A reliable method for rhythm analysis during cardiopulmonary resuscitation

Area of Science:

• Emergency Medicine
• Biomedical Engineering
• Cardiology

Background:

• Interruptions in cardiopulmonary resuscitation (CPR) reduce defibrillation success rates.
• Current rhythm analysis methods during CPR have limitations in specificity for nonshockable rhythms.

Purpose of the Study:

• To develop and evaluate a novel approach for rhythm analysis during CPR with improved specificity.
• To enhance the accuracy of shock advice algorithms (SAA) during cardiac arrest resuscitation.

Main Methods:

• A state-of-the-art CPR artifact suppression filter was combined with a novel shock advice algorithm (SAA).
• The SAA incorporated a low electrical activity detector and a support vector machine classifier using slope and frequency features.
• Analysis was performed on 1185 shockable and 6482 nonshockable 9-s segments from out-of-hospital cardiac arrest patients.

Main Results:

• The new approach achieved 91.0% sensitivity and 96.6% specificity for rhythm analysis during CPR on the test set.
• This represents a significant increase in specificity compared to previous studies.
• Sensitivity was maintained without compromise.

Conclusions:

• The developed CPR rhythm analysis approach significantly improves specificity for nonshockable rhythms.
• This advancement can lead to more effective defibrillation strategies during resuscitation.
• The findings support the integration of this algorithm in automated external defibrillators.