Material from Ciucci 2024 - Material from Ciucci 2024

This repository contains the source code to reproduce the modeling results from the paper

Paper here

Reproducible results¶

You can find a report of the results by going to the repository homepage at https://computationalphysiology.github.io/ciucci-2024/myst/report-1.

Note that the results displayed here are re-generated every time a push is made to the repository (see https://github.com/ComputationalPhysiology/ciucci-2024/actions/workflows/deploy_docs.yml). Note that all figures are also uploaded as artifacts during the build.

Also note that the results from the paper are run on a finer mesh than those displayed in the documentation.

Install dependencies¶

To run the code in this repo you need FEniCS and gmsh with OpenCascade installed. The easiest way to do this is to either use the provided docker image, i.e

docker pull ghcr.io/computationalphysiology/ciucci-2024:latest

You can create a new container using the command

docker run --name ciucci -w /home/shared -v $PWD:/home/shared -it ghcr.io/computationalphysiology/ciucci-2024:latest

which will also share your current directory with the container. If you are interested you can also check out the Dockerfile if you want to know how the image was created.

Running scripts¶

There are three types of simulation scripts.

A contracting cylinder, with only one time point
A contraction cylinder with a full twitch
An unloaded (zero pressure) and loaded (pressure = 15 kPA) contracting left ventricle.

We refer to the documentation for more information about the models and numerical experiments.

The main script for running all the commands are located inside the code directory and you should first make sure to navigate to this directory, i.e

cd code

You can list all the available options by typing

python3 main.py --help

Here is an example of how it looks inside the container

root@b2637926d9de:/home/shared/code# python3 main.py --help
usage: main.py [-h] [--dry-run] {preprocess-lv,preprocess-cylinder,run-lv,run-cylinder,run-cylinder-twitch,postprocess-lv,postprocess-cylinder,postprocess-cylinder-twitch} ...

positional arguments:
  {preprocess-lv,preprocess-cylinder,run-lv,run-cylinder,run-cylinder-twitch,postprocess-lv,postprocess-cylinder,postprocess-cylinder-twitch}
    preprocess-lv       Create left ventricle mesh
    preprocess-cylinder
                        Create cylinder mesh
    run-lv              Run simulations with left ventricle model
    run-cylinder        Run simulations with cylinder model
    run-cylinder-twitch
                        Run simulations with cylinder model
    postprocess-lv      Postprocess LV results
    postprocess-cylinder
                        Postprocess cylinder results
    postprocess-cylinder-twitch
                        Postprocess cylinder results

options:
  -h, --help            show this help message and exit
  --dry-run             Just print the command and do not run it

Below we show the basic command. Note that you can also use the flag --help to see all the different options for each command, e.g

root@b2637926d9de:/home/shared/code# python3 main.py preprocess-cylinder --help
usage: main.py preprocess-cylinder [-h] [-o MESH_FOLDER] [-c CHAR_LENGTH] [-l LENGTH] [-r RADIUS]

options:
  -h, --help            show this help message and exit
  -o MESH_FOLDER, --mesh-folder MESH_FOLDER
  -c CHAR_LENGTH, --char_length CHAR_LENGTH
  -l LENGTH, --length LENGTH
  -r RADIUS, --radius RADIUS

Cylinder¶

Preprocessing¶

python3 main.py preprocess-cylinder -o meshes/cylinder

Simulation¶

python3 main.py run-cylinder -i meshes/cylinder -o results/cylinder

Simulation (twitch)¶

python3 main.py run-cylinder-twitch -i meshes/cylinder -o results/cylinder-twitch

Postprocessing¶

python3 main.py postprocess-cylinder -i meshes/cylinder -r results/cylinder -o figures/cylinder

Postprocessing (twitch)¶

python3 main.py postprocess-cylinder-twitch -i meshes/cylinder -r results/cylinder-twitch -o figures/cylinder-twitch

LV¶

Preprocessing¶

Native¶

python3 main.py preprocess-lv -o meshes/native -c native

Transplanted¶

python3 main.py preprocess-lv -o meshes/transplanted -c transplanted

Running simulations¶

Native¶

python3 main.py run-lv -i meshes/native -c native -o results/native

When running the simulations the following files will be created in the output folder

native
├── gammas.npy
├── pressures.npy
├── results_current.h5
├── results_current.xdmf
├── results_current_smooth.h5
├── results_current_smooth.xdmf
├── results_reference.h5
├── results_reference.xdmf
├── results_reference_smooth.h5
├── results_reference_smooth.xdmf
└── volumes.npy

The xdmf files are files that can be opened in Paraview. The label reference refers to the solutions are stored on the reference mesh, while in the file with the label current the mesh has been updated to the current configuration. For the results with the label smooth we have smoothed the results using a Gaussian filter and interpolated into a first order Lagrange element function space. The files gammas.npy and pressures.npy contain the values of the gamma and pressure for each time step. The file volumes.npy contains the volume of the LV at each time step.

Transplanted¶

python3 main.py run-lv -i meshes/transplanted -c transplanted -o results/transplanted

Postprocessing¶

Native¶

python3 main.py postprocess-lv -i meshes/native -r results/native -o figures/native

Transplanted¶

python3 main.py postprocess-lv -i meshes/transplanted -r results/transplanted -o figures/transplanted

Get stats using

python3 main.py postprocess-lv -i meshes/transplanted -r results/transplanted -o figures/transplanted --print-stats

Comparison for end-systole¶

python3 main.py postprocess-lv-ES -n results/native -t results/transplanted -o figures

Citation¶

If you use the code in this repository, please cite: TBW

Author¶

Henrik Finsberg and Sam Wall

License¶

MIT

Report

Simulating mechanical forces