Application of Entropy Techniques in Analyzing Heart Rate Variabilityusing ECG Signals

Abstract

The variation of the heart rate about a mean value is the Heart Rate Variability (HRV). HRV reflects the functioning of cardio-respiratory control system. It is used as one of the diagnostic measures to detect heart disorders. In the proposed work, HRV analysis using entropy measures is carried out on healthy, Congestive Heart Failure (CHF) and Atrial Fibrillations (AF) subjects using their ECG signals. The entropy methods used in the work are Approximate entropy (ApE), Symbolic entropy (SyE) and Spectral entropy (SpE). ECG signals of 20 healthy subjects in the age group of 21 – 30 years were acquired using dry electrode at a sampling rate of 500 Hz for 10 minutes. Signal processing algorithms for removal of baseline wandering, power line interference and motion artefacts were applied for the raw ECG signal. The ECG signals for CHF and AF subjects in the age group of 30 – 75 years were obtained from the Physionet database. From the analysis it was found that values of ApE and SyE were highest for AF subjects and for SpE, the value was highest for healthy subjects. Further, values of all the three entropies were lowest for CHF subjects. In conclusion, it indicates that the entropy techniques are useful tools in diagnosing patients having heart disorders.