This is a re-implementation of Med2Vec [1] in Pytorch. It simply embeds clinical concepts into a distributed representation using skip-gram model with an aditional code loss.
To run the code first obtain the ADMISSION.CSV and DIAGNOSES_ICD.CSV from MIMIC-III database here.
Compile the data by running bash gen_data.sh make sure you set the correct paths to the files.
The directories are structured as follows:
./base: base trainer, data loader../configs: json files for experiments. This where you pass in arguments to the model/trainer and what have you../trainer: contains training logic, and anything that must be done to train the model../model: directory containing the med2vec model.
To train the model run the following:
python train_med2vec.py -c ./configs/config.json
note: make sure the directories are set appropiatly in ./configs/config.json.
[1] Choi, Edward, et al. "Multi-layer representation learning for medical concepts." Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining. ACM, 2016.