本研究の目的は，バーチャルリアリティ（VR）やウェアラブルデバイスによって自然には存在しない感覚反転環境を創出し，人間の多感覚統合脳機能に迫ることである．3か年計画の3年目となる2019年度においては，計画の遂行で必要となったVRシステムのワイヤレス化や実験時のストレス評価等を行った上で，取り纏めに向けた各種検討を行った．まず，2018年度までに得られた聴覚反転環境と前庭感覚反転環境への順応に関する知見に基づき，継続的な接触が可能な聴覚反転環境に絞って環境適応性と定着性を検討した．接触時間の進行とともに，違和感と相関する視聴覚連合野の脳リズムや手指の応答性と相関する聴覚野の誘発活動が変容したが，これらの現象は，反転環境から離れると通常の感覚環境での状態に戻り，反転環境に再接触すると反転環境での状態に再度なった．よって，新たな統合モデルを獲得すると，接触する感覚環境に応じて採用される統合モデルが切り替わることが分かった．また，視聴覚連合野の脳リズムが誤差検出に関わっていることを経頭蓋電気刺激によって検証した．一方で触覚反転環境に関しては，VRグローブを用いて手指に関する反転を検討したが，反転元の残存刺激の影響が除去しきれず今後の課題となった．しかしその他の感覚反転環境での知見から，多感覚統合の基礎となる感覚情報間の誤差検出機構が各感覚に共通して存在する可能性が示唆された．したがって，外部環境を脳内で適切に再構成できるように，多感覚統合処理は接触環境に対して最適化されることが明らかになった．従来研究の多くは，脳活動量の非線形的な増分に注目して多感覚統合脳機能を議論してきたが，本研究によって脳活動の量的な検討では得られない動的な機構の重要性が示された．
The aim of this study is to produce various reversed sensory environments that are impossible in a natural situation by using virtual reality (VR) technology and wearable devices, and to pursue multisensory integration processing in the human brain with them. In the third year (2019) of the 3-year plan, the following examinations were conducted toward the summarization after the VR system was made wireless and stress evaluation was set for the experiment that were required for promoting the plan. First, environmental adaptability and stability was examined by focusing on reversed auditory environment, to which continuous exposure is possible, based on the knowledge of adaptation to reversed auditory environment and reversed vestibular environment obtained by 2018. As the exposure time progressed, brain oscillations in the audiovisual association area correlated with a feeling of strangeness and evoked activity in the auditory area correlated with responsiveness of fingers were varied; these phenomena returned to the condition in the ordinary sensory environment after removal from the reversed environment, and turned to the condition in the reversed environment again after re-exposure to the reversed environment. Therefore, it was revealed that the employed integration model is switched according to the exposed sensory environment once the new integration model is acquired. Moreover, it was verified that the brain oscillations in the audiovisual area are associated with error detection by using transcranial electrical stimulation. As for reversed somatosensory environment, though the reversal was examined for the fingers using the VR glove, it was difficult to remove the effects of residual stimulation on the original side, which will be a future issue. From the knowledge in the other reversed sensory environments, however, the possibility is indicated that error detection mechanism across sensory information commonly exists for each sense underlying multisensory integration. Accordingly, it was revealed that multisensory integration processing is optimized for the exposed environment so that external environment can be reconstructed properly in the brain. While most of the previous studies have discussed multistory integration processing focusing on non-linear increase in the amount of brain activity, this study showed impotence of the dynamic mechanism that cannot be obtained by quantitative discussion of brain activity.