Trusted Multi-View Deep Learning Classification of Fetal Congenital Heart Disease with Feature-level and Decision-level Fusion
Researchers have developed a novel multi-view deep learning framework designed to classify fetal congenital heart disease (CHD) using echocardiographic images. This system integrates data from multiple angles and employs advanced feature extraction and attention mechanisms to enhance diagnostic accuracy. It also includes a component for uncertainty-based decision-making to manage low-quality images, aiming to provide a reliable tool for early CHD detection. AI
IMPACT This deep learning approach could enhance early detection of congenital heart disease, potentially improving clinical outcomes.