以腦波撥號技術探討

A study of brainwave-based dialing techniques

黃俊憲(Jun-Hsien Huang)

指導教授 ： 蔡偉和

國立臺北科技大學/電資學院/電腦與通訊研究所/碩士(2013年)

摘要

本研究探討如何透過腦波，令使用者下達操作指令進行撥號動作，藉以達到意念控制的目標。本文先以腦波做為選擇指令，使指令辨識問題從多重選擇簡化至是與否的是非題，例如在電話撥號系統中，共有10個數字指令，系統依序提示數字，當出現目標數字時，使用者的腦波應當呈現「是」的型態，反之則「否」，此方法稱之為二值意念選擇。
　　過去以腦波做為指令的研究中，其量測方法大多是透過多通道的電極配置，必須使用價格昂貴的腦波儀，量測地點也受限制，本文嘗試以單極記錄的方式，使用輕便、廉價之腦波儀，擷取受測者前額葉β波做為特徵訊號，並使用遞迴式類神經網路及隱藏式馬可夫模型辨識其指令，建構一個不限定使用者的腦波辨識系統，本實驗收集了11位受測者，在44個測試樣本內，使用遞迴式類神經網路辨識樣本有77.27%辨識率，而使用隱藏式馬可夫模型作為辨識工具有本研究80.68%最高辨識率。

關鍵字

腦波 ； 隱藏式馬可夫模型 ； 遞迴式類神經網路 ； 單極腦電圖

並列摘要

This thesis investigates how we can recognize a user’s intentions to select from a set of digits by measuring his/her brainwaves. This could help handicapped, hand-busy, or even eye-busy people dial by just thinking of the number to call. Our strategy is to convert a multiple-choice decision into a yes-no decision. Specifically, the proposed system prompts the user to select from each of the digits, and then measures and analyzes his/her electroencephalogram (EEG) in order to determine if a certain digit matches the user’s intention of “Yes”. Although several studies discussing such kind of binary EEG classification problems preexist, all of them use 32 or more EEG channels to develop their systems, which involve inconvenient and uncomfortable head-recording nets as well as expensive equipment, therefore unsuitable for real human-machine-interfaced applications. In contrast, this study uses a simple, portable, and cheap instrument that extracts a single-channel EEG from the user’s frontal lobe. The underlying beta waves of EEG are then distilled and used as the feature to determine the user’s intention. We develop two user-independent brainwave recognition systems, respectively, based on recurrent neural network (RNN) and hidden Markov models (HMMs). Our experiments conducted using 44 test EEG samples from 11 subjects show that the recognition accuracy obtained with RNN-based system and HMM-based system are 77.27% and 80.68%, respectively.

並列關鍵字

brainwave ； hidden Markov model ； recurrent neural network ； single-channel electroencephalogram

參考文獻

[26]鄭旺彬，利用腦波訊號識別指令，碩士論文，國立臺北科技大學電腦與通訊研究所，臺北，2012。

[1]J. R. Wolpaw, N. Birbaumer, D. J. McFarland, G. Pfurtscheller and T. M. Vaughan, “Brain-computer interfaces for communication and control,” Clin. Neurophysiology, vol. 113, 2002, pp. 767-791.

[2]A. Kübler, B. Kotchoubey, T. Hinterberger, N. Ghanayim, J. Perelmouter, M. Schauer, C. Fritsch, E. Taub and N. Birbaumer, “The thought translation device : A neurophysiological approach to communication in total motor paralysis,” Experimental Brain Research, vol.124, no.2, 1999, pp. 223-232.

[4]M. A. L. Nicolelis, and J. K. Chapin, “Controlling Robots with the Mind,” Scientific American, vol. 287, 2002, pp. 46-53.

[5]B. Obermaier, C. Neuper and G. Pfurtscheller,” Information Transfer Rate in a Five-Classes Brain–Computer Interface,” IEEE Trans. Rehab. Eng., vol. 9,no. 3, 2001, pp. 283-288.