EEG空間解析度對於腦電位分佈圖的研究

Abstract

本篇論文的目的是以腦電位分布圖來探討EEG的品質和可靠性。在這項研究中，我們將會分析資深禪定者所錄製的30-channel EEGs。首先利用離散時間小波轉換 (discretization-time wavelet transform) 將EEG訊號分解成五個頻帶分別為delta, theta, alpha, beta, gamma。然後利用Spherical-splines來重建五個頻帶的EEG訊號，最後算出EEG訊號的能量並繪出腦電位分布圖。為了觀察我們從16-channel重建的EEG腦電位分布圖的可靠性，我們計算出EEG腦電位分布圖的重心位置來做量化分析。根據我們從十位資深禪定者所錄製的EEG訊號所分析的研究結果，我們認為在腦電位分布圖中theta頻帶的能量可能是EEG五個頻帶中較具有訊號重建的可靠性。

This thesis is aimed to report our investigation of the effect of spatial resolution on the reliability and quality of EEG (electroencephalograph) topographical mapping. In the study, 30-channel Chan-meditation EEGs were analyzed. Discrete wavelet transform (DWT) was first applied to decompose EEGs into five EEG rhythms, delta, theta, alpha, beta, and gamma. Spherical-splines interpolation was then adopted to construct the brain topographical mapping of the EEG powers for each EEG rhythm. To evaluate the reliability of EEG topographical mapping reconstructed from 16-channel EEGs, we quantified the center of gravity of the reconstructed brain topographical mapping. According to the results of analyzing the EEG signals recorded from ten meditation practitioners, we concluded that brain topographical mapping of theta-rhythmic power could be most reliably reconstructed by fewer number of EEG channels.