A review of diagnosis methods for heart rhythm disorders

The human cardiovascular system exhibits complicated dynamics. Various signal analysis techniques have been utilized to study electrocardiogram (ECG) rate and morphology for their potentials in diagnosis and prediction of cardiovascular diseases. Conventional methods of linear signal theory, including power spectral analysis and time-domain analysis, have been widely adopted to analyze heart rhythm disorders due to cardiomyopathy and fibrillation. However, the success of these techniques is often hampered by the nonstationary and nonlinear character of the heart's activities. To overcome these limitations, many techniques based on nonlinear dynamics and chaos theory have been developed. In this work, we investigate various existing techniques for their potential application in the diagnosis of rhythm disorders. This comparison of different methodologies can be used as a guideline in selecting proper ECG signal analysis techniques.