Design And Development of A Signal Processing Method For The Detection of Murmurs Within Heart Sounds And The Characterization of Its Phenotypes

Abstract

Ailment of the cardiovascular valves and other cardiovascular systems often brings about unusual, turbulent blood flow inside the heart, initiating murmur. The presence of this forged components, within the Phonocardiography records, or heard through the careful routine auscultation with stethoscope, may be a valuable tool in the diagnosis of many valvular disorders. But it is difficult to identify the subtle components through manual auscultation procedure. One of the most used procedures for non-invasive cardio vascular diagnosis is automated signal analysis. newlineThe feature extraction process is one of the first steps in every artificial intelligence system. For systems designed for automated analysis, signal processing techniques used for extracting characteristics play a crucial role. Problems can be discovered, properly traced, and their nature can be diagnosed by features retrieved using relevant signal processing techniques. newlineHowever, the present approaches fall short of characterizing the murmur into four distinct classifications based on the properties of the heart signal, such as normal heart sound, systolic murmur, diastolic murmur, and continuous murmur. The majority of the approaches are designed for murmur detection rather than accurate type identification. Another significant issue with the available schemes is that they use complex classifiers, most of which are based on Neural Networks (NNs) or evolutionary computing, that use a large number of features without screening them for statistical significance, limiting their computational feasibility and embedded realization scope. newline newline