Uncovering neural correlates in human decision making based on brain decoding

2014 Fiscal Year Final Research Report

• Project/Area Number: 24300114
• Research Category: Grant-in-Aid for Scientific Research (B)
• Allocation Type: Partial Multi-year Fund
• Section: 一般
• Research Field: Bioinformatics/Life informatics
• Research Institution: Kyoto University
• Principal Investigator: ISHII Shin 京都大学, 情報学研究科, 教授 (90294280)
• Project Period (FY): 2012-04-01 – 2015-03-31
• Keywords: 意思決定 / ブレインデコーディング / ブレインマシンインターフェース / 非侵襲脳活動計測 / 強化学習

Outline of Final Research Achievements

We performed a decoding study of next scenes to be predicted by human subjects from their frontal-parietal fMRI activities when the subjects were performing a navigation task in partially observable maze environments. We performed a decoding study of various kinds of cognitive tasks by a deep neural network trained by large-scale fMRI data, and visualized high-dimensional features embedded in the trained deep network. Moreover, we constructed an online decoding system of spatial attention based on EEG (electroencephalography), which is applicable to real-world navigation. Using resting-state activities, we could calibrate the difference in neural correlates between human subjects.

Free Research Field

計算神経科学、人工知能、機械学習、ニューロインフォマティクス