Abstract 96: Real-Time Electrogram Characteristics During Human Ventricular Fibrillation for Quantification of the Antegrade Coronary Flow to Optimize Resuscitation Outcomes
Background: In patients with out of hospital cardiac arrest due to VF, CPR prior to defibrillation may improve survival. If CPR efficacy measured as antegrade coronary flow could be related to VF signal characteristics, real-time assessment of these characteristics could serve to assist EMS personnel in optimizing chest compressions and timing of defibrillation.
Objective: To derive wavelet based ECG markers during VF in human hearts to quantify the relation between antegrade coronary flow and ECG characteristics of VF.
Method: VF experiments were conducted in 5 isolated cardiomyopathic human hearts using a Langendorff setup. After the initiation of VF, the hearts were maintained in ischemia (no flow) for 3 min, followed by a 2 min reperfusion and defibrillation. The experiments were repeated at flows of 0%, 30%, and 100% of baseline perfusion and pseudo surface electrograms were recorded using plate electrodes. The electrograms were analysed using continuous wavelet transform (CWT) in 5s frames. From the CWT decomposition coefficients, Dominant Wavelet Scale (CENTER) and Scale Distribution Width (SDW) were extracted as features.
Results: We analyzed the temporal evolution of the wavelet features over ischemia and reperfusion periods. The slope of the temporal evolution of the features reversed after the reperfusion was started for 30% and 100% flows while it remained the same for the 0% flow. In order to quantify the deviation of the features for different flow rates in such a way that it is independent of the heart and time effects, we arrived at a relative quantity in the form of the angle difference between line-fitted ischemia-reperfusion slopes (phi) and the reperfusion slopes (theta) with a horizontal reference. We performed ANOVA using generalized linear model (for angel deviation versus flow rates) and observed significance levels of p= 0.0237 SDW, phi, p= 0.006 SDW, theta, p=0.0345 CENTER, phi, and p=0.0345 CENTER, theta for the wavelet features. The Bonferroni tests indicated significant differences between the flow rates.
Conclusions: Wavelet based features demonstrate significant separation between different flow rates thus could potentially be used to quantify and provide near real-time feedback of chest compression efficacy during CPR.
- © 2010 by American Heart Association, Inc.