Repository for the paper:
Continuous time recurrent neural networks : overview and application to forecasting blood glucose in the intensive care unit
As we are not data custodians we cannot publicaly share the MIMIC-IV data used. However, it is available to those with credentialed access to physionet.org. Credentialed access can be requested through your physionet account.
We used dbt to connect to the Google Bigquery MIMIC-IV database that can be autopopulated through the physionet interface.
https://docs.getdbt.com/reference/warehouse-setups/bigquery-setup
After setting up dbt run:
cd scripts/data/mimic4glucose
dbt run
Then run the notebook: scripts/data/setup_mimic_data.ipynb
After installing torchctrnn run the notebook: scripts/data/setup_simulations.ipynb
- python 3.9+
- dbt
- db-dtype
- pytorch
- torchctrnn: pip install https://github.com/oizin/torchctrnn/tarball/main
- numba
- properscoring
The full analysis is reproducible as follows:
- Run the bash scripts:
./run_simulation_experiments.sh
./run_mimic_experiments.sh
- Run the evaluation notebooks
- scripts/results/simulation_data_size.ipynb
- scripts/results/simulation_all_5000.ipynb
- scripts/results/mimic_results_from_predictions.ipynb
The following is an example of running a single experiment
python train.py --model=LinearModel --logfolder=results --test --nfolds=1 --seed 1 --data mimic