EARLY PREDICTION OF SUDDEN CARDIAC DEATH USING FRACTAL DIMENSION AND ECG SIGNALS

Sudden cardiac death (SCD) is deemed as one of the main causes of death in humans. Therefore, the prediction of an SCD event will help people to receive timely treatment, allowing saving their life. In this sense, this paper introduces a new methodology based on the adroit fusion of fractal dimension (FD) algorithms and a fuzzy logic system for predicting an SCD event automatically. Five FD implementations are investigated in this work: Katz's FD, Sevcik's FD, Box's FD, Higuchi's FD, and Petrosian's FD, in order to evaluate the geometrical complexity in electrocardiogram signals of 20 patients with SCD and 18 patients with a normal cardiac rhythm offered by Boston's Beth Israel Hospital. The results indicate that the FD-based methodology can predict an SCD event up to 60min before the onset, reaching an accuracy of 91.54%.

Automated summary

This paper introduces a new methodology for automatically predicting sudden cardiac death (SCD) events by combining fractal dimension (FD) algorithms and a fuzzy logic system. By evaluating the geometrical complexity of electrocardiogram signals, the FD-based methodology can predict an SCD event up to 60 minutes before onset with an accuracy of 91.54%.