Classification of Experimental Electrograms

Abstract

The aim of this work was to analyse and classify the experimental electrograms (EG) recording from isolated rabbit hearts during experiments with short repeated global ischemia. The morphological parameters calculated from EG were used as classification features. The discrimination ability of the features was verified and the most relevant features were then chosen to cluster the EGs into three groups (physiological, ischemic, and extrasystols) using k-means method.