A Neural Conditional Random Field Model Using Deep Features and Learnable Functions for End-to-End MRI Prostate Zonal Segmentation

The official implementation of A Neural Conditional Random Field Model Using Deep Features and Learnable Functions for End-to-End MRI Prostate Zonal Segmentation.

The positional encoding, pairwise potentials encoder, and NCRF are implemented in NCRF_module.py, and the U-Net implementation of NCRF is in NCRF_network.py.

Please feel free to reach out if you have any questions.

Credits

If you used this code or paper in your research, please kindly cite us at

@article{hung2025neural,
  title={A Neural Conditional Random Field Model Using Deep Features and Learnable Functions for End-to-End MRI Prostate Zonal Segmentation},
  author={Hung, Alex Ling Yu and Zhao, Kai and Pang, Kaifeng and Zheng, Haoxin and Du, Xiaoxi and Miao, Qi and Terzopoulos, Demetri and Sung, Kyunghyun and others},
  journal={Machine Learning for Biomedical Imaging},
  volume={3},
  number={August 2025 issue},
  pages={261--286},
  year={2025}
}