NONLINEAR ANALYSIS OF HEART RATE VARIABILITY IN CARDIOMYOPATHY AND HEALTHY SUBJECT

Abstract

An electrocardiogram (ECG) is a graphic recording of the electrical activities in the human heart that can provides significant information for medical diagnosis. In this study, nonlinear time series analysis (NTSA) was used to investigate the dynamical information and complexity of healthy, normal, hypertensive subjects, heart failure, Heart Problem (Asthma) and cardiomyopathy ECG signals. Four nonlinear parameters: Correlation Dimension, Approximate Entropy, Sample Entropy and Recurrence Plot were used to extract the dynamical information in the ECG signals from the fabricated device. The results for Correlation Dimension of healthy, heart failure, heart problem (Asthma), hypertensive subject A and hypertensive subject B subjects are 5.6238, 1.0922, 2.9337, 1.0157 and 0.9476 for the fabricated ECG device and 2.1489 for healthy, 2.0218 for cardiomyopathy for the physionet data. The nonlinear measures considered in this work showed capability to differentiate between the healthy and defective/infected hearts. The correlation dimension revealed that the heart rate variability possesses fractal dimension properties for both subject while the healthy control subject shows high value of fractal dimension than the defective heart diseases subjects. The Approximate and Sample Entropies show that for healthy control ECG signals the degree of complexity measured is higher than the defective heart diseases subject, indicating that for diseased heart beat, complexity measure is reduced. The recurrence plot revealed more deterministic structure for defective heart compared to the Healthy subjects which indicate possibilities of predictability in the heart rate variability for Heart diseases signals. The fabricated ECG device record the electrical activity of the heart which is displayed digitally.