Filtering chest compression artifacts improves the performance of VF-detection parameters

Cardiopulmonary resuscitation (CPR) artifact filtering techniques have not been successfully combined with commercial shock advice algorithms (SAA) to diagnose the rhythm during CPR. Recently, a promising new approach based on using SAAs especially designed to diagnose the filtered ECG has been introduced. This study evaluates the impact of filtering CPR artifacts on the shock/noshock decision for several well-known VF-detection parameters.The detection accuracy of 22 VF-detection parameters was calculated for artifact-free ECG segments and for ECG segments corrupted by chest compressions before and after filtering. Performance was measured in terms of: area under the curve of the receiver operating characteristic curve, and sensitivity/specificity for the shocklnoshock decision. Filtering the CPR artifact improved the detection capacity of most parameters, and showed that combining features after filtering may be a successful strategy.