Independent component analysis based trial-to-trial variations reduction in fNIRS signal

Functional near-infrared spectroscopy (fNIRS) is emerging optical brain imaging technique. It measures the hemodynamic changes that effectively reflect the brain states. However, the fNIRS signal analysis is sensitive to the trial-to-trial variability (TTV), and the source of the TTV remains elusive. Previous functional magnetic resonance image (fMRI) study suggested that the TTV can be attributed to spontaneous low frequency hemodynamic fluctuation in fMRI signal. Meanwhile, recent fNIRS studies confirm that the spontaneous fluctuation in fNIRS signal can be robustly detected by functional near infrared spectroscopy (fNIRS). In this paper, we investigate the relationship between TTV and the spontaneous fluctuations in fNIRS signal. The subject is asked to complete a right finger tapping experiment. Independent component analysis as well as functional connectivity is used to reduce the TTV in the finger tapping experiment. The result suggests that the low frequency spontaneous fluctuation contribute significantly to the TTV in fNIRS signal.