次世代智慧型加護病房照護系統-子計畫九：應用於智慧型加護病房之腦神經活動監測系統( I )

Abstract

獨立成份分析法被廣泛用於去除腦電磁波訊號中的雜訊成份，其分離出的時序訊號具有獨立的特性，因此被稱為獨立成份，每個獨立成份有一組對應的頭皮分佈，由於頭皮分佈是由該獨立時序分佈於各個感測器上的強度值，亦即同一個獨立成份可能是由腦中具有相同時序波形但分佈於不同腦區的訊號源所構成，故本計劃利用各獨立成份的皮質分佈來找出有相同活動的腦區，在進一步整合所有獨立成份的皮質活動分佈資訊後，希望可找出有高度相關活動的區域以建構神經網路。在本計畫執行過程中，我們已開發出一個獨立成份時空造影演算法，並利用模擬資料來驗證該演算法的準確性，接下來我們將開發神經網路建構技術，該技術將利用此我們開發的獨立成份時空造影演算法所估算之獨立成份皮質分佈，據此資訊進行神經網路之估算，最後將應用於性別辨認實驗中所蒐集的腦磁波資料以驗證方法的可行性。

Independent component analysis (ICA) has been widely used to alleviate interference caused by noise components from the electromagnetic recordings of brain activity. Each independent component includes a temporal waveform and its scalp topography. The temporal waveforms decomposed from electromagnatic signals by ICA are independent. The scalp distribution is the strength of the temporal waveform on each sensor location. In other words, an independent component may be constituted by multiple sources in different brain areas with the same source activity. Thus the independent component can provide the information for estimating brain network. In this project, we proposed to develop a method to estimate the cortical mapping of an independent component, which is then used to statistically guide the selection of cortical brain areas with highly similar temporal activity and the estimation of brain functional network. Until now, we have developed an algorithm to calculate the cortical source distribution of an independent component. Moreover, we have evaluated the accuracy of the proposed method by using simulation studies. In the second half year, we will focus on developing the algorithm of functional neural network estimation. To evaluate the feasibility of the proposed algorithm, it will be applied to the magnetoencephalography data acquired in the gender experiment.