Functional near-infrared spectroscopy short-channel regression improves cortical activation estimates of working memory load

Functional near-infrared spectroscopy (fNIRS) is a noninvasive technique commonly used to examine cognitive functions such as working memory (WM). However, fNIRS signals are often interfered with by extracerebral activity, such as scalp hemodynamics. Short separation channels (SSCs) allow direct measurement of these signals. Short-channel regression (SCR) is widely used to reduce scalp interference, but its added value in WM paradigms remains underexplored. We aimed to examine the effect of SCR on improving the validity of fNIRS measurements for WM load (WML). We used the N -Back task to induce WML-dependent brain activation by varying the " n " level. Data from 20 participants were collected using fNIRS with SSC. Hemodynamic responses were analyzed with generalized linear models and linear mixed models to assess SCR's effect on the sensitivity of cortical activation measures. SCR enhanced the statistical effects of N -Back levels on measured hemodynamic responses at both group and subject levels, improving the validity and sensitivity of fNIRS. SCR improves fNIRS measurement sensitivity and validity, even in tasks with minimal motor requirements.