注意・知覚統合/空間・時間認知 Attention and Perceptual Integration／Spatial and Temporal Cognition
O1-9-1-1 LORETA独立成分分析を用いた安静時回路とその加齢性変化の検出 Detection of EEG-Resting State Networks and its aging-related changes by LORETA-ICA method ○青木保典1, 石井良平1池田俊一郎1, 畑真弘1, 今城郁4, 松崎晴康5, 武者利光5, 朝田隆6, 岩瀬真生1, 武田雅俊1 ○Yasunori Aoki1, Ryouhei Ishii1, Roberto D. Pascual-Marqui2,3, Shunichiro Ikeda1, Masahiro Hata1, Kaoru Imajo4, Haruyasu Matuzaki5, Toshimitsu Musha5, Takashi Asada6, Masao Iwase1, Masatoshi Takeda1 大阪大学大学院　医学系研究科　精神医学1, 滋賀医大　精神医学3, 日本光電4, 脳機能研究所5, 筑波大学　神経精神医学6 Dept Phychi, Univ of Osaka, Osaka1, The KEY Inst for Brain-Mind Research, Univ Hosp of Psychi, Zurich, Switzerland2, Dept of Psychi, Shiga Univ of Medical Science3, Nihon Kohden Corp4, Brain Functions Lab Inc, Yokohama5, Dept of Neuropsychi, Inst of Clinical Medicine, Univ of Tsukuba, Tsukuba6 Recent functional magnetic resonance imaging (fMRI), electroencephalography (EEG) and magnetoencephalography (MEG) studies has revealed that the brain works together between distant regions and construct cortical networks to perform fundamental functions and during complex cognitive action, several cortical networks correlates positively or negatively. Furthermore, more recent studies revealed that this set of cortical networks can be obtained by applying independent component analysis (ICA) to resting-state fMRI data or by applying Source Reconstruction and ICA to resting-state MEG data and which is called resting state networks (RSNs). It was also revealed that cortical electromagnetic activity in RSN regions shifts its frequency from lower sensory regions to higher-order regions. But, most RSN studies used fMRI data that is derivative of electromagnetic cortical activity and without frequency information. In this study, we applied low resolution brain electromagnetic tomography -ICA (LORETA-ICA) with five frequency bands to resting state EEG data in 68 healthy subjects and identified four RSNs; visual network, right ventral attention network (right VAN) negatively correlated with left posterior dorsal attention network (DAN), anterior hub of default mode network (DMN) negatively correlated with right VAN, and left VAN negatively correlated with posterior hub of DMN across alpha and beta frequency bands for the first time. Next, to detect aging-related changes of four RSNs, we divided healthy subjects into three age groups and performed the Student's t test. Finally, we found that left VAN activation from younger to middle aged group, right VAN deactivation and anterior hub of DMN deactivation from middle aged to elderly group. Overall findings indicate that LORETA-ICA can detect RSNs with frequency information which represents cortical roles and also aging-related changes of RSN.	O1-9-1-2 自閉症児における腕交差時の時間順序判断 A crossed hand illusion task with autistic children ○和田真1, 鈴木繭子2, 東江浩美2, 高木晶子3, 宮尾益知4, 神作憲司1 ○Makoto Wada1, Mayuko Suzuki2, Hiromi Agarie2, Akiko Takaki3, Masutomo Miyao4, Kenji Kansaku1 国リハ研・脳機能・脳神経科学1, 国リハ研・発達情報セ2, 国リハ・自立支援局・秩父学園3, 成育医療研究セ・こころ・発達心理4 Sys Neurosci Sect, Dept of Rehab for Brain Func, Res Inst of NRCD, Tokorozawa1, Info Cent for Development Disabil, Res Inst of NRCD, Tokorozawa2, Chichibu Gakuen, Rehab Serv Bureau of NRCD, Tokorozawa3, Div of Development Neuropsychol, Dept of Psychol Med, Natl Cent for Child Health and Develop, Tokyo4 It has been suggested that autistic children may have difficulty with implicitly switching between an egocentric and an abstract allocentric stance and that this may lead to real-life problems in social interactions (Frith & De Vignemont, 2005); however, little empirical evidence is available in this regard. This study used a crossed hand illusion task with autistic children. Neurotypicial individuals experience a subjective reversal of temporal order judgments when the two hands are stimulated while they are crossed (Yamamoto & Kitazawa, 2001) and it may be caused by a conflict between an egocentric and an allocentric stance (Shore et al., 2002). Autistic children (n = 10, 11.8 ± 0.7 y.o.) and neurotypical children (n = 10, 11.9 ± 0.7 y.o.) were required to judge temporal order of two tactile stimuli that were delivered to their both ring fingers in the range of ± 1500ms stimulus onset asynchronies with their arms uncrossed and crossed. To evaluate judgment probabilities of each participant, degree of the reversal illusion was calculated as the sum of differences between correct response rates of the arms crossed condition and those of the arms uncrossed condition.We showed that the reversal illusion was significantly smaller in autistic than in neurotypical children (P < 0.05, t-test). In contrast, no difference in temporal resolution was observed between groups when participants' arms were not crossed. These results imply that an egocentric body image is given priority over the allocentric spatial location of the hand in autistic children, and the disturbance may interfere with the development of an external frame of reference in real-life situations.
O1-9-1-3 海馬場所細胞と内嗅野の空間周期表現を持つ細胞によるhomingの計算論モデル A Model of homing computation with CA1 place cells and "stripe cells" as upper-stream elements of grid cells ○山口陽子1,2, 三津沢将司1,2 ○Masashi S Mitsuzawa1,2, Yoko Yamaguchi1,2 東京大院・情報理工1, 理研BSI 神経情報基盤センター2 Grad. school of Information Science and Technology, Univ of Tokyo, Tokyo1, Neuroinfomatics Japan Center, RIKEN BSI, Wako2 In rodent studies, the hippocampal formation has been thought to be the main area of representing the cognitive map with spatially selective cells including place cells and grid cells. On the other hand, computational theories on the cognitive map based on the developing experimental findings are little clarified. Our question is whether simple computation of homing from a given start position to the goal position with place cells and spatially selective cells in the entorhinal cortex(EC) is possible or not. Firstly, we propose a neuronal model of grid cells consists of component cells whose firing fields are periodic in a form of stripes in the environmental space termed "stripe cells". We simply assume that the activity of the stripe cells is given by a pair of integrate-and-fire neurons with directional preference. Their superposition in triple direction gives the firing field of grid cells. To elucidate the homing computation, we focus on the computation in CA1 and EC by assuming the memory of place cells are stored in the hippocampus. We assume the synfire relationship between a population of CA1 place cells and EC stripe cells during encoding, which enables a synfire population of stripe cells to be indexed by a CA1 place cells.Accordingly, Internal activation of stripe cells corresponding to a number of grid cells is sent to CA1 to cause superposition of traveling stripe patterns from the start to the goal. Homing vector should be uniquely given by an intersection of different goal candidates of different stripe cells. We found the unique answer is found only when grid cells consist of stripe cells with an extra redundant direction: triple stripes are required for obtaining two dimensional homing vectors. Furthermore, we found that the triple direction must be asymmetry.We conclude that our model can compute homing with stripe cells as directional axes and that a redundant direction has a critical role in determining a unique homing vector.	O1-9-1-4 Information integration and phase-locking (synchrony) of frontal-parietal EEG signals ○Steven Phillips1, Yuji Takeda1, Singh Archana K.2 National Institute of Advanced Industrial Science and Technology (AIST)1, Advanced Telecommunications Research (ATR/NIA/DBI)2 Integrating multiple sources of information is essential for complex cognitive behaviour. A category theory approach to cognitive complexity accounts for differences in terms of the arity of the underlying (co)product (Phillips, et al, 2009). By parametrically varying product arity in a visual search task (as the number of feature dimension--one, two, or three--needed to identify the target of search), we identified greater phase-locking (PLV), a measure of synchrony, between frontal and parietal EEG electrodes with arity (Phillips, et al 2012). Here, we test the duality principle, a uniquely category-theoretic account of complexity, by parametrically varying the arity of the underlying coproduct (dual to product) in an object categorization task. Participants categorized objects as vehicles or nonvehicles in three arity conditions, depending on the number of image fragments that must be pieced together for identification: one, two, or three. Consistent with our previous study, we observed greater PLV with coproduct arity in the 30-34 Hz, 200-400 ms post-stimulus frequency-time band for F3-P3, Fz-Pz, F4-P4 electrodes. This band overlaps with the 30-38 Hz, 175-225 ms band observed in our previous study for products. Although further analysis is needed to assess the significance of this coproduct effect, these initial results suggest a common (category-theoretic) computational principle underlying complex cognition that is realized in the brain as phase-locking between frontal and parietal lobes. References Phillips, S., Takeda, Y., & Singh, A. (2012). Visual feature integration indicated by phase-locked frontal-parietal EEG signals. PLoS ONE, 7(3), e32502. doi:10.1371/journal.pone.0032502 Phillips, S., Wilson, W. H., & Halford, G. S. (2009). What do Transitive Inference and Class Inclusion have in common? Categorical (co)products and cognitive development. PLoS Computational Biology, 5(12), e1000599. doi:10.1371/journal.pcbi.1000599

上部に戻る

前に戻る