Signal regression in frequency-domain diffuse optical tomography to remove superficial signal contamination

Significance: Signal contamination is a major hurdle in functional near-infrared spectroscopy (fNIRS) of the human head as NIR signal contaminated with changes corresponding to superficial tissue, therefore occluding information originating from cerebral region. For continuous wave, this generally handled through linear regression shortest source-detector (SD) distance intensity measurement all signals. Although phase measurements utilizing frequency domain (FD) provide deeper tissue sampling, use SD for can lead misleading results, suppressing cortical Aim: An approach FD fNIRS that utilizes short-separation directly regress both and measurements, providing better data-types, proposed. Approach: Simulated data realistic models are used, using phase-based components evaluated. Results: Intensity-based achieves suppression by 68% whereas only 13%. Phase-based also shown generate false-positive signals cortex, which not desirable. Conclusions: provides methodology minimizing fNIRS.