AugmentA: Patient-specific Augmented Atrial model Generation Tool

We propose a patient-specific Augmented Atrial model Generation Tool (AugmentA) as a highly automated framework which, starting from clinical geometrical data, provides ready-to-use atrial personalized computational models. AugmentA consists firstly of a pre-processing step applied to the input geometry. Secondly, the atrial orifices are identified and labelled using only one reference point per atrium. If the workflow includes fitting a statistical shape model (SSM) to the input geometry, this is first rigidly aligned with the given mean shape and finally a non-rigid fitting procedure is applied. AugmentA automatically generates the fiber orientation using a Laplace-Dirichlet-Rule-based-Method.

Files and Folders

• main.py: AugmentA's main script
• mesh/: contains the exemplary mesh and the statistical shape model
• standalones/: standalone tools used in the pipeline
• template/: template for non-rigid fitting process
• Atrial_LDRBM/: Laplace-Dirichlet-Rule-based-Method to annotate anatomical regions and generate atrial fiber orientation in the atria

Setup

Create a python virtual environment to install AugmentA dependencies: python -m venv ~/myEnv source ~/myEnv/bin/activate pip install -r requirements.txt

• Install PyMesh
• Install meshtool
• Install carputils

Go to the carputils folder and re-install carputils' requirements (assuming that carputils was installed in the home folder): cd ~/carputils pip install -r requirements.txt

Usage

Remember to source to myEnv before using the pipeline: source ~/myEnv/bin/activate Show all options: python main.py --help Example using an MRI segmentation to produce a bilayer atrial model: python main.py --mesh mesh/LA_MRI.vtp --closed_surface 0 --use_curvature_to_open 1 --atrium LA --open_orifices 1 --MRI 1 Example opening the atrial orifices using the surface curvature to identify the veins of a closed geometry, it expects the valve region to be tagged on the atrial surface with a value > 0.5 (see the scalar "valve" in mesh/LAEAM.vtp): ``` python main.py --mesh mesh/LAEAM.vtp --openorifices 1 --MRI 0 Example manually opening the atrial orifices of a closed geometry: python main.py --mesh mesh/LAEAM.vtp --openorifices 1 --MRI 0 --usecurvaturetoopen 0 Example using a closed surface derived from a MRI segmentation to produce a volumetric atrial model: python main.py --mesh mesh/mwk05bi.vtp --closedsurface 1 --usecurvaturetoopen 0 --atrium LARA ```

Q&A

• Selection of appendage apex: the selected point will be used as boundary condition for a Laplacian problem. Therefore, the point at the center of the appendage is the most suitable to identify the whole appendage body
• FiberLA: LAA labeling (check LPVs identification functions distinguishPvs and optimizePVs in lagenerate_fiber.py)
• FiberRA: PMs (check step in function Method.downsamplepath in rageneratefiber.py), bridges (boolean operations and normal directions of original mesh)

Citation

When using this work, please cite

AugmentA: Patient-specific Augmented Atrial model Generation Tool Computerized Medical Imaging and Graphics 2023;108:102265 Luca Azzolin, Martin Eichenlaub, Claudia Nagel, Deborah Nairn, Jorge Sánchez, Laura Unger, Olaf Dössel, Amir Jadidi, Axel Loewe doi:10.1016/j.compmedimag.2023.102265

License

All source code is subject to the terms of the Academic Public License. Copyright 2021 Luca Azzolin, Karlsruhe Institute of Technology.

Contact

Luca Azzolin
Institute of Biomedical Engineering
Karlsruhe Institute of Technology
www.ibt.kit.edu