Classification of Functional Near-Infrared Spectrocopy Signals during Passive and Combinatory Exercises for Neurorehabilitation

Abstract

In this study, we evaluated and classified hemodynamic responses induced by conventional passive exercise for neurorehabilitation and combined exercise strategy (passive exercise with active motor execution or motor imagery). Functional near infrared spectroscopy (fNIRS) was recorded while eight healthy subjects conducted three different tasks (passive motor execution alone, passive motor execution with motor imagery, and passive motor execution with active motor execution). From the results, stronger and broader activation around the sensorimotor cortex was observed when subjects performed the combinatory exercises. Results of pattern classification showed classification accuracy higher than 80 %, demonstrating that fNIRS could be used as a potential tool to assess users' cognitive engagement in the combinatory neurorehabilitation strategy. ? 2019 IEEE.