多通道行動與無線EEG系統之新式設計

Abstract

在過去的研究中，大多數的學者使用包含強大功能的標準化量測儀器來收錄高訊號品質的腦電位(EEG)訊號；然而，由於此套腦電位系統較笨重而巨大，這也造成在日常生活當中收錄腦波變成一項困難的任務。因此我們提出一套多通道行動與無線腦電位系統足以容易應用於醫療用途、居家照護控制介面，以及科學認知實驗等等。此套系統包含三個部份：(1)乾電極、(2)多通道行動無線腦電位擷取電路裝置，和(3) 腦電位展示程式。其中乾電極不需要依靠電極膠來降低皮膚與電極的阻抗值的前置準備工作，卻比起濕電極有更好的長時間偵測品質。而實作出來的電路裝置能為16個輸入通道提供服務，並且體積小、重量輕巧。此外，後端展示程式能透過無線通訊介面接收腦電位訊號，並且將數位資料存成檔案以利線下分析。 在本研究中，我們也設計出一套系統驗證的方法逐步為我們的系統驗證出效能。從最後的結果中，我們得知此套系統有72-95%的程度相似於Neuroscan公司出產的標準化量測系統；而結合在多通道腦電位擷取電路裝置的乾電極不僅可以量測到與濕電極相似的腦電位訊號特徵，也能在oddball認知實驗當中量測到相當微小的事件相關電位(ERP)訊號，並與濕電極所量到的訊號有93-95%的相似程度。總結來說，我們宣稱此套系統量測腦電位確實有可靠的訊號品質，並且可以方便地應用於日常生活當中。

In past study, most of researchers used standard measurement instrument which produced strong functions to monitor EEG activity with high signal quality. However, recording EEG trends to be a difficult task in dairy life during to hardly moving the heavy, huge, and wired EEG system. Hence, we mention a multi-channels mobile and wireless EEG system which can be easily used for medical application, home care control interface, cognitive experiments, and so on. The system includes three parts: (1) dry sensors, (2) multi-channels mobile and wireless EEG acquisition circuitry, and (3) EEG display program. Dry sensor needs not any skin preparation with conductive gels to maintain small skin-sensor impedance but still make the better long-term monitoring performance than wet sensor. The circuitry is implemented to serve 16-channel inputs and has a miniature area and light. In addition, back-end display program receives EEG data via wireless communication and can also save digital data as a file for off-line analysis. In this study, we also design the method of system verification to verify the performance of our system step by step. Finally, the results show that our circuitry and program is 72-95% correlate to standard measurement instrument like Neuroscan system. Dry sensor applied in multi-channels device can not only monitor the similar EEG feature with wet sensor but also detect pretty tiny EEG signal even ERP in Oddball task with 93-95% correlation with wet sensor. In conclusion, we demonstrate our proposed system is exactly reliable and can be conveniently applied in dairy life.