脳波の時間周波数解析システムの開発とその解析結果

概要

論文の詳細を見る
Presently, the short time Fast Fourier Transform (FFT) is applied to analyze EEG as a major frequency analyzing method. The time-frequency analyzing methods have been developed recently as the Cohen class based on the Wignar distribution. In this paper, a modified time-frequency analysis for EEG is presented with development for the field application and some given results. In the first stage of this study, the EEG analyzing system, code name KOMA 04, is developed on the MATLAB platform for time-frequency analysis of EEG. To verify the properties of the KOMA 04 system, a signal generation system to simulate EEG is also developed and the specification of the KOMA 04 is evaluated exhaustively. As the second stage, actual EEG were analyzed by using the KOMA 04 system, where EEG were collected from university students (five males and four females) with 15〜30 minutes record time by stimulation given 10〜20 oral questions and special sound. Some results are described in this paper. We observed new phenomena which will become a problem about the relation between EEG and psychological processes. The precise time-frequency analysis of EEG is developed as shown and further studies will be directed to the causal relation between human psychological state and EEG as problems by given results.
駒澤大学の論文