Characterization of Mayer wave oscillations in functional near-infrared spectroscopy using a physiologically informed model of the neural power spectra

Abstract

OverviewO_ST_ABSSignificanceC_ST_ABSMayer waves are spontaneous oscillations in arterial blood pressure that can mask cortical hemodynamic responses associated with neural activity of interest. AimTo characterize the properties of oscillations in the fNIRS signal generated by Mayer waves in a large sample of fNIRS recordings. Further, we aim to determine the impact of short-channel correction for the attenuation of these unwanted signal components. ApproachMayer wave oscillation parameters were extracted from 310 fNIRS measurements using the Fitting Oscillations & One-Over-F (FOOOF) method to compute normative values. The effect of short-channel correction on Mayer wave oscillation power was quantified on 222 measurements. The practical benefit of the short-channel correction approach for reducing Mayer waves and improving response detection was also evaluated on a subgroup of 17 fNIRS measurements collected during a passive auditory speech detection experiment. ResultsMayer-wave oscillations had a mean frequency of 0.108 Hz, bandwidth of 0.04 Hz, and power of 3.5 M2/Hz. The distribution of oscillation signal power was positively skewed, with some measurements containing large Mayer waves. Short-channel correction significantly reduced the amplitude of these undesired signals; greater attenuation was observed for measurements containing larger Mayer-wave oscillations. ConclusionsA robust method for quantifying Mayer-wave oscillations in the fNIRS signal spectrum was presented and used to provide normative parameterization. Short-channel correction is recommended as an approach for attenuating Mayer waves, particularly in participants with large oscillations.