VDT2023_MRI2CT: Conversion Between CT and MRI Images Using Diffusion Model

Training and sampling procedure:

First create venv and isntall all requirements

conda create -n [env_name] python=3.9
conda activate [env_name]
cd VDT2023_MRI2CT
pip install -r requirments.txt

Download dataset from here and put it inside .\dataset . Download checkpoints from here and put them into \eperiments\mri2ct_230603_193325\checkpoint\

To start the training process

python sr.py -c config/mri2ct.json 

The gpu_id is set as '0' by default. To infer

python infer.py

Some qualitative results

Acknoledgement

Lastly, I want to thank janspiry for such a wonderful and quality repository. I learned a lot from his/her repo design.