FNIRS-Based Binary Communication Using Mental Imagery

Abstract

Brain-computer interfacing in patient populations should be straightforward and enable in-session communication. An fNIRS-based binary communication paradigm is presented that requires limited preparation time and employs merely nine optodes on the left parietal-frontal hemisphere. Eighteen healthy participants performed two mental imagery tasks, mental drawing and spatial navigation, during auditorily cued time windows to answer yes/no questions. Participants’ answers were decoded in simulated real-time with a general linear model. A multi-trial group accuracy of 67% was reached using the oxygenated hemoglobin signal, whereas seven participants showed significant response in the deoxygenated hemoglobin signal. A simulated real-time multivariate approach was explored and showed discernibility of the two tasks, based on spatial patterns solely, in seven subjects. Given the short preparation time and limited optode set-up, the presented paradigm has potential to enable communication in affected patient populations.