Assessment of Valve Regurgitation Severity via Contrastive Learning and Multi-view Video Integration
- Under construction will be updated soon.
This repo contains the code for our paper Assessment of Valve Regurgitation Severity via Contrastive Learning and Multi-view Video Integration .
Acquisition of various color Doppler imaging in standard echocardiographic views from varying probe positions.
Overall architecture of the proposed method. A supervised contrastive learning to learn representations using a contrastive loss but uses label information to sample positives and negative pairs. A feature extractor which employs a pre-trained encoder shared across all input video. The optimization step involves the use of inter-intra contrastive loss in conjunction with classification loss.
- Envrionment Setting
pip install -r requirements.py
If you found our paper is contributed to your research, please consider starring โญ us on GitHub and citing ๐ us in your research!