Contents and Labels “memories”, “emotional thoughts”, “future thoughts”, “people”, and “current events” “future plans”, “feelings”, “memories”, “past events”, “past experiences”, and “experiences” “picturing”, “planning”, “coaching my self”, “trying”, “planning to discuss about my thoughts”, “imagining”, “remembering”, “couldn’t find a label”, and “fantasizing” “questioning”, “thinking about thinking”, “evaluating of my performance”, “assessing negative or positive thoughts”, “verbal thoughts”, “anticipation”, “hopefulness”, “awareness”, “chattering to my self”, and “awareness of my body” “memories”, “future plans”, “recent events”, fantasies”, “future ambitions”, today’s events”, “dreams”, “logic puzzles”, “ monologues”, “logic problem”, “math equations”, “daydreams”, “future predictions” “curious, and abstract”, “contemplating”, “thinking about the graphs”, “memories”, and “contemplating my thoughts”  3.5.2.2 Feedback In general subjects found the feedback panels helpful for regulating ROI activity, but distracting from task performance. Some reported that focusing on the feedback would influence its direction (subject three and seven), while others experienced some degrees of frustration and distraction when they received negative feedback (subject four, five and six). Subject two found the feedback panels distracting; nevertheless, they used the thermometer occasionally to monitor their performance. They also found the posttraining session more difficult than the previous sessions for not receiving any feedback. Subject three also found the feedback distracting; they also noticed that focusing on the feedback itself would affect the activation level. Subject four found the feedback panels helpful. They used the history bar graph more often since it was motivating and enabled  48  them to focus on the task. Subject four also found the continuous feedback panels distracting when they went to the opposite direction of what expected. Subject five used both the history bar graph and the thermometer throughout the scanning session. They reported some degree of frustration when receiving the opposite feedback through the thermometer. Subject six tried not to focus on the feedback panels, since they received negative feedback. Subject seven found the feedback helpful, but distracting at times. In some sessions they tried not to focus on the thermometer since “thinking about it made it go the wrong way”. They also noticed that using the feedback panels “the more you think about what you think the easier it is to control”.  49  4 Discussion The results of this study provide support for the first hypothesis that RLPFC activation is linked to the evaluation of internally-generated thought processes. Group analyses demonstrated significant clusters of activation in left RLPFC when subjects were reflecting on their own thought processes compared to when they were engaged in perceptual/body awareness during all sessions (pre-training, four training, and posttraining sessions). These findings are in line with previous neuroimaging studies showing that RLPFC is activated during tasks requiring mentalizing (Gilbert et al., 2006) and internally-oriented attention (Christoff, Ream, Geddes, & Gabrieli, 2003). The results are also consistent with the findings of anatomical studies suggesting that by not interacting with low-level perceptual areas, RLPFC does not regulate attention toward external stimuli (Petrides & Pandya, 2007). The results also provide evidence in support of the second hypothesis that healthy volunteers can achieve enhanced modulation of their RLPFC activation through a course of real-time fMRI feedback training. Group analysis of individual sessions revealed a gradual increase in the extent of left RLPFC activation throughout the feedback training. Also a direct comparison of post-training and pre-training sessions showed a significant cluster of greater activation within left RLPFC during post-training session, suggesting the persistence of the feedback training effect in the absence of feedback. Despite these encouraging results at the group level, single-subject analyses revealed considerable individual differences in subjects’ ability to activate their RLPFC. In particular, subject 3 and 5 showed the opposite pattern of expected activation (deactivation of RLPFC when directing their attention internally) during all scanning  50  sessions. This was in the face of relative improvement (lesser deactivation in RLPFC) between pre- and post-training sessions. Comparisons of personality measures between these two subjects and others showed that both of them had higher rumination scores than the rest (although no statistical difference in rumination scores was observed between subject 3 and 5 (M = 4.09), and other subjects (M = 2.95); t(5) = 1.77, ns). High scores in rumination which has been defined as “neurotic self-attentiveness” (as opposed to “intellectual self-attentiveness” or “reflection”) have been associated with depression and anxiety (Trapnell & Campbell, 1999). It has been suggested that in depressed patients, rumination causes cognitive deficits through occupying central executive resources (Watkins & Brown, 2002). Rumination has also been associated with perseveration and cognitive inflexibility (Davis & Nolen-Hoeksema, 2000). In one study (Davis & NolenHoeksema, 2000), ruminators exhibited impaired performance on the Wisconsin Cart Sorting Task (WCST), a task that requires cognitive flexibility and feedback evaluation, and has been found to elicit activation in RLPFC (Berman et al., 1995; Goldberg et al., 1998; Nagahama et al., 1996). The authors concluded that “Ruminators become mentally stuck in a style of relating to the environment even when the adaptiveness of that style has been invalidated by negative feedback” (Davis & Nolen-Hoeksema, 2000). Hence, while other factors may have been involved, it is possible that subjects with high rumination scores had difficulty switching between cognitive strategies, and making effective use of the feedback. All subjects found the up-regulation condition more difficult than downregulation. This finding is consistent with previous findings which showed that RLPFC activation is associated with increasing task difficulty (Baker et al., 1996; Christoff &  51  Owen, 2006; Dagher, Owen, Boecker, & Brooks, 1999; van den Heuvel et al., 2003) However, it has been shown that task difficulty alone cannot account for the activation of RLPFC (Christoff, Keramatian, Gordon, Smith, & Mädler, in press; Christoff, Ream, Geddes, & Gabrieli, 2003; Koechlin, Basso, Pietrini, Panzer, & Grafman, 1999). It is therefore unlikely that difficulty alone can explain the differential activation observed between the two conditions. It is possible that the improved RLPFC modulation observed in this study is at least partially attributed to factors such as the practice effect, and not necessarily to the presentation of real-time information. Control conditions using an identical training paradigm but with sham feedback from other region of the brain, or using no feedback at all are underway to verify the effect of real-time feedback training. Indeed previous realtime fMRI studies using such control conditions reported an enhanced learning effect in subjects who received genuine feedback from the ROI, and not in subjects in the control groups (Caria et al., 2007; deCharms et al., 2004; deCharms et al., 2005). In addition to RLPFC, a few other brain regions including anterior insula, medial prefrontal cortex, and precuneus showed increased activation during up-regulation versus down-regulation. This co-activation might indicate an existing pattern of functional connectivity (defined as temporal correlation between anatomically remote areas; (Friston et al. 1993; McIntosh, 2000) between RLPFC and those brain areas during introspective thought processes. Also, these findings confirm previous fMRI experiments implicating a distributed network of brain regions (also known as the “default mode network”; Gusnard and Raichle, 2001), including medial prefrontal cortex and precuneus, in stimulus-independent thought processes (Mason et al., 2007). More recently, Christoff  52  and colleagues observed a parallel recruitment of default network and executive network (including RLPFC) regions during introspective thought processes (Christoff et al., 2009). Further data analysis is in progress to elucidate the exact relation between activation in RLPFC and other brain areas.  4.1 ROI Definition As discussed earlier, several theories exist regarding the function of RLPFC. The investigation of these theories requires a proper definition of this brain region. In this study, ROIs were defined by using two different anatomical approaches: For subject 1, 2, 3 and 5, ROIs were defined bilaterally as the region of intersection between Brodmann area 10 and the middle frontal gyrus, as done by Christoff et al (2003). For these subjects, ROIs were specified on a normalized brain image transferred into the subjects’ anatomical space. To avoid the inaccurate overlap of functional regions caused by spatial normalization (Brett, Johnsrude, & Owen, 2002), ROIs from the other subjects (subjects 4, 6, and 7) were selected visually using anatomical landmarks identified on high-resolution anatomical images of each individual subject’s brain. The latter approach has been used in one of the previous real-time fMRI experiments targeting other brain regions (Caria et al., 2007). While providing an unbiased and relatively straightforward approach to defining the ROI, anatomical definitions, especially for the purpose of real-time feedback training, can impose some limitations: First, there is no consensus on the anatomical boundaries of RLPFC; with definitions ranging from “any activations that occur within BA10” to “the intersection between BA10 and middle frontal gyrus” (Smith, Keramatian, & Christoff,  53  2007). Second, ROI definition solely based on anatomical information does not take into account the possible functional subdivisions within an anatomically-defined region. A recent meta-analysis of functional neuroimaging studies has suggested the possibility of functional variation within RLPFC, with relatively more caudal subregions within RLPFC being activated in mentalizing (reflection on one’s own mental states) and more rostral subregions being activated in multiple task coordination (Gilbert et al., 2006). Third, anatomical definition of ROI does not allow for the possibility of hemispheric lateralization of activation within a given ROI. There is no hypothesis about RLPFC lateralization (to the best of my knowledge), nevertheless, the results of individual subjects activation maps in this study show that while RLPFC activation in some subjects was bilateral, it was predominantly left-sided in other subjects. One alternative approach to address the above-mentioned limitations is to define the ROI based on a combination of anatomical and functional information. In this approach, individual subjects perform a task known to activate the ROI (functional localizer task); the resultant activation map combined with anatomical information from the region is then used in order to identify significantly activated voxels that fall within the anatomically defined ROI. These voxels form the new functionally-anatomically defined ROI. There are two possible ways to design a functional localizer task for a realtime fMRI study. The first method would be to use the first real-time session with no feedback, a method that was used by deCharms and colleagues to localize the rACC (deCharms et al., 2005). One drawback of using this method for functionally localizing high-order areas of association cortex such as RLPFC is that due to the inherent difficulty of gaining control over the activation of these regions, there is a high chance that the ROI  54  will not be activated in the first real-time session, as was the case for a few subjects in this study. One way to overcome this difficulty would be to specify a full scanning day for each subject to define an individualized functional ROI prior to the real-time fMRI training. Another concern with using the first real-time session with no feedback is that (as the results in this study show) the extent of activation within the ROI may increase throughout the course of feedback training; as a result, specifying the ROI based on the first training session might cause subjects to receive less than optimal feedback. An alternative functional localizing method is to use a separate task that has been shown to activate the ROI as a functional localizer task. Weiskopf and colleagues (2004) used this method to localize the supplementary motor area (SMA) and the parahippocampal place area (PPA) prior to the real-time feedback training sessions. In a block design fMRI experiment, Smith and colleagues (2007) employed a simple reasoning task with the aim of functionally localizing RLPFC. The results showed consistent activation of RLPFC in all eleven subjects. Altogether, although anatomical definition of ROI provides an unbiased and less time-consuming localizing method for the real-time fMRI training, ROI definition based on the combination of functional and anatomical information could offer a more individualized, subject-specific approach. Since real-time fMRI is a relatively new area of research, additional studies are needed to determine which localizing method produces the most desirable ROI for feedback training.  55  4.2 Feedback Display Subjects received both continuous and discrete forms of feedback from the level of activation in their left and right RLPFC. A combination of continuous and discrete feedback was used in order to allow subjects to receive instantaneous feedback during task performance, and also to provide them with information regarding their long-term performance trends. Continuous feedback for subject one to four was presented in a scrolling time-course graph which has previously been used by deCharms et al (2005), and also a thermometer bar graph adapted from Caria et al (2007). The scrolling graph panel was removed for subject five to seven due to distraction reported by previous subjects (see below). Although no real-time fMRI study comparing discrete and continuous feedback has been reported, EEG feedback training experiments suggest that using continuous feedback leads to superior performance in relatively simple tasks such as an imagined motor task (Guger et al., 2001). However, evidence for such superior performance in higher cognitive functions is lacking. Theoretically, continuous feedback can be beneficial to task performance by maintaining the subject’s interest in the task, and providing continual motivation. Conversely, continuous feedback can be detrimental to performance by distracting the subject from the task at hand (McFarland, McCane, & Wolpaw, 1998). This is of particular concern in higher cortical regions for which successful regulation may require substantial attentional resources. Most subjects in the present study reported being distracted by continuous feedback while trying to perform the task. In addition, strategies used to regulate the region can interfere with attending to the feedback thereby causing further deleterious effects on task performance. In the present study, while during up-regulation subjects are supposed to direct their attention to  56  their own thoughts, simultaneous attention to the visual aspects of the feedback could down-regulate the region. During down-regulation, subjects may engage in introspective thought processes by thinking about the meaning of the feedback which in turn could cause increased activation in RLPFC. One study using a Tower of London task showed that receiving feedback on performance is associated with increased activation in right RLPFC (Elliott, Frith, & Dolan, 1997). In the present study, two subjects (subject three and seven) noticed that focusing on the feedback itself can affect the activation level. Finally, emotional responses can potentially interfere with task performance, and affect the activation; for instance, undesired feedback might elicit frustration (as in subject four, five, and six), whereas positive feedback could lead the subject to anticipate hitting the target (McFarland, McCane, & Wolpaw, 1998). One way to minimize potential issues with continuous feedback from higher cognitive regions is to use a discrete feedback method. In a real-time fMRI study by Yoo and colleagues, subjects were presented with functional brain activation maps every 60 seconds following performing a hand movement task and rest. All subjects showed an increase in the extent of activation in somatosensory and motor areas following the fMRI feedback training (Yoo & Jolesz, 2002). In another experiment by Posse et al, subjects received feedback from intensity and extent of amygdala activation at the end of 30second blocks of sadness induction (Posse et al., 2003). Feedback in that study was used solely to reinforce sadness induction; it is not clear, therefore, whether feedback affected brain activation. In the present study, bar graphs indicating average ROI activation during blocks were presented at the end of each block. This feedback panel design allowed subjects to assess their performance during the previous block and avoided  57  interference with the task. The bar graphs remained on the screen throughout the whole session providing subjects with information regarding their long-term performance trend. Since all subjects received both continuous and discrete forms of feedback during all sessions, it is difficult to ascertain which form of feedback is more beneficial. Future studies are needed to directly compare the effect of discrete and continuous feedback on real-time fMRI training outcome.  4.3 Sample Size A total of ten subjects were recruited to participate in this study and seven subjects were included in the final data analysis. The number of subjects who completed real-time feedback training in previous studies ranges from 1 to 11. This range of sample size is significantly lower than recent conventional fMRI studies, most likely due to the novelty of real-time fMRI technique, the exploratory nature of many of these studies, technical complications with real-time feedback training and possibly higher rate of subject dropout. Because of a relatively small sample size of the current study, a fixed effects analysis at the group level was performed. As such, the results of the group analysis are limited since no inference can be made regarding the population. A random effects analysis is a more suitable analysis for which a minimum of 8 to 16 subjects is needed to achieve sufficient statistical power to detect an effect (Friston, Holmes, & Worsley, 1999).  58  4.4 Global Signal Change Two subjects were excluded from the study due to extreme global signal-task correlations. Milder degrees of such global signal modulation were observed in some sessions of most included subjects. Changes in respiration during scanning have been shown to cause global changes in BOLD signal intensity (Birn, Diamond, Smith, & Bandettini, 2006). Two distinct mechanisms have been proposed to account for the effect of breathing on BOLD signal: fluctuation of the magnetic field strength due to the bulk motion of the thoracic cavity (Glover, Li, & Ress, 2000; Raj, Anderson, & Gore, 2001), and also alterations in cerebral blood flow caused by changes in blood CO2 concentration (Kastrup, Li, Glover, & Moseley, 1999). It has been shown that intentional breath-holds of as little as 3 to 6 seconds can result in increased activation in brain regions such as ACC, PCC, insula and caudate (Abbott, Opdam, Briellmann, & Jackson, 2005), whereas longer duration breath-holds cause more widespread, global signal increase in grey matter (Li, Kastrup, Takahashi, & Moseley, 1999). Of particular importance is the possibility of subconscious task-correlated respiratory changes thought to more likely occur when the task requires more concentration or effort (Abbott, Opdam, Briellmann, & Jackson, 2005), or when no explicit task is being performed (rest condition) (Birn, Diamond, Smith, & Bandettini, 2006). The resultant task-correlated signal modulations can be a serious confounding factor in real-time fMRI studies (Weiskopf, Scharnowski et al., 2004), as subjects might learn to modify their breathing pattern to gain control over the feedback they receive. There are some potential ways to address this issue: First, subjects should be informed regarding the potential confounding effect of respiration pattern on the signal  59  and be instructed to breathe normally throughout the scanning session. Second, information regarding depth and rate of respiration during the scanning session can be obtained by placing a respiration belt around subjects’ abdomen (Birn, Diamond, Smith, & Bandettini, 2006). Such information can then be used by experimenters to warn subjects in case a task-correlated respiration pattern is observed. Alternatively, subjects themselves can be provided with respiratory information in real-time in order to become self-conscious regarding their breathing pattern and its potential correlation with the task. Finally, global signal can be presented to subjects as an additional form of feedback; subjects would then be instructed to attempt to modulate the ROI activation while trying to keep the global activation relatively constant using both forms of feedback. Yet another solution might be to present subjects with a differential feedback calculated by subtracting the ROI signal from a large background signal to increase the feedback specificity by cancelling out the global signal effect (deCharms et al., 2004; Weiskopf, Scharnowski et al., 2004; Weiskopf et al., 2003).  4.5 Future Direction This work represents the first application of real-time fMRI technique to investigate the functional characteristics of RLPFC. Future research should address some of the methodological shortcomings of the present study such as small sample size, the confounding effect of respiratory changes, and the lack of control groups to ensure that the observed improvement is not due to practice alone. In addition, the differential effect of receiving continuous and discrete feedback on subjects’ performance should be directly investigated to determine the most beneficial form of feedback for higher  60  cognitive regions such as RLPFC. Finally, as all real-time fMRI experiments have studied only the short term effects of training, the possible long term effect of training is unclear. As such, follow-up scanning sessions could be used to investigate the long term effects of real-time fMRI feedback training on subjects.  4.6 Conclusion The results of this study expand our understanding of the role of RLPFC in human cognition by providing further evidence for the involvement of this region in introspective evaluation of thought processes. Although the overall results suggest the feasibility of using real-time fMRI feedback training to achieve enhanced modulation of RLPFC, considerable individual differences, small sample size, and lack of control groups preclude any definitive conclusions in this regards. As such, the results serve as a starting point for ongoing and future real-time fMRI studies of higher cognitive regions, especially RLPFC. In addition, the results raise important questions with respect to the methodological issues in real-time fMRI experiments such as global signal modulation and potential undesirable effects of feedback on task performance. By highlighting these issues and furthering our knowledge of how they may influence the implementation and outcome of real-time fMRI training experiments, this work provides a valuable contribution to existing research in this field.  61  5 Bibliography Abbott, D. F., Opdam, H. I., Briellmann, R. S., & Jackson, G. D. (2005). Brief breath holding may confound functional magnetic resonance imaging studies. Hum Brain Mapp, 24(4), 284-290. Baer, R. A., Smith, G. T., & Allen, K. B. (2004). Assessment of Mindfulness by SelfReport: The Kentucky Inventory of Mindfulness Skills. Assessment, 11(3), 191. Baker, S. C., Rogers, R. D., Owen, A. M., Frith, C. D., Dolan, R. J., Frackowiak, R. S., et al. (1996). Neural systems engaged by planning: a PET study of the Tower of London task. Neuropsychologia, 34(6), 515-526. Barsalou, L. W. (1999). Perceptual symbol systems. Behavioral and Brain Sciences, 22(4), 577. Benet-Martinez, V., & John, O. P. (1998). Los Cinco Grandes across cultures and ethnic groups: Multitrait-multimethod analyses of the Big Five in Spanish and English. Journal of Personality and Social Psychology, 75(3), 729. Berman, K. F., Ostrem, J. L., Randolph, C., Gold, J., Goldberg, T. E., Coppola, R., et al. (1995). Physiological activation of a cortical network during performance of the Wisconsin Card Sorting Test: a positron emission tomography study. Neuropsychologia, 33(8), 1027-1046. Birn, R. M., Diamond, J. B., Smith, M. A., & Bandettini, P. A. (2006). Separating respiratory-variation-related fluctuations from neuronal-activity-related fluctuations in fMRI. Neuroimage, 31(4), 1536-1548. Brett, M., Johnsrude, I. S., & Owen, A. M. (2002). The problem of functional localization in the human brain. Nat Rev Neurosci, 3(3), 243-249. Bunge, S. A., Helskog, E. H., & Wendelken, C. (2009). Left, but not right, rostrolateral prefrontal cortex meets a stringent test of the relational integration hypothesis NeuroImage, 46(1), 338-342. Burgess, P. W., Dumontheil, I., & Gilbert, S. J. (2007). The gateway hypothesis of rostral prefrontal cortex (area 10) function. Trends Cogn Sci, 11(7), 290-298. Burgess, P. W., Quayle, A., & Frith, C. D. (2001). Brain regions involved in prospective memory as determined by positron emission tomography. Neuropsychologia, 39(6), 545-555.  62  Caria, A., Veit, R., Sitaram, R., Lotze, M., Weiskopf, N., Grodd, W., et al. (2007). Regulation of anterior insular cortex activity using real-time fMRI. Neuroimage, 35(3), 1238-1246. Christoff, K., & Gabrieli, J. D. (2000). The frontopolar cortex and human cognition: Evidence for a rostrocaudal hierarchical organization within the human prefrontal cortex. Psychobiology, 28(2), 168. Christoff, K., Gordon, A. M., Smallwood, J., Smith, R., & Schooler, J. W. (2009). Experience sampling during fMRI reveals default network and executive system contributions to mind wandering. Proc Natl Acad Sci U S A, 106(21), 8719-8724. Christoff, K., & Keramatian, K. (2008). Abstraction of mental representations: theoretical considerations and neuroscientific evidence. In S. A. Bunge & J. D. Wallis (Eds.), Neuroscience of rule-guided behavior (pp. 107-126). New York: Oxford University Press. Christoff, K., Keramatian, K., Gordon, A. M., Smith, R., & Mädler, B. (in press). Prefrontal organization of cognitive control according to levels of abstraction. Brain research. Christoff, K., & Owen, A. M. (2006). Improving reverse neuroimaging inference: cognitive domain versus cognitive complexity. Trends Cogn Sci, 10(8), 352-353. Christoff, K., Prabhakaran, V., Dorfman, J., Zhao, Z., Kroger, J. K., Holyoak, K. J., et al. (2001). Rostrolateral prefrontal cortex involvement in relational integration during reasoning. Neuroimage, 14(5), 1136-1149. Christoff, K., Ream, J. M., Geddes, L. P., & Gabrieli, J. D. (2003). Evaluating selfgenerated information: anterior prefrontal contributions to human cognition. Behav Neurosci, 117(6), 1161-1168. Dagher, A., Owen, A. M., Boecker, H., & Brooks, D. J. (1999). Mapping the network for planning: a correlational PET activation study with the Tower of London task. Brain, 122 ( Pt 10), 1973-1987. Damasio, A. R. (1996). The somatic marker hypothesis and the possible functions of the prefrontal cortex. Philos Trans R Soc Lond B Biol Sci, 351(1346), 1413-1420. Davis, R. N., & Nolen-Hoeksema, S. (2000). Cognitive inflexibility among ruminators and nonruminators. Cognitive Therapy and Research, 24(6), 699. deCharms, R. C. (2008). Applications of real-time fMRI. Nat Rev Neurosci, 9(9), 720729. deCharms, R. C., Christoff, K., Glover, G. H., Pauly, J. M., Whitfield, S., & Gabrieli, J. D. (2004). Learned regulation of spatially localized brain activation using realtime fMRI. Neuroimage, 21(1), 436-443.  63  deCharms, R. C., Maeda, F., Glover, G. H., Ludlow, D., Pauly, J. M., Soneji, D., et al. (2005). Control over brain activation and pain learned by using real-time functional MRI. Proc Natl Acad Sci U S A, 102(51), 18626-18631. den Ouden, H. E., Frith, U., Frith, C., & Blakemore, S. J. (2005). Thinking about intentions. Neuroimage, 28(4), 787-796. Duncan, J., Burgess, P., & Emslie, H. (1995). Fluid intelligence after frontal lobe lesions. Neuropsychologia, 33(3), 261-268. Duncan, J., & Owen, A. M. (2000). Common regions of the human frontal lobe recruited by diverse cognitive demands. Trends Neurosci, 23(10), 475-483. Elliott, R., Frith, C. D., & Dolan, R. J. (1997). Differential neural response to positive and negative feedback in planning and guessing tasks. Neuropsychologia, 35(10), 1395-1404. Friston, K. J., Holmes, A. P., & Worsley, K. J. (1999). How many subjects constitute a study? Neuroimage, 10(1), 1-5. Friston, K. J., Holmes, A. P., Worsley, K. J., Poline, J. P., Frith, C. D., & Frackowiak, R. S. J. (1995). Statistical parametric maps in functional imaging: A general linear approach. Hum Brain Mapping, 2, 189:210. Friston, K. J., Frith, C. D., Liddle, P. F., & Frackowiak, R. S. (1993). Functional connectivity: the principal-component analysis of large (PET) data sets. J Cereb Blood Flow Metab, 23(10), 475-483. Fuster, J. M. (1980). The prefrontal cortex: Anatomy, physiology, and neuropsychology of the frontal lobe. New York: Raven. Gilbert, S. J., Spengler, S., Simons, J. S., Steele, J. D., Lawrie, S. M., Frith, C. D., et al. (2006). Functional specialization within rostral prefrontal cortex (area 10): a meta-analysis. J Cogn Neurosci, 18(6), 932-948. Glover, G. H., Li, T. Q., & Ress, D. (2000). Image-based method for retrospective correction of physiological motion effects in fMRI: RETROICOR. Magn Reson Med, 44(1), 162-167. Goldberg, T. E., Berman, K. F., Fleming, K., Ostrem, J., Van Horn, J. D., Esposito, G., et al. (1998). Uncoupling cognitive workload and prefrontal cortical physiology: a PET rCBF study. Neuroimage, 7(4 Pt 1), 296-303. Goldman-Rakic, P. S. (1987). Circuitry of primate prefrontal cortex and regulation of behavior by representational memory. In F. P. V. B. Mountcastle (Ed.), Handbook of physiology: Sec. 1. The nervous system (Vol. 5. Higher functions of the brain., pp. 373-417). Bethesda, MD: American Physiological Society.  64  Guger, C., Schlogl, A., Neuper, C., Walterspacher, D., Strein, T., & Pfurtscheller, G. (2001). Rapid prototyping of an EEG-based brain-computer interface (BCI). IEEE Trans Neural Syst Rehabil Eng, 9(1), 49-58. Gusnard, D. A. & Raichle, M. E. (2001). Searching for a baseline: functional imaging and the resting human brain. Nat Rev Neurosci, 2(10) 685-694. Jonides, J., Schumacher, E. H., Smith, E. E., Lauber, E. J., Awh, E., Minoshima, S., et al. (1997). Verbal working memory load affects regional brain activation as measured by PET. Journal of Cognitive Neuroscience, 9, 462-475. Kastrup, A., Li, T. Q., Glover, G. H., & Moseley, M. E. (1999). Cerebral blood flowrelated signal changes during breath-holding. AJNR Am J Neuroradiol, 20(7), 1233-1238. Koechlin, E., Basso, G., Pietrini, P., Panzer, S., & Grafman, J. (1999). The role of the anterior prefrontal cortex in human cognition. Nature, 399(6732), 148-151. Koechlin, E., Ody, C., & Kouneiher, F. (2003). The architecture of cognitive control in the human prefrontal cortex. Science, 302(5648), 1181-1185. Li, T. Q., Kastrup, A., Takahashi, A. M., & Moseley, M. E. (1999). Functional MRI of human brain during breath holding by BOLD and FAIR techniques. Neuroimage, 9(2), 243-249. Mason, M. F., Norton, M. I., Van Horn, J. D., Wegner, D. M., Grafton, S. T. & Macrae, C. N. (2007). Wandering minds: the default network and stimulus-independent thought. Science, 315(5810), 393-395. McFarland, D. J., McCane, L. M., & Wolpaw, J. R. (1998). EEG-based communication and control: short-term role of feedback. IEEE Trans Rehabil Eng, 6(1), 7-11. McIntosh, A. R. (2000) Towards a network theory of cognition. Neural Netw, 13(8-9), 861-870. Miller, E. K., & Cohen, J. D. (2001). An integrative theory of prefrontal cortex function. Annu Rev Neurosci, 24, 167-202. Nagahama, Y., Fukuyama, H., Yamauchi, H., Matsuzaki, S., Konishi, J., Shibasaki, H., et al. (1996). Cerebral activation during performance of a card sorting test. Brain, 119 ( Pt 5), 1667-1675. Okuda, J., Fujii, T., Yamadori, A., Kawashima, R., Tsukiura, T., Fukatsu, R., et al. (1998). Participation of the prefrontal cortices in prospective memory: evidence from a PET study in humans. Neurosci Lett, 253(2), 127-130. Ongur, D., Ferry, A. T., & Price, J. L. (2003). Architectonic subdivision of the human orbital and medial prefrontal cortex. J Comp Neurol, 460(3), 425-449. 65  Petrides, M. (1991). Monitoring of selections of visual stimuli and the primate frontal cortex. Proc Biol Sci, 246(1317), 293-298. Petrides, M., & Pandya, D. N. (2004). The frontal cortex. In G. Paxinos & J. K. Mai (Eds.), The human nervous system (2nd ed., pp. 950-972). San Diego: Elsevier Academic Press. Petrides, M., & Pandya, D. N. (2007). Efferent association pathways from the rostral prefrontal cortex in the macaque monkey. J Neurosci, 27(43), 11573-11586. Posse, S., Fitzgerald, D., Gao, K., Habel, U., Rosenberg, D., Moore, G. J., et al. (2003). Real-time fMRI of temporolimbic regions detects amygdala activation during single-trial self-induced sadness. Neuroimage, 18(3), 760-768. Pruessmann, K. P., Weiger, M., Scheidegger, M. B., & Boesiger, P. (1999). SENSE: sensitivity encoding for fast MRI. Magn Reson Med, 42(5), 952-962. Raj, D., Anderson, A. W., & Gore, J. C. (2001). Respiratory effects in human functional magnetic resonance imaging due to bulk susceptibility changes. Phys Med Biol, 46(12), 3331-3340. Ramnani, N., & Owen, A. M. (2004). Anterior prefrontal cortex: insights into function from anatomy and neuroimaging. Nat Rev Neurosci, 5(3), 184-194. Rolls, E. T. (1996). The orbitofrontal cortex. Philos Trans R Soc Lond B Biol Sci, 351(1346), 1433-1443; discussion 1443-1434. Rypma, B., Prabhakaran, V., Desmond, J. E., Glover, G. H., & Gabrieli, J. D. (1999). Load-dependent roles of frontal brain regions in the maintenance of working memory. Neuroimage, 9(2), 216-226. Semendeferi, K., Armstrong, E., Schleicher, A., Zilles, K., & Van Hoesen, G. W. (2001). Prefrontal cortex in humans and apes: a comparative study of area 10. Am J Phys Anthropol, 114(3), 224-241. Simons, J. S., Scholvinck, M. L., Gilbert, S. J., Frith, C. D., & Burgess, P. W. (2006). Differential components of prospective memory? Evidence from fMRI. Neuropsychologia, 44(8), 1388-1397. Smith, R., Keramatian, K., & Christoff, K. (2007). Localizing the rostrolateral prefrontal cortex at the individual level. Neuroimage, 36(4), 1387-1396. Trapnell, P. D., & Campbell, J. D. (1999). Private self-consciousness and the five-factor model of personality: Distinguishing rumination from reflection. Journal of Personality and Social Psychology, 76(2), 284.  66  Tulving, E., Markowitsch, H. J., Craik, F. E., Habib, R., & Houle, S. (1996). Novelty and familiarity activations in PET studies of memory encoding and retrieval. Cereb Cortex, 6(1), 71-79. van den Heuvel, O. A., Groenewegen, H. J., Barkhof, F., Lazeron, R. H., van Dyck, R., & Veltman, D. J. (2003). Frontostriatal system in planning complexity: a parametric functional magnetic resonance version of Tower of London task. Neuroimage, 18(2), 367-374. Velanova, K., Jacoby, L. L., Wheeler, M. E., McAvoy, M. P., Petersen, S. E., & Buckner, R. L. (2003). Functional-anatomic correlates of sustained and transient processing components engaged during controlled retrieval. J Neurosci, 23(24), 8460-8470. Watkins, E., & Brown, R. G. (2002). Rumination and executive function in depression: an experimental study. J Neurol Neurosurg Psychiatry, 72(3), 400-402. Weiskopf, N., Mathiak, K., Bock, S. W., Scharnowski, F., Veit, R., Grodd, W., et al. (2004). Principles of a brain-computer interface (BCI) based on real-time functional magnetic resonance imaging (fMRI). IEEE Trans Biomed Eng, 51(6), 966-970. Weiskopf, N., Scharnowski, F., Veit, R., Goebel, R., Birbaumer, N., & Mathiak, K. (2004). Self-regulation of local brain activity using real-time functional magnetic resonance imaging (fMRI). J Physiol Paris, 98(4-6), 357-373. Weiskopf, N., Veit, R., Erb, M., Mathiak, K., Grodd, W., Goebel, R., et al. (2003). Physiological self-regulation of regional brain activity using real-time functional magnetic resonance imaging (fMRI): methodology and exemplary data. Neuroimage, 19(3), 577-586. Yoo, S. S., & Jolesz, F. A. (2002). Functional MRI for neurofeedback: feasibility study on a hand motor task. Neuroreport, 13(11), 1377-1381.  67  6 Appendices 6.1 Appendix A: Debriefing Questions The following questions were presented visually to subjects at the end of each scanning session. Only one question was displayed at a time. When subjects indicated that they had finished answering a question, the next question was displayed. Subjects’ verbal responses were recorded and later transcribed. Questions marked with an asterisk were displayed for sessions with feedback only. Questions regarding up-regulation Where were you looking during up-regulation?* Were you able to simply observe your thoughts or did you find yourself concentrating on a particular thought? As you were monitoring your thoughts, what labels did you find yourself using most often? Rate on a scale of 1-7 how much mental energy up-regulation required. 1 = “I was very relaxed” 7 = “I was very concentrated” Rate on a scale of 1-7 how much you used the feedback in order to assist you in completing the up-regulation task. 1 = “I did not use it at all” 7 = “I used it very frequently”* Did you notice any link between what you were doing and the activation signal? If yes, describe what you noticed and how this helped you perform the task?* Questions regarding down-regulation Where were you looking during down-regulation?* Did you experience any intruding thoughts during down-regulation? Rate on a scale of 1-7 how much mental energy down-regulation required 1 = “I was very relaxed” 7 = “I was very concentrated”  68  Rate on a scale of 1-7 how much you used the feedback in order to assist you in completing the down-regulation task. 1 = “I did not use it at all” 7 = “I used it very frequently”*  69  6.2 Appendix B: Questionnaires and Personality Measures Table 6.1: Mean scores for RRQ  Subject  Rumination  Reflection  One  2.00  2.58  Two  3.58  3.33  Three  4.17  3.67  Four  3.25  3.92  Five  4.00  4.25  Six  3.83  3.58  Seven  2.08  3.92  Mean  3.27  3.61  Table 6.2: Mean scores for KIMS  Subject  Mindfulness Factors Act with Awareness  Observe  Describe  Accept without Judgment  One  30  32  31  38  Two  49  34  27  38  Three  44  21  36  20  Four  39  22  26  32  Five  44  33  31  21  Six  47  21  20  20  Seven  53  31  37  36  Mean  43.72  27.72  29.71  29.29  70  Table 6.3: Mean scores for BFI-44  Subject  Extraversion  Personality Traits Agreeableness Conscientiousness  Neuroticism  Openness  One  2.38  2.89  3.11  1.13  3.80  Two  4.88  2.89  3.89  3.13  4.00  Three  4.25  3.00  3.44  3.00  3.60  Four  2.63  4.11  4.56  3.75  3.80  Five  2.88  4.33  3.11  3.50  4.50  Six  2.75  3.67  2.67  3.75  4.70  Seven  4.00  4.22  4.56  1.25  4.30  Mean  3.39  3.59  3.62  2.79  4.10  71  6.3 Appendix C: Ethics Approval Certificate  72  73  

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.24.1-0070823/manifest

Comment

Related Items