Tunable-Q wavelets transform based methodologies for analysis and classification of cardiac signals

Abstract

After cancer, heart disorders are the second major cause of mortality and morbidity around the globe. The heart valve disorders, septal defects and coronary artery disease are the most commonly occurring heart disorders. Timely diagnosis of heart disorders is generally required for prevention and treatment of these disorders to ensure contented, happier and longer life of patients. The cardiac auscultation and electrocardiogram (ECG) are the important means of assessing the activity of cardiovascular system. These procedures are commonly used for reliable diagnosis of heart disorders. Heart disorders especially heart valves cause changes or additional sounds to normal heart sounds that can be useful for diagnosis. These heart sounds can be analysed non-invasively using traditional cardiac auscultation with conventional stethoscope. However, analysing these heart sounds by listening, requires sophisticated interpretive skills and expertise in diagnosis. Moreover, the heart sounds often last for a short period of time and pathological splitting of the heart sound is di cult to judge because human ears lack desired sensitivity towards heart sounds and murmurs. The cardiac sound signals represent digital recording of the heart sounds by placing an electronic stethoscope at the appropriate location on the subject's chest. These signals can be used to extract valuable diagnostic features for diagnosis of the heart valve disorders. Electrocardiography is also a non-invasive measure of the electrical activity of the heart against time. It records electrical potentials of the contractile heart cells by placing electrodes on the surface of the chest and on the limb. Generally, electrocardiography involves recordingof ECG waveform onto a graph paper that runs at a constant speed or visual display on a screen. ECG waveform analysis is carried out by evaluating the morphological changes in shape, amplitude, period, segments, and intervals. The subtle changes in these features of ECG waveforms cannot be deciphered precisely on visual inspection. Moreover, the clinical interpretations of ECG waveform are based on observation or experimental knowledge. On the other hand, the digitally recorded ECG signals can provide valuable diagnostic features for automatic diagnosis of the CAD.