Reducing CPR artefacts in ventricular fibrillation in vitro

Area of Science:

• Cardiology
• Biomedical Engineering
• Signal Processing

Background:

• Cardiopulmonary resuscitation (CPR) generates electrocardiogram (ECG) artifacts, necessitating pauses in chest compressions for rhythm analysis.
• These hands-off intervals during CPR compromise vital organ perfusion.
• Existing fixed coefficient filters are ineffective in humans due to overlapping frequency spectra of CPR artifacts and ventricular fibrillation (VF) signals.

Purpose of the Study:

• To develop and evaluate a novel digital adaptive filter for removing CPR-induced artifacts from human ECG signals.
• To compare the performance of the adaptive filter against a fixed coefficient filter in the presence of CPR artifacts.
• To assess the filter's efficacy across a range of signal-to-noise ratios (SNRs).

Main Methods:

• CPR artifacts from porcine models were mixed with human VF signals at SNRs ranging from -10 dB to +10 dB.
• A digital adaptive filter was developed to remove CPR artifacts.
• The adaptive filter's performance was compared to a fixed coefficient filter.

Main Results:

• The adaptive filter significantly outperformed the fixed coefficient filter in removing CPR artifacts across all tested SNR levels.
• At an initial SNR of 0 dB, the adaptive filter restored SNRs to 9.0 ± 0.7 dB, compared to 0.9 ± 0.7 dB for the fixed filter (P < 0.0001).
• The adaptive filter demonstrated superior performance in enhancing signal quality even at low SNRs.

Conclusions:

• A digital adaptive filter can effectively remove CPR artifacts from ECG signals during ongoing CPR.
• This technology has the potential to eliminate the need for hands-off intervals during rhythm analysis.
• Implementing this adaptive filter could lead to improved resuscitation outcomes by ensuring continuous chest compressions.