Novel Approach for Brain Computer Interfacing

Brain computer interfacing (BCI) plays an important role as an assistive device for various neural disorder people. In this paper, attempts have been made to interface thought process generated in photoplethysmography (PPG) with machine/computer. PPGs were collected by using PPG electrodes on fingertips corresponding to two types of brain activities, planning and relax. The proposed scheme is based on discrete wavelet transform (DWT) analysis. Features like entropy, energy and standard deviation at each node of the decomposition tree were calculated followed by application of SVM for classification. Clinical PPG signal recorded from planning as well as relax state are used to test the performance of this new scheme. The testing accuracy of this wavelet based SVM classifier is found to be 90% accurate. The proposed model is user friendly, cost effective and more accurate compared to its EEG counterpart.