言語機能 Language
P3-2-151 事象関連電位N400成分にみられる個人差の起源に関する検討 The origin of interindividual variability in the N400 component ○秦政寛1, 保前文高1, 萩原裕子1 ○Masahiro Hata1, Fumitaka Homae1, Hiroko Hagiwara1 首都大学東京人文科学研究科言語科学1 Department of Language Sciences, Graduate School of Humanities, Tokyo Metropolitan University, Tokyo, Japan1 N400 is an event-related potential (ERP) component reflecting the processing of semantic information. Since the semantic relationships of words are subjective, the judgment of these relationships and the magnitude of N400 would vary among individuals. However, the cause of this interindividual variability at the neural level is still unclear. We aimed to investigate the origin of this variability by focusing on the time course of word processing. We used noun word pairs (prime and target) with different degrees of semantic relatedness as visual stimuli. The electroencephalogram was recorded from 123 Ag/AgCl electrodes placed on participants (n = 30; age, 18-24 years) performing a semantic-relatedness judgment task. The N400 effect (unrelated condition - highly related condition) was observed between 350 and 450 ms after target onset (average number of events, maximum 192 in each participant). We divided the participants into 3 groups based on the nature of N400 in the first 24 events: the amplitude was more than -2.5 µV in Group A, ranged between -2.5 and 0 µV in Group B, and was positive in Group C. Group A exhibited the N400 effect both in the first 24 events and in the average of all 192 events, whereas Group C did not show these effects. This type of intergroup difference was not observed in an early negative component (N100), but in the following positive shift (P150). A two-way ANOVA of the P150 component [Group (A, B, and C) × the number of averages (24 and 192)] showed the main effects of both factors (p<0.05) and the interaction (p<0.005), where the amplitude of Group A was smaller than that of Groups B and C in the first 24 events. These results indicated that the amplitude of P150 would affect the intensity of the N400 component; the lesser the P150 component, the greater the N400 effect. Our findings suggest that the P150 component reflects early stages of lexical-semantic processing.	P3-2-152 前部側頭葉底面は意味判断に深く関わる：高頻度皮質電気刺激／local field potentialによる研究 Ventral anterior temporal lobe actively engages in semantic judgment: direct evidence from electrical cortical stimulation and local field potential recording ○下竹昭寛1, 松本理器1, 上野泰治2國枝武治3, 福山秀直4, 宮本享3, 高橋良輔1池田昭夫1 ○Akihiro Shimotake1, Riki Matsumoto1, Taiji Ueno2, Paul Hoffman2, Takeharu Kunieda3, Hidenao Fukuyama4, Susumu Miyamoto3, Ryosuke Takahashi1, Matthew A Lambon Ralph2, Akio Ikeda1 京都大院・医・臨床神経1, 京都大院・医・脳神経外科3, 京都大院・医・脳機能統合研究センター4 Dept. of Neurology, Grad. Sch. of Med., Kyoto Univ.1, NARU, Sch.of Psychol Sci, Univ. of Manchester, UK2, Dept. of Neurosurg, Grad. Sch. of Med., Kyoto Univ.3, HBRC, Grad. Sch. of Med., Kyoto Univ.4 Introduction: Recently, neuropsychological and neuroimaging studies have increasingly implicated the ventral part of the anterior temporal lobe (vATL) as potentially crucial in pan-modal,semantic processing. We had a rare opportunity to prove the function of the vATL directly in patients with partial epilepsy and brain tumor who underwent presurgical evaluations with subdural electrodes. We attempted to delineate the semantic function in the vATL by recording cortical local field potentials (LFPs) while the patients complete a semantic judgment task and by stimulating the functionally-identified regions to evaluate the necessity of these area in semantic processing.Methods: Two tumor or epilepsy patients who underwent chronic electrode implantation in the language-dominant temporal lobe were studied (IRB#C533). LFPs were recorded time-locked to a synonym judgment task and a control difficulty-matched number judgment task. Direct cortical stimulation was delivered to a part of vATL where the most robust LFP was recorded. Weaker stimuli (7 mA, 3 sec) were delivered time-locked to the task presentation to analyze both the accuracy and reaction time.Results: For the synonym judgment task, robust LFPs were recorded at and around the anterior fusiform with a significant positive peak at 420-550 ms, while the non-semantic number judgment task did not produce any clear LFPs in this area. Compared with the control task, cortical stimulation at the anterior fusiform reduced accuracy of the synonym judgment task from 100% at baseline to 48 or 65%. In addition, RT was significantly prolonged in one patient. Conclusions: The present findings provided direct evidence that the anterior fusiform gyrus plays an active role in semantic judgment. Taken together with patient studies with semantic dementia, distortion corrected fMRI studies and our conventional cortical stimulation study, the present findings indicate that vATL is crucial for semantic processing.
P3-2-153 TMSとMEGを用いた言語優位半球判定法 A novel method to determine the language dominance combining TMS and MEG ○進士美沙子1, 柳澤琢史1,2, 後藤哲1,2, 平田雅之1,2, 菅田陽怜2, 荒木俊彦1, 細川祥代1, 岡村友美子1, 長谷川侑香1, 井原綾3, 梅原広明3, 依藤史郎1 ○Misako Shinshi1, Takufumi Yanagisaawa1,2, Tetsu Goto1,2, Masayuki Hirata1,2, Hisato Sugata2, Toshihiko Araki1, Sachiyo Hosokawa1, Yumiko Okamura1, Yuka Hasegawa1, Aya Ihara3, Hiroaki Umehara3, Shiro Yorifuji11 大阪大院・医・保・機能診断科学1, 大阪大院・医・脳神経外科学2, (独)NICT・未来ICT・脳情報通信研究室3 Dept.Funct.Diagnost.Sci, Sch Med, Osaka Univ, Osaka1, Dept.Neurosurgery, Osaka Univ, Osaka2, Brain ICT Lab, Adv ICT Res Inst, NICT, Japan3 Objective: Determining the language dominant hemisphere is imperative in the planning of neurosurgical procedures. The standard method is Wada test: intra-arteiral barbiturate injection. However, considering its invasiveness and risk of complications, an alternative noninvasive method must be necessary. In this study, we proposed a novel method to determine the language dominance by combining TMS and MEG. Notably, the TMS has been shown to have the ability to site-specifically suppress the language function and the MEG is known to sensitively detect the cerebral local oscillatory changes related to the language function. Here, we examined how the combination of these two modalities improves the accuracy of determination of language dominance. Method: Twelve healthy right-handed volunteers performed two types of language tasks: 1) Picture naming with TMS, 2) Verb generation with MEG. In the first task. TMS was applied on the right or left inferior frontal gyrus (IFG) at five different time points after the presentation of pictures. The times to answer the names of the pictures (reaction time, RT) were examined. In the second task, MEG was measured when subjects silently hit upon verbs corresponding to serially presented noun words. Language dominance was determined by using event-related desynchronization of low-gamma band activities on frontal lobes. Results: In 10 out of 12 subjects, TMS on left IFG significantly lengthened RT at 300ms and 375ms after the picture presentation, though the TMS on the right IFG showed no significant changes of RT. Using the MEG signals, language dominance was estimated as the left side in 9 out of 12 persons. For the remaining 3 subjects, language dominance was successfully estimated on left hemisphere according to the results of the TMS task, though the MEG could not determine the dominance. Interpretation: It was suggested that combining the TMS and the MEG could improve the accuracy to diagnose the language dominance noninvasively.	P3-2-154 意味記憶構造の神経科学的基盤の検討 The investigation for the semantic memory structure in the brain ○松本敦1, 井原綾1, 藤巻則夫1, 糸井誠司1, 梅原広明1 ○Atsushi Matsumoto1, Aya Ihara1, Norio Fujimaki1, Seishi Itoi1, Hiroaki Umehara1 独立行政法人情報通信研究機構1 Brain ICT Laboratory, Advanced ICT Research Institute, National Institute of Information and Communications Technology, Japan1 In this study, we investigated the neural basis of semantic network. We used fMRI and semantic priming paradigm to identify the brain regions showing the change of activity accompanied by the change of strength of semantic relationship between word pairs. Participants conducted the semantic decision task to the target word. The target word was preceded by a prime word. Semantic relatedness between the prime and target word was systematically manipulated. Semantic relatedness was treated as a continuum (rather than a discrete variable) to better reflect the nature of the semantic relation between words that we encounter in the world. We examined the regions that showed modulation in activity as a function of the degree of semantic relatedness. In this model, we entered the degree of semantic relatedness between prime and target (as determined by our pretests) as a parametric modulator with linear expansion for each item in the target conditions. Semantic priming was associated with enhanced activation in multiple bilateral fronto-temporal areas, i.e. semantic enhancement. The results suggest that semantic representation of word stored in several distinct brain regions.

上部に戻る

前に戻る