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.

Data

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.

Code

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

Cite this work

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}
}

Questions

To get help on how to use the data or code, simply open an issue in the GitHub "Issues" section.