A Visual Domain Transfer Learning Approach for Heartbeat Sound Classification
This work addresses early detection of heart disease, a major global health issue, by improving automated classification of heart sounds, though it is incremental as it adapts existing visual methods to a new domain.
The paper tackled automated heart sound classification by converting sounds to mel spectrograms and applying visual domain transfer learning with CNNs like ResNet and MobileNetV2, achieving ~90% categorical accuracy and ~0.97 AUROC on two datasets from the PASCAL collection.
Heart disease is the most common reason for human mortality that causes almost one-third of deaths throughout the world. Detecting the disease early increases the chances of survival of the patient and there are several ways a sign of heart disease can be detected early. This research proposes to convert cleansed and normalized heart sound into visual mel scale spectrograms and then using visual domain transfer learning approaches to automatically extract features and categorize between heart sounds. Some of the previous studies found that the spectrogram of various types of heart sounds is visually distinguishable to human eyes, which motivated this study to experiment on visual domain classification approaches for automated heart sound classification. It will use convolution neural network-based architectures i.e. ResNet, MobileNetV2, etc as the automated feature extractors from spectrograms. These well-accepted models in the image domain showed to learn generalized feature representations of cardiac sounds collected from different environments with varying amplitude and noise levels. Model evaluation criteria used were categorical accuracy, precision, recall, and AUROC as the chosen dataset is unbalanced. The proposed approach has been implemented on datasets A and B of the PASCAL heart sound collection and resulted in ~ 90% categorical accuracy and AUROC of ~0.97 for both sets.