Functional near-infrared spectroscopy based discrimination of mental counting and no-control state for development of a brain-computer interface.

In this paper we propose to apply functional near-infrared spectroscopy (fNIRS) to measure the brain activity during mental counting and discriminate it from the no-control (rest) state, which could potentially lead to a two-choice brain-computer interface (BCI) application. fNIRS is a relatively new optical brain imaging modality that can be used for BCI. The major advantages using fNIRS are its relatively low cost, safety, portability, wearability and overall ease of use. In the present research, five healthy subjects are asked to perform mental counting during the activity period. Signals from the prefrontal cortex are acquired using a continuous-wave imaging system. The mental counting and no-control states are classified, using linear discriminant analysis (LDA), with an average accuracy of 80.6%. These classified signals can be translated into control commands for a two-choice BCI. These results show fNIRS to be a potential candidate for BCI.