Detecting abnormal heart sound using mobile phones and on-device IConNet

Given the global prevalence of cardiovascular diseases, there is a The cardiovascular disease screening process detects abnormalities pressing need for easily accessible early screening methods. Typically, such as heart murmur, which is an irregular sound audible during this requires medical practitioners to investigate heart auscultations the heartbeat cycle through a stethoscope. Detection of a heart for irregular sounds, followed by echocardiography and electrocardiography murmur suggests underlying cardiac issues, prompting further evaluation tests. To democratize early diagnosis, we present a through echocardiography and electrocardiography tests user-friendly solution for abnormal heart sound detection, utilizing to pinpoint the specific heart disease. To enhance the accessibility mobile phones and a lightweight neural network optimized for of early diagnosis, we introduce a novel system for detecting abnormal on-device inference. Unlike previous approaches reliant on specialized heart sounds using mobile phones and an on-device neural stethoscopes, our method directly analyzes audio recordings, network. Our system does not require extra equipment, a server, facilitated by a novel architecture known as IConNet. IConNet, an or a specific data preprocessing pipeline, which is an advantage Interpretable Convolutional Neural Network, harnesses insights compared to existing works.