📄 Paper: SynMSE: A multimodal similarity evaluator for complex distribution discrepancy in unsupervised deformable multimodal medical image registration
✍️ Authors: [Jingke Zhu] et al.
📍 Published in: Medical Image Analysis, 2025
🧱 Link: https://www.sciencedirect.com/science/article/pii/S1361841525001677
- You need to pre-train a CycleGAN model for distribution transfer and replace the path with your own generative model in trainer/cyclegan_syn.
- Start visdom：
python -m visdom.server -p 8097
If other port parameters are used, you need to modify the port in yaml.
- Train:
python main.py
You can train the evaluator, train the registration model, and test the registration model separately by modifying the code in main.py.