Neuroscience Letters

Transcription

1 Neuroscience Letters 499 (2011) Contents lists available at ScienceDirect Neuroscience Letters j our nal ho me p ag e: The neural organization of perception in chess experts Daniel C. Krawczyk a,b,, Amy L. Boggan a,m. Michelle McClelland a, James C. Bartlett a a Center for BrainHealth and School of Behavioral and Brain Sciences, The University of Texas at Dallas, Richardson, TX , USA b Department of Psychiatry, University of Texas Southwestern Medical Center, Dallas, TX , USA a r t i c l e i n f o Article history: Received 3 February 2011 Received in revised form 2 May 2011 Accepted 15 May 2011 Keywords: Perception Face processing Chess Expertise fmri a b s t r a c t The human visual system responds to expertise, and it has been suggested that regions that process faces also process other objects of expertise including chess boards by experts. We tested whether chess and face processing overlap in brain activity using fmri. Chess experts and novices exhibited face selective areas, but these regions showed no selectivity to chess configurations relative to other stimuli. We next compared neural responses to chess and to scrambled chess displays to isolate areas relevant to expertise. Areas within the posterior cingulate, orbitofrontal cortex, and right temporal cortex were active in this comparison in experts over novices. We also compared chess and face responses within the posterior cingulate and found this area responsive to chess only in experts. These findings indicate that the configurations in chess are not strongly processed by face-selective regions that are selective for faces in individuals who have expertise in both domains. Further, the area most consistently involved in chess did not show overlap with faces. Overall, these results suggest that expert visual processing may be similar at the level of recognition, but need not show the same neural correlates Elsevier Ireland Ltd. All rights reserved. Expertise can be developed through extreme levels of practice resulting in behavior considered to be outstanding relative to the general population. Uncommonly effective performance within a domain remains the clearest marker of expertise [10,7]. Recent neuroimaging explorations of expertise using have begun to provide insights into the neural basis of expertise [8,26]. Among expert domains, chess is widely regarded to be one in which a select few experts perform at an exceptional level [6,15]. In the process of becoming outstanding at chess, a Master level player accumulates massive visual experience with chess configurations. This experience confers distinct advantages to experts over novices when encountering situations that commonly appear in games. These expertise effects are limited to game configurations, as the perceptual and memory advantages of experts are greatly reduced when tasks are not chess game specific [10,6]. Meanwhile, the brain organization of perceptual recognition in chess experts has remained unclear. The perception of faces is a skill at which nearly everyone is considered to be an expert. Face perception has been associated Abbreviations: FFA, fusiform face area; OFA, occipital face area; MRI, magnetic resonance imaging; MNI, montreal neurological institute; ROI, region of interest; HRF, hemodynamic response function; GLM, general linear model. Corresponding author at: Center for BrainHealth and School of Behavioral and Brain Sciences, The University of Texas at Dallas, Richardson, TX , USA. Tel.: ; fax: address: daniel.krawczyk@utdallas.edu (D.C. Krawczyk). activation of the fusiform gyrus [18]. More broadly, the fusiform is considered to be a neural marker of visual expertise, as other studies have reported selective fusiform activity when car experts and bird experts perceive cars and birds and when radiologists examine scans [12,27,17]. Such findings have spawned the hypothesis that the fusiform gyrus can support expert processing in a variety of domains. However, cars, birds, and body scans share properties with faces, including similar features, similar configurations, and biological characteristics in the case of birds and radiology scans. Chess allows a critical test for theories of visual expertise, as chess configurations bear little featural or configural resemblance to faces, cars, or birds and also lack biological characteristics. If chess experts process chess patterns similarly to faces, it would challenge the view that common visual or biological characteristics are necessary for different classes of stimuli to be perceived in the same way [9]. The idea that face-selective fusiform cortex can become adapted to process chess patterns is a compelling one, and there have been reports in the expertise literature that the fusiform may be involved in processing chess patterns [21,24,2]. A recent documentary film showed a neuroimaging clip with a chess expert and suggested that the face-selective fusiform can be hijacked to process chess patterns [23]. However, there has not yet been a published study comparing expert chess perception to that of faces and other visual categories. The present study addressed the central question of how large amounts of practice at chess alters the functional organization of the brain. Specifically, we tested whether face selective areas /$ see front matter 2011 Elsevier Ireland Ltd. All rights reserved. doi: /j.neulet

2 D.C. Krawczyk et al. / Neuroscience Letters 499 (2011) Fig. 1. (A) Examples of each category shown in the experimental task. Conditions included blocks of chess, random chess, faces, outdoor scenes, and objects. (B) Regions of significant difference within the experts over the novices on the chess > random chess contrast. (C) Regions significantly greater for novices over experts on the chess minus random chess contrast. become adapted to support chess expertise at an early perceptual level and whether there are other regions that become more active when chess expertise has been achieved. We compared the activation of chess experts and novices when viewing faces, chess boards, and other stimuli to determine whether chess and face perception activate common regions using fmri. We included a comparison of chess board recognition to scrambled chess board recognition, as scrambled boards tend to reduce the performance advantage typical of chess experts [10,6]. By including this comparison, we are able to address a secondary question: whether chess expertise is limited to game-specific configurations at the neural level, or whether this expertise extends to non-game configurations using the same spatial and featural information. Furthermore, we were interested in whether such areas would show chess selectivity relative to faces and other visual categories including scenes and objects. Subjects were twelve healthy, right-handed males. Six were chess experts recruited from the UT Dallas Chess Program, age (M = 23 years). These subjects ranked within the top one percent of tournament players (five International Masters, one Grandmaster). Their expertise was substantiated by their competitive ratings (Elo range = ; M = 2515), years playing (M = 16 years), and tournament activity (M = 17 per year). The remaining six subjects were healthy males who were chess novices age (M = 25 years). These subjects reported that they rarely played chess and had not participated in tournaments. This experiment was approved by the Institutional Review Boards of UT Dallas and UT Southwestern Medical Center. Informed consent was obtained in accordance with the 1964 Declaration of Helsinki. Subjects viewed blocks of items and judged whether each was a repeat or a new image. Stimuli consisted of images of chess boards from games, randomly positioned chess boards that could not occur in real games, objects [14], and outdoor scenes (see Fig. 1A). Images were presented in five runs of 8 blocks, 12 images per block, 2 s per image, and a 5000 ms inter-stimulus interval. We used longer exposure times and inter-stimulus intervals than standardly appear in the face literature to ensure that novices could perform the task given the complexity of chess boards. Images were presented offset from center to the right or left in an alternating sequence to avoid apparent motion effects in the chess conditions between non-matching items in sequence. Two image repeats occurred per block, and subjects were instructed to press buttons for each repeat. Each block, presented in a pseudo-randomized order, contained one image category or was a fixation block (lasting 30 s). Images were acquired using a 3T Philips MRI scanner with a gradient echoplanar sequence (TR = 2000 ms, TE = 28 ms, flip angle = 20 ) sensitive to BOLD contrast. Each volume consisted of tilted axial slices (3 mm thick, 0.5 mm slice gap) that provided nearly whole brain coverage. Anatomical T1-weighted images were acquired in the following space: TR = 2100 ms, TE = 10, slice thickness = 4 mm with no gap at a 90 flip angle. FMRI block design analyses were conducted using multiple regression. Preprocessing was conducted using SPM5 ( EPI images were realigned to the first volume and then smoothed (8 mm 3D Gaussian kernel). Separate regressors were used to model each block, convolved with a canonical hemodynamic response function (HRF), and entered into a modified general linear model (GLM). Parameter estimates were extracted from this analysis for each regressor. At an individual subject level, contrasts between conditions were computed by performing one-sample t-tests on the contrasted images. A faces minus scenes and objects contrast was used to functionally define ventral temporal and occipital regions of interest (ROIs) using a Family-Wise Error (FWE) corrected threshold (p <.01). In some instances False Discovery Rate (FDR) (p <.05) or uncorrected (p <.005) thresholds were used to localize as many of the face regions as possible in each subject (minimum of 10 voxels per cluster). While we did not run an independent face localizer to isolate fusiform face area (FFA) regions, we did not include chess or random chess to localize FFAs, thereby leaving chess as an independent category to be evaluated. We also isolated chess regions using a chess minus random chess contrast between groups. To carry out a subject-specific ROI analyses, we ran this contrast on each group independently (p <.001 uncorrected, 10 voxel cluster minimum). This contrast showed no significant clusters in novices. In experts, this contrast resulted in two clusters within the posterior cingulate (MNI coordinates: x = 32, y = 10, z = 12) and the right insula (x = 12, y = 50, z = 10). To further isolate chess responses we defined ROIs at the individual level. Five of the experts showed significant activation within the

3 66 D.C. Krawczyk et al. / Neuroscience Letters 499 (2011) posterior cingulate (between x = 12 to 13; y = 66 to 24, and z = 2 to 39). These individual subject ROIs were corrected between FWE p <.01 and FDR p <.05. We did not perform an ROI analysis on the right insula as there were no spatially consistent clusters within this region for a majority of the experts. ROI data were extracted and converted to percent-signal change for each subject using MarsBar (sourceforge.net/projects/marsbar). These data were then analyzed using 2 group 5 condition ANOVAs and pairwise contrasts to evaluate face perception versus each of the other four conditions for the FFAs [10] and occipital face areas (OFAs) [27] and to evaluate chess processing versus face, object, and scene processing for the posterior cingulate (Bonferroni corrected t-tests within each ROI). The experts indicated that they could perceive all or most of the chess boards within 2 s. Several reported that random chess was more difficult to perceive than real chess. Novices reported that they could rarely perceive all of the pieces and none reported that they were aware that the random games were impossible according to the rules of chess, a distinction all experts readily reported. Accuracy was similar for both groups for all categories except for chess. Experts and novices exhibited high accuracy on faces (experts M = 97.92%, SD = 5.10; novices M = 95.83%, SD = 7.57), scrambled chess (experts M = 91.67%, SD = 3.23; novices M = 91.67%, SD = 7.57), scenes (experts M = 76.04%, SD = 20.70; novices M = 94.79%, SD = 7.31) and objects (experts M = 95.83%, SD = 7.57; novices M = %, SD = 0). The only significant group difference was that experts (M = 97.92%, SD = 3.23) were more accurate at detecting repeats of chess boards than novices (M = 87.50%, SD = 13.69). This was confirmed by an independent samples t-test, t(10) = 1.81, p <.05. No other performance differences were significant between the groups, though the experts were numerically lower on scene recognition performance than the novices. FMRI activation comparisons of each category were carried out using 2 5 ANOVAs for each ROI. All four ANOVAs reached significance for both group and category effects, but few showed group by category interactions. Within the Left FFA the effect of group was significant with novices showing higher mean percent signal change (M = 1.03) than experts (M = 0.69), F(1, 9) = 33.97, p <.001 (see Fig. 2A). Additionally, the effect of category was also significant, F(4, 36) = 34.03, p <.001. The right FFA showed similar effects with novices showing higher percent signal change (M = 0.94) than experts (M = 0.80), F(1, 10) = 56.17, p <.001. There was a significant category effect, F(4, 40) = 31.45, p <.001. In the left OFA novices showed higher percent signal change (M = 1.48) than experts (M = 1.16), F(1, 8) = 47.36, p <.001 with a significant category effect, F(4, 32) = 17.20, p <.001. The left OFA showed a significant group by category interaction F(4, 32) = 6.81, p <.001. This interaction indicated that experts showed greater face selectivity in the left OFA, while novices showed more left OFA activation to both of the chess categories. Lastly, the right OFA showed a significant group effect with experts showing a higher mean percent signal change (M = 1.20) than novices (M = 1.11), F(1, 7) = 53.95, p <.001. There was also a significant effect of category F(4, 28) = 26.72, p <.001. To further understand the within category effects we followed up with post hoc comparisons. In the experts, post hoc comparisons (p <.05) indicated that face perception activated the face-selective ROIs more than each of the other categories, and all ROIs responded more strongly to faces than chess game configurations, driving the effects. In the left FFA activation to faces (M = 1.27) was significantly greater than to chess (M = 0.48), scenes (M = 0.20) and objects (M = 0.82). The face and random chess (M = 0.70) comparison did not reach significance (p =.067). In the right FFA, face activation (M = 1.47) was significantly greater than activation for chess (M = 0.73), scenes (M = 0.21) and objects (M = 0.79). The comparison of face and random chess (M = 0.80) conditions did not reach significance when corrected for multiple comparisons (p =.019). In the left OFA, face activation (M = 2.02) was significantly greater than for chess (M = 0.86), random chess (M = 1.15), scenes (M = 0.54), and the objects (M = 1.24). The right OFA showed greater activation to faces (M = 1.93) over chess (M = 1.22) and objects (M = 1.18) (see Fig. 2). Post hoc comparisons for novices (p <.05) revealed that the left FFA activation to faces (M = 1.47) was significantly greater than to chess (M = 0.98), scenes (M = 0.59) and objects (M = 0.93). The face and random chess (M = 1.18) comparison did not reach significance (p =.10). The right FFA was significantly more active in response to faces (M = 1.48) than chess (M = 0.85), random chess (M = 1.03), scenes (M = 0.54) and objects (M = 0.82). In the left OFA, face activation (M = 1.48) was significantly greater than for scenes (M = 0.60) and objects (M = 0.81). The right OFA showed a similar pattern, with greater activation to faces (M = 1.32) over scenes (M = 0.38) and objects (M = 0.86). Thus faces were predominantly active over other categories leading to the significant category effect. We directly compared the groups on the chess minus random chess contrast using a random-effects fmri group analysis (uncorrected p <.001, 20 voxel minimum) (refer to Fig. 1B and C). Significant differences emerged with experts showing greater activation in the left orbitofrontal cortex (x = 4, y = 58, z = 2), in the left (x = 10, y = 50, z = 12) and right posterior cingulate (x = 14, y = 52, z = 2), and in the left anterior temporal cortex (x = 50, y = 8, z = 26). By contrast, the novices showed greater activation than experts in two distinct clusters in right parietal cortex (x = 32, y = 52, z = 40) and (x = 38, y = 54, z = 46). To further analyze the chess-related areas we performed contrasts on each group independently, subtracting activation of random chess from chess. This analysis revealed two clusters within the posterior cingulate and the right insula within the experts (uncorrected p <.001), but no significant clusters in the novices. We then conducted individual ROI analyses on the posterior cingulate areas of the experts. Percent signal change comparisons in this area resulted in a significant overall repeated measures ANOVA, F(4, 20) = 4.29, p <.05. Post hoc comparisons (evaluated at p <.05) revealed greater activation for chess (M =.28) compared to faces (M =.48), random chess (M =.78), objects (M =.56), and scenes (M =.46) (refer to Fig. 2B). Though expert chess players possess high levels of visual familiarity with chess configurations, the brain areas most sensitive to face perception showed no evidence of greater relative activation in response to such configurations. Experts showed an advantage over novices in detecting repeats of valid chess configurations. This is consistent with the advantage previously observed between experts and novices in memory for chess configurations [10,6,15]. While the detection task was simple, this difference was likely significant due to the exceptionally high performance of the experts. Notably, both groups performed well at recognition overall, but this difference alone differentiated the two. Experts also reported that they were able to perceive most or all of the valid chess boards within the brief exposure period, while novices did not report this ability. There was a numerical trend toward lower accuracy in matching scene stimuli within the experts compared to novices. While this was non-significant, it suggests that there may be some differentiation in recognition of non-face and non-chess stimuli within experts and novices, which may be addressed in future research. Chess board configurations appear to be different from facial configurations in terms of both perception and neural organization for experts. While there have been prior reports that chess board processing and face processing involve overlapping brain areas, notably the right FFA, our findings do not support this view. There may be instances in which face-sensitive areas respond to chess stimuli [21,24], but our results indicate that this does not reflect

4 D.C. Krawczyk et al. / Neuroscience Letters 499 (2011) Fig. 2. (A) ROI results from face selective areas, the FFA and OFA. Graphs present brain activation averages from each visual category plotted by region. All four face-selective areas showed significantly greater activation for faces than for chess boards. (B) ROI results from the posterior cingulate defined by subtracting random chess from real chess displays. The activation was significantly greater for chess boards over each of the other categories. basic perceptual processing of the type that was examined in the present study and in most prior studies that have established the existence of the fusiform face area [13,18]. While processing configural patterns is known to be important in both face perception and expert chess board perception [2,23], such processing need not rely on the same neural mechanisms. Further, we did not observe clear reductions in FFA activation either in chess experts or in chess novices. Such a finding would be consistent with neural plasticity dedifferentiating standard face expertise responses [8,22]. Similarly, within the OFA areas our results showed face selectivity, absent of modulation based on any other category strongly within experts, and approximately equivalent activation toward faces, chess, random chess, and objects over scenes in the novice group, possibly due to differences in attention demands among the

5 68 D.C. Krawczyk et al. / Neuroscience Letters 499 (2011) categories in the task. Our results are most consistent with the position that chess and face expertise are processed independently as measured by modulation of fusiform responses in experts. We also found evidence among the chess experts of areas related to processing chess boards over random boards within the posterior cingulate, an area of the left orbitofrontal cortex, the left anterior temporal cortex, and the right insula at the group level. The orbital and anterior temporal regions have been associated with emotional and motivational processing, as well as linking emotion to reasoning [16]. The region of interest analysis of the posterior cingulate suggested that this area was reliably associated with only chess board processing relative to the other categories including faces. The posterior cingulate has previously been associated with an fmri comparison of partial chess game boards to geometric shape displays in chess experts [3]. It has also been associated with the default mode network of the brain [16,25] which relates to selfdirected thinking as well as semantic memory retrieval. Given the simplicity of our task instruction and the extensive expertise of our subjects, it is not possible to determine the precise role the posterior cingulate plays in chess-related cognition, but activation of this area is consistent with memory retrieval of game configurations, or self-directed thinking about chess during the time these displays were evaluated. Notably, the posterior cingulate ROIs were not significantly active in novices and in experts were not active during perception and evaluation of any of the other stimulus categories we used in this experiment including faces. Future work will be needed to clarify the cognitive functions associated with this region in chess experts. It will also be valuable to determine whether this same area is active in other more demanding chess tasks. There was some differential activation of the left parietal cortex in novices over experts associated with real game chess processing. The parietal lobes have been associated with spatial processing previously [3], suggesting that this activation may have been related to a greater level of visual location search than occurred for the expert group. The left temporo-parietal junction showed evidence of chess modulation in experts. This is an area that has been associated with integration of visual features, thus providing the clearest evidence of a perceptually driven neural change related to chess expertise. To put our results in the context of the broader question of how the neural basis of expertise operates, it is worthwhile to consider studies from the literature on word perception and literacy. Words have been shown to produce highly specific neural effects that depend on specific configurations, such as the presence of vowels versus consonants [5]. Likewise, our comparison of scrambled versus real chess board configurations elicited expert effects within the posterior cingulate that were specific to real chess games only, consistent with the classic behavioral effects showing effects limited to chess games over scrambled boards. Further, priming effects vary depending on specific material types [20]. Similarly, cortical regions sensitive to face processing showed face selectivity without overlap with chess. On the other hand, other recent work has indicated a different pattern brought about by reading experience. Dehaene et al. [8] recently reported that reading experience shows evidence of domain general effects in V1, temporal, and fusiform cortex, as literacy increased overall neural responses in these areas. These contrasting results indicate that caution must be taken when considering the neural basis of expertise, as domain specificity within neural responses appears to vary based on the specific skills acquired. The results we report are limited to basic chess recognition and are preliminary. Future work varying levels expertise and a greater range of chess tasks may further clarify neural changes for chess expertise. These results leave open a variety of intriguing routes for future investigations. Chess configurations, unlike face configurations, are movable and almost surely differ from faces in the manner in which attention is allocated toward areas of the stimulus, with experts tending to plan a next move [11], assess which side is winning [1], and recognize patterns that they have experienced before [4]. Several of these cognitive factors may lead to perceptual differences between faces and chess at both behavioral and neural levels of analysis. It may be particularly interesting to test individuals with varying degrees of chess experience in future work, as the neural representations of chess may undergo alterations accompanying the acquisition of greater fluency with chess configurations [19]. The neural basis of chess perception may also change with variations in the capacity and need for visualization of the patterns occurring within chess games. Acknowledgments We thank James Stallings and the Chess Program at UT Dallas for their continued support of our research. We also thank Michael Motes for neuroimaging assistance. This work was supported by a UT Dallas Catalyst Grant. References [1] O. Amidzic, H.J. Riehle, T. Elbert, Focal magnetic gamma bursts as a signature of memory chunks and the aptitude of chess players, J. Psychophysiol. 20 (2006) [2] C.M. Bukach, I. Gauthier, M.J. Tarr, Beyond faces and modularity: the power of an expertise framework, Trends Cogn. Sci. 10 (2006) [3] G. Campitelli, F. Gobet, K. Head, M. Buckley, A. Parker, Brain localization of memory chunks in chess players, Int. J. Neurosci. 117 (2007) [4] G. Campitelli, A. Parker, K. Head, F. Gobet, Left lateralization of autobiographical memory: an fmri study using the expert archival paradigm, Intern. J. Neurosci. 118 (2008) [5] M. Carreiras, J.A. Duñabeitia, N. Molinaro, Consonants and vowels contribute differently to visual word recognition: ERPs of relative position priming, Cereb. Cortex 19 (2009) [6] W.G. Chase, H.A. Simon, Perception in chess, Cogn. Psychol. 4 (1973) [7] A.D. de Groot, Thought and Choice in Chess, 2nd ed., Mouton De Gruyter, New York, [8] S. Dehaene, F. Pegado, L.W. Braga, P. Ventura, G.N. Filho, A. Jobert, G. Dehaene- Lambertz, R. Kolinsky, J. Morais, L. Cohen, How learning to read changes the cortical networks for vision and language, Science 330 (2010) [9] R. Diamond, S. Carey, Why faces are and are not special: an effect of expertise, J. Exp. Psychol.: Gen. 115 (1986) [10] K.A. Ericsson, J. Smith, Prospects and limits of the empirical study of expertise: an introduction, in: D.J. Levitin (Ed.), Foundations of Cognitive Psychology, Allyn & Bacon, Boston, 2011, pp [11] D. Ferrari, A. Didierjean, E. Marmeche, Dynamic perception in chess, Q. J. Exp. Psychol. 59 (2006) [12] I. Gauthier, P. Skudlarski, J. Gore, A. Anderson, Expertise for cars and birds recruits brain areas involved in face recognition, Nat. Neurosci. 3 (2000) [13] I. Gauthier, M.J. Tarr, J. Moylan, P. Skudlarski, J.C. Gore, W.A. Anderson, The fusiform face area is part of a network that processes faces at the individual level, J. Cogn. Neurosci. 12 (2000) [14] J.M. Geusebroek, G.J. Burghouts, A.W.M. Smeulders, The Amsterdam library of object images, Int. J. Comput. Vision 61 (2005) [15] F. Gobet, H.A. Simon, Templates in chess memory: a mechanism for recalling several boards, Cogn. Psychol. 31 (1996) [16] D.A. Gusnard, E. Akbudak, G.L. Shulman, M.E. Raichle, Medial prefrontal cortex and self-referential mental activity: relation to a default mode of brain function, Proc. Natl. Acad. Sci. U.S.A. 98 (2001) [17] E.M. Harley, W.B. Pope, J.P. Villablanca, J. Mumford, R. Suh, J.C. Mazziotta, D. Enzmann, S.A. Engel, Engagement of fusiform cortex and disengagement of lateral occipital cortex in the acquisition of radiological expertise, Cereb. Cortex 19 (2009) [18] N. Kanwisher, J. McDermott, J. Chun, The fusiform face area: a module in human extrastriate cortex specialized for face perception, J. Neurosci. 17 (1997) [19] A. Keisel, W. Kunde, Playing chess unconsciously, J. Exp. Psychol.: Learn. Mem. Cogn. 35 (2009) [20] M. Perea, J.A. Duñabeitia, A. Pollatsek, M. Carreiras, Does the brain regularize digits and letters to the same extent? Q. J. Exp. Psychol. 62 (2009) [21] G. Righi, M.J. Tarr, Are chess experts any different from face, bird, or Greeble experts? J. Vis. 4 (2004) 504a. [22] B. Rossion, C.C. Kung, M.J. Tarr, Visual expertise with nonface objects leads to competition with the early perceptual processing of faces in the human occipitotemporal cortex, Proc. Natl. Acad. Sci. U.S.A. 101 (2004) [23] K. Scott, (Producer/Director) My Brilliant Brain: Make Me a Genius [Television series episode]. London, 2007, Windfall Films.

6 D.C. Krawczyk et al. / Neuroscience Letters 499 (2011) [24] D.L. Sheinberg, M.J. Tarr, in: I. Gauthier, D. Bub (Eds.), Perceptual Expertise: Bridging Brain and Behavior, Oxford Univ. Press, New York, 2010, pp [25] M. van Buren, T.E. Gladwin, B.B. Zandbelt, R.S. Kahn, M. Vink, Reduced functional coupling in the default-mode network during self-referential processing, Hum. Brain Mapp. 31 (2010) [26] X. Wan, H. Nakatani, K. Ueno, T. Asamizuya, K. Cheng, K. Tanaka, The neural basis of intuitive best next-move generation in board game experts, Science 331 (2011) [27] Y. Xu, Revisiting the role of the fusiform face area in visual expertise, Cereb. Cortex 15 (2005)