Automatic analysis of ECG signals based on their fractal and multifractal properties

Abstract

Automatic analysis of electrocardiographic (ECG) signals, including the heart rate variability (HRV), makes it possible to assess the health status of patients, by reducing the likelihood of human error and providing an optimal and relatively accurate result. HRV is an important information indicator for diagnosing and predicting cardiovascular disease, which is based on measuring the intervals between heartbeats (known as RR-time intervals) derived from ECG signals. HRV has been used for research in various scientific fields, including information technology where it is used to create software products for automatic analysis of ECG signals in order to gain additional knowledge about the behavior of RR fluctuations in normal and disease states. ECG signals are non-stationary and one of the most suitable methods for analysis are the fractal methods. This article presents the results of the study of the fractal and multifractal properties of real ECG signals, combined into two groups: ECG signals of healthy subjects as well as patients with cardiovascular disease (arrhythmia), using the methods: Rescaled range (R/S) analysis and Multifractal Detrended Fluctuation Analysis (MFDFA). The obtained values of the studied parameters are used to distinguish healthy subjects from sick ones, by applying statistical analysis. The statistical analysis is performed by applying a t-test to determine the statistical significance of the studied ECG signals and Receiver Operating Characteristic (ROC) analysis to assess the quality of the selected methods. The obtained results show that the fractal methods used are suitable for analysis of the dynamics of RR intervals and for distinguishing the healthy subjects from those with pathological diseases.