Identification of heart beat abnormality using heart rate and power spectral analysis of ECG

Abstract

Tachycardia and bradycardia are the most common type of heart beat abnormalities. These are caused by disruption of normal heart impulses which makes flow of blood and oxygen to vital body organs too rapid in tachycardia and too slow in bradycardia. This may lead to heart, brain and other organs damage. An attempt has been made in this paper to distinguish the normal heart rhythm from that of irregular ones. Electrocardiogram (ECG) signal processing provides information regarding the functionality of heart and thus helps in detecting cardiac abnormalities. Heart rate analysis of ECG provides the associated heart rate, an indicator of heart beats per minute. However, it does not give any information regarding strength of variations and thus distribution of power. Power spectral analysis using the fast Fourier transform (FFT) is performed to detect and characterize strength of variations in ECG with respect to frequency. The algorithm was tested on normal, bradycardia and tachycardia ECG rhythms loaded from MIT-BIH arrhythmia database of Physiobank ATM. Distinct heart beat abnormalities are found to be correlated with distinct frequency components in respective power spectrum. The inferences drawn from results demonstrate the potential of proposed algorithm that could be developed as an effective computer-aided diagnostic tool to identify heart beat abnormalities in ECGs.