3D MRI Series Classifier

Official repository for "Classification of Multi-Parametric Body MRI Series Using Deep Learning", published in IEEE Journal of Biomedical and Health Informatics.

[IEEE JBHI paper]

1. Installation

You may need Conda environment. You can download Anaconda from this link. Then, please run:

conda env create –f mriclassifier_env.yml
conda activate mriclassifier

2. Dataset

This code runs using NIFTI (.nii.gz) files. Please put your NIFTI files into ./data. For preprocessing data, please run:

python data_preprocessing.py

Then, the processed data will be saved in ./data/preprocessed.

• If you use our pre-trained model, please preprocess the data using the above code, which resamples the 3D MRI volumes.

3-1. Train

To train our model, run the following command:

sh train.sh

The checkpoints of the model will be saved in ./checkpoint/.

3-2. Test

To test the trained our model, run:

sh test.sh

The results of the MRI classification such as the confusion matrix will be saved in ./results.

• Here, we have provided our pre-trained model which was trained on the Siemens scanner data and then finetuned on the Philips scanner data (Please see Strategy 2 in Table 2 of the paper).
• To use our pre-trained model, please download the model weights from here and put the weights in ./pretrained_model.

Citation

@article{kim2024classification,
  title={Classification of Multi-Parametric Body MRI Series Using Deep Learning},
  author={Kim, Boah and Mathai, Tejas Sudharshan and Helm, Kimberly and Pinto, Peter A and Summers, Ronald M},
  journal={IEEE Journal of Biomedical and Health Informatics},
  year={2024},
  publisher={IEEE}
}