Detection of Abnormal Heart Sounds Using MFCC and Deep Neural Networks

Heart murmurs are abnormal heart sounds that point out the presence of heart disorders such as valve disorders. It is very important to diagnose such anomalies in advance for effective treatment. We propose a deep learning-based system that analyzes different samples of heart murmurs and classifies them as normal or abnormal. The model is trained using recorded sounds from the CirCor DigiScope dataset, which comprise a massive collection of labeled heart murmurs. In order to achieve a precise murmur detection, different signal processing techniques are employed to extract desired features, which is followed by classification using a Deep Neural Network (DNN). The model learns murmur-related patterns and demonstrates its potential to discriminate between normal and abnormal heart sounds. The proposed system provides a reliable and accessible tool for cardiovascular monitoring.