應用視覺誘發腦波之高傳輸率腦波人機介面設計

The Design of Visual Evoked Potential-Based Brain Computer Interface with High Information Transfer Rate

李柏磊(Po-Lei Lee) ； 徐國鎧(Kuo-Kai Shyu)

《科儀新知》 197期 (2013/12) Pp. 82-93

摘要

在本文中筆者將介紹以視覺誘發電位為基礎的大腦人機介面，這個系統藉由使用LED或LCD螢幕作為產生閃光刺激的來源，以呈現多重時序的閃光刺激的方式，來控制此大腦人機介面。使用者可以利用眼睛注視所欲選擇的選項，以引發特定的視覺誘發電位，這些特定的視覺誘發電位訊號在經過判別之後，將可送出對應的指令給輸出裝置，以達到溝通與資訊傳輸的目的。這種視覺誘發腦波人機介面，只需要簡單數個試驗的平均，就可以將視野中央的閃光所誘發的視覺誘發電位萃取出來，並衰減視野周圍的閃光刺激所引發的訊號。經過判別可以選擇具最大視覺誘發訊號的閃光選項為目前受測者正在注視的閃光選項，並輸出其對應的指令。本大腦人機介面系統可以讓肢體殘障的病人，能夠利用腦波快速且正確地與其外在環境溝通。

並列摘要

In this article, we introduce a novel visual evoked potential (VEP)-based brain computer interface (BCI) which allows paralyzed patients to communicate with external environments, independent of peripheral nerves and muscles. In the present systems, flashing stimuli, such as digits or letters, are displayed on a LED array or a LCD screen to induce user's visual evoked potentials (VEPs). User can generate desired commands one after one while he/she stares at target stimulus. Since VEPs are time-locked and phase-locked to the flash onsets of visual stimuli, VEPs induced by gazed targets, having greater amplitudes due to cortical magnification, can be extracted using a simple averaging process to remove task-unrelated noise. The commands corresponding to the detected gazed targets are then identified for paralyzed patients to communicate with external environments. The present VEP-based BCI systems establish a reliable communication channel for users to control environmental devices via their brain waves.