Time-integrated and time-resolved wavelet coherence analysis to quantify cerebral autoregulation: Algorithm review, new development, and prediction of pediatric cardiac arrest

Abstract Wavelet transform and wavelet transform coherence (WTC) are advanced data analysis tools in the time-frequency domain for studying nonstationary time series and the relationship between two synchronous time series. These advanced tools can be used to quantify cerebral autoregulation (CAR) by calculating the mean arterial pressure (MAP) and regional cerebral oxygenation (StO2). The goal of this study was to develop a novel WTC analysis method that enables the quantification of disrupted CAR in pediatric cardiac arrest. Accordingly, this paper first provides a narrative explanation of a well-developed WTC computational package using a graphical flowchart for clinicians and life scientists to better comprehend the analysis rationale. Next, we introduce a new analysis strategy to quantify time-resolved significant coherence (trSC) which facilitates the visualization of temporal changes in CAR, followed by a clinical example using trSC to detect impaired CAR in pediatric cardiac arrest. Specifically, we quantified 25-hour temporal traces of trSC using concurrently recorded MAP and StO2 and examined CAR impairment during pre-, peri-, and post-cardiac arrest in 20 arrest events in 16 pediatric patients. This study demonstrated that trSC is a promising parameter for investigating CAR and can be developed as a bedside monitoring tool for clinicians.