Skip to content

Bilkent-CYBORG/ESCADA

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

3 Commits
 
 
 
 
 
 
 
 
 
 

Repository files navigation

ESCADA: Efficient Safe and Context Aware Dose Allocation for Precision Medicine

Author: Ilker Demirel (ilkerd1997@gmail.com)

The repository for the manuscript "ESCADA: Efficient Safe and Context Aware Dose Allocation for Precision Medicine", (NeurIPS 2022).

Requirements

pip install -r requirements.txt

UVA/PADOVA T1DM simulator

We use the open source implementation of the UVa/PADOVA simulator, which is available under MIT License at simglucose. We obtained the simulator outputs for different patient, meal event, insulin intake triples used in the experiments and saved the results in the following directories for faster reproduction,

/experiments/calc_res/*
/experiments/calc_res_clinician_data/*

Reproduction

Our experimental results are classified under three main categories: single meal event (SME) scenario, multiple meal event (MME) scenario, and comparison against a clinician. We have separate __.py files for each setting, and the experiments can be run for each scenario by running the corresponding scripts in the following directories,

/experiments/SME/SME.py
/experiments/MME/MME.py
/experiments/clinician_comparison/clinician_comparison.py

Plots

The necessary data to obtain the plots and the numerical results in the paper is already available in the following directories,

/experiments/SME/ppbg
/experiments/SME/ppbg_tc
/experiments/MME/ppbg
/experiments/clinician_comparison/test_res

You can run the experiments as described before to replace these results. Finally, to obtain the plots and the numerical results in the table, use the following notebook,

/experiments/plot.ipynb

About

No description, website, or topics provided.

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published