Monitoring Chest Compression Rate in Cerebral Oximetry Signals during Cardiopulmonary Resuscitation using Wavelet Analysis.
Annual International Conference of the IEEE Engineering in Medicine and Biology Society. IEEE Engineering in Medicine and Biology Society. Annual International Conference · 2024-07-01 · Method & validation
Abstract
The high temporal resolution cerebral oximetry signal obtained through near-infrared spectroscopy reflects fluctuations due to chest compressions (CCs) during cardiopulmonary resuscitation (CPR). The aim of this study was to develop a method for CC rate calculation in out-of-hospital cardiac arrest (OHCA) patients using cerebral oximetry signals. The study database comprised 284 segments extracted from 30 OHCA patients, with each segment including concurrent cerebral oximetry and thoracic impedance signals. The proposed method analyzed 10-s nonoverlapping windows of the oximetry signal. First, the stationary wavelet transform was used to extract the CC-related component from cerebral oximetry. Then, quality control was performed to avoid artifacts and/or low quality signals. Finally, the CC frequency, fCC, was computed as the maximum argument of the spectrum. The performance of the method was assessed in terms of: median (interdecil range, IDR) of the absolute error between the estimated fCC and the reference CC rate calculated from the thoracic impedance in compressions per minute (cpm), and Bland-Altman analysis together with its corresponding limits for the 90% level of agreement, LOA90%. The median (IDR) absolute error was 0.62 (0.10-3.95) cpm with a LOA90% of -2.72-4.74 cpm that covers 90.27% of the total windows that met the quality control criterion. These results demonstrate the robustness and accuracy of the method that could be integrated into cerebral oximetry monitoring systems during CPR.