Universidade Federal de Minas Gerais (UFMG) / Children's National Hospital (Martins et al.)
Training code and model weights not public. Contact creators for more information.
Video-based deep learning system for automated screening of rheumatic heart disease (RHD) from echocardiography, aimed at low-resource settings where RHD -- the most common acquired heart disease in children and young adults worldwide -- is endemic but echocardiography expertise is scarce. A 3D convolutional neural network (C3D) classifies each echo video clip, explicitly modeling temporal information across frames; a supervised meta-classifier then aggregates the per-video predictions from an exam (which may contain many video clips from different views) into a single exam-level RHD diagnosis. Evaluated on 11,646 echocardiography videos from 912 screening exams collected in underserved areas of Brazil and Uganda, the 3D C3D network significantly outperformed a comparison 2D CNN (VGG16) that ignores temporal information, and the learned aggregation model reached 72.77% exam-level diagnostic accuracy, exceeding simple majority voting across a patient's videos.
Architecture
CNN (3D)
C3D (3D convolutional neural network) for per-clip video classification, with a separate supervised meta-classifier aggregating per-video predictions from a multi-video echocardiography exam into a single diagnosis
Added to catalog
2026-08-12
PROVAR/ATMOSPHERE Rheumatic Heart Disease Echocardiography Screening Cohort
11,646 echocardiography videos from 912 screening exams collected during rheumatic heart disease (RHD) screening campaigns in underdeveloped areas of Brazil and Uganda, as part of the PROVAR and ATMOSPHERE studies.
Exam-level binary diagnosis of rheumatic heart disease from a set of echocardiography video clips