Comparative Analysis of Machine Learning Models for Predicting Depressive States from EEG Signals

カテゴリ: 部門大会

論文No: SS1-6

グループ名: 【C】2024年電気学会電子・情報・システム部門大会

発行日: 2024/08/28

タイトル(英語): Comparative Analysis of Machine Learning Models for Predicting Depressive States from EEG Signals

著者名: Li Pengcheng（東京工業大学）,Connelly Akima（東京工業大学）,Rangpong Phurin（東京工業大学）,Nakatani Hironori（東海大学）,Yagi Tohru（東京工業大学）

著者名(英語): Pengcheng Li (Tokyo Institute of Technology),Akima Connelly (Tokyo Institute of Technology),Phurin Rangpong (Tokyo Institute of Technology),Hironori Nakatani (Tokai University),Tohru Yagi (Tokyo Institute of Technology)

キーワード: depressive state|electroencephalogram (EEG)|feature extraction|machine learning|resting state

要約(日本語): Depression affects over 322 million people globally, with its prevalence increasing with age. Electroencephalography (EEG) is a non-invasive tool essential for identifying biomarkers in depression research. Using EEG signals and machine learning techniques such as neural networks, support vector machines, and logistic regression, research in this field aims to predict depression treatment outcomes and diagnose disorders effectively. However, it remains uncertain whether these techniques could detect depressive states in healthy individuals. This study investigates whether machine learning could classify features from one-minute, eyes-open EEG signals to effectively identify depressive states, potentially aiding early detection and prevention of depression.