左右聴覚野の音声言語処理における機能分化―多チャネル近赤外分光法（NIRS）による検討―

Cerebral Lateralization in Spoken Language Processing Measured by Multi-Channel Near-Infrared Spectroscopy（NIRS）

要旨　聴力の正常な健常者において，左右感覚性言語野の脳反応を多チャネル近赤外分光法により測定した。最後の母音（音韻）のみ異なる2単語を聴取させる音韻対比課題と，単語末の抑揚のみ異なる2単語を聴取させる抑揚対比課題において，ヘモグロビン反応のピークから左右差を側化指数として算出し比較すると，右利きでは85％の被験者で音韻対比の左優位が示された。右利き以外では50％の被験者で音韻対比における右優位が示され，残り50％では有意差がなかった。本法は無侵襲で被験者の負担も少ない脳機能測定法であり，感覚性の言語優位側の判定などに応用可能であると思われる。

Language processing is lateralized to the left hemisphere in 95％ of people as assessed by invasive techniques like Wada test（intracarotid amobarbital test） and cortical electric stimulation. Here we report a new non-invasive method of assessing language dominance by measuring speech-evoked local cerebral responses in hemoglobin（Hb） concentration with near infrared spectroscopy（NIRS）. The activation of the bilateral auditory areas to phoneme and pitch contrasts was investigated in twenty right-handed（RH） and ten non-right-handed（NRH） normal-hearing native speakers of Japanese with a 24-channel NIRS system using 830 and 780nm lasers. Three analysis-synthesized, inflected forms of a Japanese verb／iku／;（a）／itta／, （b）／itte／, and （c）／itta?／, were used as stimuli, delivered diotically at a rate of one word per second on average. In phoneme and pitch contrast conditions, a baseline block containing（a） and a test block including （a） and （b）（phoneme-contrast） or （a） and （c）（pitch contrast） were alternated and repeated several times. Laterality indices, LI＝（L－R）／（L＋R）, calculated from the peaks of the selectively averaged Hb responses in the left（L） and right（R） auditory areas, were compared between the two contrast conditions. A significant leftward shift of LI in the phoneme condition relative to the pitch condition was observed in 85 ％ of the RH subjects, with a significant group difference in the same direction. In contrast, 50％ of the NRH subjects showed an inverse shift of LI between the two conditions, but no significant overall shift of LI was observed in this group. These results suggest that multi-channel NIRS could help non-invasively determine the dominant temporal cortex in speech perception, independent of speech production.