janghana/synthetic-bile-duct

AI Technology Development for Bile Duct Imaging Synthesis Based on CT

AI Technology Development for Bile Duct Imaging Synthesis Based on CT

Figure

Adversarial Diffusion Process

Adversarial Diffusion Projector

CT Modality Synthesis Process

MRT1 Modality Synthesis Process

Abstract

Medical imaging plays a crucial role in enhancing the accuracy of diagnosis. In particular, some tissues can be more accurately identified through multiple imaging modalities. If an image obtained from one imaging technique can be transformed into another modality, it would lead to a more efficient diagnostic process in the medical field. In this study, we conducted research on synthesizing the bile duct using CT scans into MRI-T1 images. Due to differences in imaging techniques, the bile duct is less distinguishable in CT scans but appears more prominently in MRI-T1 images. By obtaining MRI-T1 image information solely from CT scans, we anticipate time and cost savings in the medical field. This research highlights the limitations of prior neural network-based studies on the translation from CT to MRI and proposes a solution using a diffusion model.

Train

python3 train.py --image_height 400 --image_width 480 --exp exp_syndiff --num_channels 2 --num_channels_dae 64 --ch_mult 1 1 2 2 4 --num_timesteps 4 --num_res_blocks 2 --batch_size 1 --contrast1 CT --contrast2 MRT1 --num_epoch 100 --ngf 64 --embedding_type positional --use_ema --ema_decay 0.999 --r1_gamma 1. --z_emb_dim 256 --lr_d 1e-4 --lr_g 1.6e-4 --lazy_reg 10 --num_process_per_node 1 --save_content --local_rank 0 --input_path ../dataset/train --output_path ./output/for/results --port_num 5000

Test

python3 test.py --image_height 400 --image_width 480 --exp exp_syndiff --num_channels 2 --num_channels_dae 64 --ch_mult 1 1 2 2 4 --num_timesteps 4 --num_res_blocks 2 --batch_size 1 --embedding_type positional  --z_emb_dim 256 --contrast1 CT --contrast2 MRT1 --which_epoch 100 --gpu_chose 1 --input_path ../dataset/test --output_path /output/for/results

Citation

@misc{özbey2023unsupervised,
      title={Unsupervised Medical Image Translation with Adversarial Diffusion Models}, 
      author={Muzaffer Özbey and Onat Dalmaz and Salman UH Dar and Hasan A Bedel and Şaban Özturk and Alper Güngör and Tolga Çukur},
      year={2023},
      eprint={2207.08208},
      archivePrefix={arXiv},
      primaryClass={eess.IV}
}

Author

Han Jang - Korea Institute of Science and Technology, Seoul, South Korea.

Reference

• https://github.com/icon-lab/SynDiff
• https://arxiv.org/abs/2207.08208