FemurSurfacePrediction

This is the implementation of MICCAI 2024 paper "Prediction of Disease-Related Femur Shape Changes Using Geometric Encoding and Clinical Context on a Hip Disease CT Database."

Dataset

After correspondence construction using GBCPD, create a tabular file with the following format as in src/data/template_data.csv.

Training

python position_inr.py --batch_size ${batch_size} --n_iteration ${num_iterations} --lr ${lr} \
--inr_model ${inr_model} --exp_name ${exp_name} --use_cosine_similarity --alpha 0. --beta 0. --data_root ${data_root} \
--domain_name ${domain_name} --info_root ${info_root} --gamma 1. --delta 1. --grading_diff --fold ${fold}

or run the shell script "run_train.sh" with your parameters.

Evaluation

python simple_position_evaluation.py

Citation

@InProceedings{Li_Prediction_MICCAI2024,
        author = { Li, Ganping and Otake, Yoshito and Soufi, Mazen and Masuda, Masachika and Uemura, Keisuke and Takao, Masaki and Sugano, Nobuhiko and Sato, Yoshinobu},
        title = { { Prediction of Disease-Related Femur Shape Changes Using Geometric Encoding and Clinical Context on a Hip Disease CT Database } },
        booktitle = {proceedings of Medical Image Computing and Computer Assisted Intervention -- MICCAI 2024},
        year = {2024},
        publisher = {Springer Nature Switzerland},
        volume = {LNCS 15003},
        month = {October},
        page = {368 -- 378}
}