2021 Fiscal Year Final Research Report
Investigation on brain networks in task processing based on dissipation from default modes
| Project/Area Number |
18K11450
|
|---|
| Research Category |
Grant-in-Aid for Scientific Research (C)
|
| Allocation Type | Multi-year Fund |
|---|
| Section | 一般 |
|---|
| Review Section |
Basic Section 61030:Intelligent informatics-related
|
|---|
| Research Institution | Fukui University of Technology |
|---|
Principal Investigator |
|
|---|
| Co-Investigator(Kenkyū-buntansha) |
高橋 哲也 福井大学, 学術研究院医学系部門, 客員准教授 (00377459)
信川 創 千葉工業大学, 情報科学部, 准教授 (70724558)
|
|---|
| Project Period (FY) |
2018-04-01 – 2022-03-31
|
| Keywords | 脳波 / デフォルトモード / 同期の逸脱 / 数字想起 / アルツハイマー型認知症 / 機械学習 |
|---|
| Outline of Final Research Achievements |
This study is to classify the machine in EEG analysis. At first, open data (3 electrodes) of event-related potential EEG with numbers from 1 to 9 were analyzed using Multiscale Entropy as a measure of complexity, and Gaussian process regression was performed for each subject to calculate the regression loss. Then, all subjects were found to have the potential to discriminate visibility at any one electrode. Next, we classified cognitive severity and mildness based on EEG of Alzheimer's disease. The Phase Lag Index as a measure of synchrony was calculated from 16-electrode measurements, and a combination of electrode-specific arithmetic mean and the t-SNE method was found to significantly improve the classification loss value.
|
| Free Research Field |
知能情報学
|
| Academic Significance and Societal Importance of the Research Achievements |
事象関連電位の3電極脳波で1から9までの数字視認では，どの被験者も複雑性指標のMultiscale Entropy値と機械学習のガウス過程回帰によって，3電極のうちどれか1つの電極で視認の判別の可能性があることが分かった。これは脳波からブレインマシンインターフェイスの可能性があることを示せた。
アルツハイマー型認知症の脳波からの認知重軽度の分類では，同期性の指標であるPhase Lag Index値を求め電極別算術平均と機械学習のt-SNE法を組み合わせることで，機械分類の損失値が大幅に改善された。この解析手法は認知機能の状態を脳波から定量的に評価する上で有効な手段になり得ることを示せた。
|
