This is the implementation of Radiology Report Generation via Multi-objective Preference Optimization.
- pip install -r requirement.txt
You can download the models we pre-trained model for each dataset from here 提取码：2nq7
you can download the best models for each dateset from here 提取码：52tc
We use two datasets (IU X-Ray and MIMIC-CXR) in our paper.
For IU X-Ray, you can download the dataset from here and place the files in the specified path within scripts/iu-xray/run_rl.sh."
For MIMIC-CXR, you can download the dataset from here and then the specified path in scripts/mimic-cxr/run_rl.sh. You can apply the dataset here with your license of PhysioNet.
Run bash scripts/iu_xray/run_rl.sh to train a model on the IU X-Ray data.
Run bash scripts/mimic-cxr/run_rl.sh to train a model on the MIMIC-CXR data.
You can download the pre-trained model for CheXbert from here: Chexbert. Then place it in the MPO_IU\MPO_TRAIN\CheXbert\checkpoint directory.
For using RadGraph, you can refer to the following link: RadGraph. The specific model checkpoint can be downloaded from here: model_checkpoint. Place the related files in my MPO_IU\MPO_TRAIN\RadGraph directory.
在PMO_TEST
Run bash MPO/MPO_TEST/test_iu_xray.sh to test a model on the IU X-Ray data.
Run bash MPO/MPO_TEST/test_mimic_cxr.sh to test a model on the MIMIC-CXR data.