closed_loop_fontan

Closed-loop Fontan circulation models and calibration/reference data workflows.

The repository is organized for three model families:

Model family	Status	Description
models/full_0d	active	Current full 0-D closed-loop PhysioBlocks model.
models/quasi_0d_1d	active	Calibrated PhysioBlocks-only quasi 0-D/1-D model with R-L-C aortic and Fontan chains.
models/coupled_0d_1d	planned	Future coupled 0-D/1-D model with the aorta and TCPC represented by 1-D models.

Each model family owns its README, schematic, configs, implementation notes, and standardized technical reference PDF/source. Model changes must keep those artifacts synchronized.

Setup

python -m venv .venv
. .venv/bin/activate
python -m pip install -r requirements.txt

Current Full 0-D Model

Run the smoke case:

python scripts/run_one.py models/full_0d/configs/fontan_0d_smoke.jsonc --series Smoke

Compute metrics from a generated run:

python scripts/metrics.py runs/simulations/Smoke/*/main.csv models/full_0d/configs/fontan_0d_smoke.jsonc --out models/full_0d/reference_outputs/smoke_metrics.json

See models/full_0d/README.md for the topology, scenarios, and model caveats.

Current Quasi 0-D/1-D Model

Run the smoke case:

.venv/bin/python scripts/run_one.py models/quasi_0d_1d/configs/fontan_quasi_smoke.jsonc --series QuasiSmoke

See models/quasi_0d_1d/README.md for the implemented quasi-chain topology, config generator, metrics workflow, scenarios, and model caveats.

Aramburu 2024 Data

The raw Aramburu_et_al_2024_Heliyon_e30404.zip archive is kept locally under data/raw/ and is not committed. Standardized calibration/reference outputs are tracked under data/processed/aramburu_2024/.

Regenerate the processed data:

python scripts/data/prepare_aramburu_2024.py
python scripts/calibration/extract_targets.py

The processed package includes canonical measurement tables, paper-result tables, model-input tables, converted Nektar 1-D .mat outputs, .fo outputs, shared calibration targets, metadata, and source checksums.

Tests

pytest -q

License

Repository code is MIT licensed. Third-party data derived from Aramburu et al. 2024 is provided with provenance metadata and is not relicensed as MIT.