Neurovascular coupling (NVC) in newborns using processed EEG versus amplitude-EEG

Abstract

Abstract There is a critical need for development of real time physiological biomarkers for birth asphyxia that constitutes a major global public health burden. Our recent study (Scientific Reports, V10:9183, 2020) established a novel non-invasive neurovascular coupling (NVC) assessment in newborns using dynamic wavelet transform coherence (WTC) analysis irrespective of different aEEG algorithms. As an extended study, the current paper examines whether the variability in processed EEG and amplitude-EEG (aEEG) outputs would impact the determination of NVC in newborns with encephalopathy. Concurrent processed EEG tracings and regional near infrared spectroscopy (NIRS)-based cerebral tissue oxygen saturation (SctO2) readings during a period of twenty hours in their first day of life were selected and processed in this study. After bandpass-filtered in 2–15 Hz, rectified, and down-sampled at 0.21 Hz, the processed EEG tracings along with NIRS-SctO2 (0.21 Hz) were used to perform WTC analysis, followed by comparison of WTC-metrics between SctO2-processed EEG coherence and SctO2-aEEG coherence using Bland–Altman statistics. Our results demonstrated high and significant correlation (R2 = 0.96, p < 0.001) between NVC assessments by SctO2-processed EEG and SctO2-aEEG coherence, confirming that band-passed, rectified, and down-sampled processed EEG, or aEEG, can be paired with NIRS-SctO2 to assess NVC in newborns with encephalopathy. Findings indicate the feasibility of a simpler approach to NVC in neonates by using directly processed EEG, instead of aEEG.