AI-empowered System for Diagnosis of Congenital Heart Disease in Infants

Among birth defect diseases in infants, congenital heart disease has the highest incidence rate. The diagnosis of this disease mainly relies on echocardiography data. However, diagnosing congenital heart disease in infants remains highly challenging due to poor ultrasound data quality and heavy reliance on doctors' skills and clinical experience. This project aims to construct, for the first time, a database of echocardiograms for over 500 patients with infantile congenital heart disease worldwide. Based on this database, the project aims to develop a novel AI-powered diagnostic system for infantile cardiac defects. This system integrates three main sub-modules: an infant ultrasound data standard section acquisition sub-module, a detection and segmentation sub-module for key tissues, and a disease classification and quantitative analysis sub-module. We believe that the system cannot only promote the research on AI-empowered medical image analysis in mainland China and Hong Kong but also has great commercial potential and social impacts.