Automatic heart and lung sounds classification using convolutional neural networks

We study the effectiveness of using convolutional neural networks (CNNs) to automatically detect abnormal heart and lung sounds and classify them into different classes in this paper. Heart and respiratory diseases have been affecting humankind for a long time. An effective and automatic diagnostic method is highly attractive since it can help discover potential threat at the early stage, even at home without a professional doctor. We collected a data set containing normal and abnormal heart and lung sounds. These sounds were then annotated by professional doctors. CNNs based systems were implemented to automatically classify the heart sounds into one of the seven categories: normal, bruit de galop, mitral inadequacy, mitral stenosis, interventricular septal defect (IVSD), aortic incompetence, aorta stenosis, and the lung sounds into one of the three categories: normal, moist rales, wheezing rale.