Functional near-infrared spectroscopy sensor using a combined short-channel regression and frequency-division multiplexing

We designed and developed an fNIRS system that suppresses this noise by integrating a short-channel-based signal correction method and frequency division multiplexing (FDM). The results indicated that after applying short-channel regression, the effect size (Cohen’s d) increased by approximately 0.9 compared to the measurements taken before applying short-channel regression. Additionally, a significant difference in hemodynamic responses was observed between the digit span task and the control task. These findings suggest that the proposed fNIRS device is unaffected by ambient light and that short-channel regression effectively reduces noise interference. Consequently, this system can deliver accurate and stable measurements of brain activity, even in conditions where managing ambient light and surface noise is complicated.