Analysis of scalp electroencephalogram used in BCI

Abstract

In late years, a study of BCI (Brain Computer Interface) which can perform the apparatus control of the computer by using a signal occurring with the activity of the brain is performed flourishingly. It is expected that BCI technique will make possible to support of communication with general-paralysis patients. There are two types of BCI. One is an invasive type which can gain in high quality and stable brain signals, and the other is a non-invasive type which uses EEG (Electroencephalogram) measured from scalp. However, invasive type BCI requires a surgery which becomes cause of patient’s burden. Therefore, the objective of this study is a development of a BCI system using EEG which is non-invasively measured from a scalp. The EEG measurement system for BCI was experimentally developed and assessed. In order to find features of EEG, the EEG which was measured when a subject held his/her hand and the EEG which was measured when a subject imagined holding his/her hand were analyzed by the system. As a result, we could confirm the differences between right hand holding and left hand holding. And we also found that wide area of the brain tended to be activated when a subject held not the dominant hand. A similar tendency was found when a subject imagined holding his/her hand. Furthermore, the tendency for frequency band of μ-wave to decay was observed in around motor cortex. From these results, we think that the development of the BCI system which uses these experimental results is possible I near future.