Probabilistic classification approaches for cardiac arrest rhythm interpretation during resuscitation

Our ultimate objective is to develop methodology for resuscitation data analysis that involves monitoring of the patients response, the quality of therapy, and to understand the interplay between therapy and response. To this end, methods to reliably detect the rhythm types during a resuscitation episode are needed. The objective of this study was to develop machine learning algorithms to recognize the rhythms appearing during a resuscitation episode. In this study, we used a probabilistic framework to classify different cardiac arrest rhythms. We propose two different classifiers; naïve Bayes and logistic regression classifier.