Stimuli Effect Identification

Package to identify stimuli model from clinical data. It is intended to be use as a plugin for the Python Anesthesia Simulator to calibrate the model on a given dataset. In this repository the identification procedure is applied to as subset of the VitalDB dataset.

Usage

Install Environement

with uv:

uv sync

With pip:

python -m venv your_environment
source your_environment/bin/activate
pip install .

Exemple files

Create database from VitalDB

• example/select_cases.py Select the caseid to download
• example/create_database.py Download the data, perform the simulations with raw model and save the data

Identification of the stimuli and noise

• example/identify_bis_dist.py Compute the median error over the intubation period and approximate the results with a polinomial function of time.
• example/identify_stimuli.py Run the identification prdecure as describe in the paper.
• example/identify_noise.ipynb Identify noise for MAP, HR, and BIS over a steday state period of clinical data.

Compare clinical data and model output

• example/bis_model_comparison.ipynb Compare all bis model with clinical data.
• example/raw_model_vs_data.ipynb Compare raw model with clinical data at intubation and surgery time.
• example/compare_stimuli.ipynb Compare Stimuli model versus Non-stimuli model versus Clinical data.