Denoising of neuronal signal from mixed systemic low-frequency oscillation using peripheral measurement as noise regressor in near-infrared imaging

Abstract

Functional near-infrared spectroscopy (fNIRS) is a noninvasive functional imaging technique measuring hemodynamic changes including oxygenated (O 2 Hb) and deoxygenated (HHb) hemoglobin. Low frequency (LF; 0.01 to 0.15 Hz) band is commonly analyzed in fNIRS to represent neuronal activation. However, systemic physiological artifacts (i.e., nonneuronal) likely occur also in overlapping frequency bands. We measured peripheral photoplethysmogram (PPG) signal concurrently with fNIRS (at prefrontal region) to extract the low-frequency oscillations (LFOs) as systemic noise regressors. We investigated three main points in this study: (1) the relationship between prefrontal fNIRS and peripheral PPG signals; (2) the denoising potential using these peripheral LFOs, and (3) the innovative ways to avoid the false-positive result in fNIRS studies. We employed spatial working memory (WM) and control tasks (e.g., resting state) to illustrate these points. Our results showed: (1) correlation between signals from prefrontal fNIRS and peripheral PPG is region-dependent. The high correlation with peripheral ear signal (i.e., O 2 Hb) occurred mainly in frontopolar regions in both spatial WM and control tasks. This may indicate the finding of task-dependent effect even in peripheral signals. We also found that the PPG recording at the ear has a high correlation with prefrontal fNIRS signal than the finger signals. (2) The systemic noise was reduced by 25% to 34% on average across regions, with a maximum of 39% to 58% in the highly correlated frontopolar region, by using these peripheral LFOs as noise regressors. (3) By performing the control tasks, we confirmed that the statistically significant activation was observed in the spatial WM task, not in the controls. This suggested that systemic (and any other) noises unlikely violated the major statistical inference. (4) Lastly, by denoising using the task-related signals, the significant activation of region-of-interest was still observed suggesting the manifest task-evoked response in the spatial WM task.

The impact of physiological noise on hemodynamic-derived estimates of directed functional connectivity

“…In summary, the high temporal resolution of fNIRS renders Granger-causality analyses of hemodynamic measurements possible and allows the comparison of different conditions provided that physiological parameters like aBP are controlled. In line with previous studies we demonstrated that (i) peripheral measurements of systemic hemodynamic processes can be used to correct functional connectivity estimates for physiological noise (Frederick et al 2012;Tong et al 2013;Sutoko et al 2019) and that (ii) ICA stenosis impairs functional network organization (Avirame et al 2015).…”

“…Furthermore, the PDC estimates (i.e. the connectivity corrected for aBP influences) were similar for the oxyHb, dxyHb, and CBSI data suggesting that the deviating results for the oxyHb-derived DC estimates were due to the higher susceptibility of the oxyHb measurement to physiological noise as previously reported (Obrig et al 2000;Zhang et al 2009;Kirilina et al 2012;Sutoko et al 2019). This again corroborates our finding that including the aBP signal in PDC estimation effectively controlled for bias induced by physiological noise.…”

“…Optical measurements of blood oxygenation at the periphery (e.g. the finger) provide an easy way to capture a wide spectrum of systemic hemodynamic processes and can be used to reduce physiological noise in functional measurements (Frederick et al 2012;Tong et al 2013;Sutoko et al 2019).…”

Short-channel regression in functional near-infrared spectroscopy is more effective when considering heterogeneous scalp hemodynamics

“…The GLM method is a commonly known technique that allows the regression of short channels and that is relatively easy to implement. 12 , 38 , 45 – 47 It is known that the regression of different regressor signals may alter the shape of the recovered signal. 36 Therefore, we applied five regressor sets incorporating different assumptions on the spatial distribution of systemic interference.…”

“…Systemic artifacts are typically not constrained locally, but they affect the whole brain and extracerebral tissues, and are thus considered “global.” As a consequence, it is often assumed that global noise is distributed homogeneously over the entire scalp layer and thus that a single global signal can be used for the short-channel regression. 26 – 30 In contrast, it was repeatedly shown that scalp hemodynamics follow spatially heterogeneous patterns, 9 , 17 , 31 – 38 i.e., different short-channel signals are measured at different locations on the scalp. An additional complexity is observed when considering that the physiological artifacts also show frequency-dependent spatial variations, 36 with a noteworthy effect observed for the MW-frequency band (i.e., average time lags up to 2 s between ipsilateral head regions 36 ).…”