CREATIS, University of Lyon (Leclerc et al.) / University of Sherbrooke (vitalab pretrained models)
The original U-Net baseline segmentation network introduced alongside the CAMUS (Cardiac Acquisitions for Multi-structure Ultrasound Segmentation) dataset, one of the largest fully open-access, expert-annotated 2D echocardiography benchmarks. The network segments the left ventricle endocardium (LVEndo), left ventricle epicardium/myocardium (LVEpi), and left atrium (LA) from apical 2-chamber and 4-chamber echo views at end-diastole (ED) and end-systole (ES). In the original ten-fold cross-validation benchmark comparing U-Net, U-Net++, Stacked Hourglass, Anatomically Constrained Neural Networks, and classical methods, the U-Net variant (18M parameters) achieved the best overall accuracy, reaching Dice scores of 0.939 (ED) / 0.916 (ES) for LVEndo and 0.954 (ED) / 0.945 (ES) for LVEpi, approaching inter-observer variability. A pretrained checkpoint of this baseline U-Net is distributed via the University of Sherbrooke's vitalab CASTOR project as part of a broader library for building anatomically-constrained cardiac segmentation pipelines.
Architecture
CNN (2D)
U-Net (2D CNN encoder-decoder) for multi-class semantic segmentation of the left ventricle endocardium, epicardium/myocardium, and left atrium from single 2D echocardiography frames
Framework
PyTorch
Parameters
18,000,000
Reported for the best-performing U-Net variant ("U-Net 1") in the original CAMUS benchmark paper
Added to catalog
2026-08-12
Apache 2.0
License for model weights only. Associated code may be licensed seperately, check code source for specific terms.
500 patients imaged with 2D transthoracic echocardiography (A2C/A4C views) at University Hospital of St Etienne; roughly half with LVEF < 45%.
Multi-class segmentation of the left ventricle endocardium, left ventricle epicardium/myocardium, and left atrium from 2-chamber and 4-chamber apical echocardiography frames at end-diastole and end-systole