This is a official PyTorch implementation of the MedGTX in "Graph-Text Multi-Modal Pre-training for Medical Representation Learning" (Sungjin Park, Seongsu Bae, Jiho Kim, Tackeun Kim, Edward Choi, CHIL 2022) PAPER
We distribute the experimental enviroments via Docker
docker pull sjpark9503/dev_env:chil2022_medgtx_env
- Code for MIMIC-III pre-processing in
preprocessing
- Code for running experiments in
gtx/run_DxPx.py
andgtx/run_Rx.py
- Please log in to W&B for logging results.
- Pre-trained models in
gtx/pretrained_models/Pre/
@inproceedings{park2022medgtx,
title={Graph-Text Multi-Modal Pre-training for Medical Representation Learning},
author={Park, Sungjin and Bae, Seongsu and Kim, Jiho and Kim, Tackeun and Choi, Edward},
booktitle={The Conference on Health, Inference, and Learning},
year={2022}
}