AI-Enhanced Compact Ultrasound for Autonomous Newborn Heart Monitoring and Congenital Disease Detection

The project aims to develop a next generation of echocardiography using autonomous ultrasound (US) scanning and artificial intelligence-based diagnostics of congenital heart diseases (CHD) in neonates, overcoming the limited accuracy of current standard practice using perinatal US imaging and pulse oximetry. We will optimize an US scanner and research framework will be established. Extensive clinical study will collect volumetric echocardiographic data from neonates with and without CHD, which will train artificial intelligence for guided probe placement, optimized scanning sequence, and autonomous CHD identification. Complementary expertise in AI development (Kitware), biomedical instrumentation (Johns Hopkins University), and clinical cardiology, neonatology, and pediatric intensive care (Johns Hopkins Medicine) will bring a unique opportunity to embody a rapid and accurate autonomous echocardiography.