A Noninvasive Method for Determining Elastic Parameters of Valve Tissue Using Physics-Informed Neural Networks
The data and code for the paper W. Wu, M. Daneker, C. Herz, H. Dewey, J.A. Weiss, A.M. Pouch, L. Lu & M.A. Jolley. A Noninvasive Method for Determining Elastic Properties of Valve Tissue Using Physics-Informed Neural Networks, Acta Biomaterialia, 200, 283-298, 2025.
All data are in the folder data. The name preceding ".npy" indicates the data for a specified example. For example, "HLHS_TV_data.npy" contains data for the HLHS tricuspid valve example.
All code are in the folder src. The code depends on the deep learning package DeepXDE v1.12.1.
- 2D hollow cylinder
- 2D deflected circular plate
- 3D cone
- 3D HLHS tricuspid valve with Neo-Hookean material model
- 3D HLHS tricuspid valve with Lee-Sacks material model
To run the code:
python example1_hollow_cylinder.py
If you use this data or code for academic research, you are encouraged to cite the following paper:
@article{wu2025noninvasive,
author = {Wu, Wensi and Daneker, Mitchell and Herz, Christian and Dewey, Hannah and Weiss, Jeffrey A. and Pouch, Alison M. and Lu, Lu and Jolley, Matthew A.},
title = {A Noninvasive Method for Determining Elastic Parameters of Valve Tissue Using Physics-Informed Neural Networks},
journal = {Acta Biomaterialia},
volume = {200},
pages = {283-298},
year = {2025},
doi = {https://doi.org/10.1016/j.actbio.2025.05.021}
}
To get help on how to use the data or code, simply open an issue in the GitHub "Issues" section.