정합필터 기반 신호처리를 활용한 동작 상상 기반 fNIRS 신호의 사용자 상태 구분

Abstract

Functional near-infrared spectroscopy (fNIRS) has the potential for the application of the brain - computer interface (BCI). Statistical features are often extracted from NIRS measurements to classify a signal in fNIRS-BCI. This paper introduces a matched filter-inspired feature extraction method and compares it to the conventional method. Experimental results demonstrate that the proposed method shows a higher accuracy on average than the conventional method in motor imagery signal classification.