A Preliminary Study of Brain-Computer Interface Paradigm Based on Electrical Somatosensory Modality

Abstract

Brain-computer interfaces( BCIs) usually apply the brain signals in responding to various sensory inputs in visual or auditory modalities to provide direct communication pathways to external devices.No early study shows the possibility of using electrical stimuli as BCI input. In this study,we adopted the electrical somatosensory stimuli as the BCI input to elicit event-related potential( ERP). Three-condition experiment was conducted using visual,auditory and electrical stimuli individually in each condition. We compared the ERP components of each condition as well as the classification accuracy of these three conditions. Results show that electrical stimuli could provide relatively high amplitude and stable latencies of ERP components. It also enjoys higher class-discriminative information and classification accuracy than auditory paradigms. Thus,electrical paradigms with different stimuli intensities could be a good choice for BCI applications,which will enlarge the options for BCI purpose.

大腦對各類感覺輸入(視、聽模態)會產生不同的響應信號,腦-機接口正是利用這一響應信號實現大腦與外部設備間直接的通訊.然而以電刺激作為腦-機接口的輸入模態還未有報道.本研究嘗試通過使用體感電刺激作為腦-機接口的輸入,從而誘發事件相關電位(ERP).在整個實驗中,分別使用視覺、聽覺以及電刺激作為誘發因素,針對每種條件下的事件相關電位及其分類準確率開展對比分析.結果顯示電刺激所誘發的事件相關電位幅值較高且具有相對穩定的潛伏期,其分類準確率高于聽覺刺激范式.也就表明了以不同刺激強度作為參數的電刺激范式作為腦-機接口應用的可行性,這將進一步拓展腦-機接口的應用領域.