Investigating the impact of feature extraction methods on prediction accuracy of neurological recovery levels in comatose patients post-cardiac arrest

Abstract

Cardiac arrest can cause irreversible Post-Cardiac Arrest Brain Injury (PCABI), but predicting PCABI with certainty remains challenging. This study aims to improve prognostication by predicting neurological recovery using EEG data from the 'I-CARE: International Cardiac Arrest Research Consortium Database.' Data were preprocessed with an FIR Equiripple Bandpass Filter, and three feature extraction methods were applied. Decision Tree, KNN, SVM, and Ensemble Learning algorithms were evaluated using F1-Score, Accuracy, and ROC-AUC. The highest accuracy, 0.89, was achieved with Hamming-windowed streamline feature extraction and Decision Tree after feature selection.