Under Construction

Determining heterogeneous mechanical properties of biological tissues via PINNs

The data and code for the paper W. Wu, M. Daneker, K. T. Turner, M. A. Jolley, & L. Lu. Identifying heterogeneous micromechanical properties of biological tissues via physics-informed neural networks. Small Methods, 2400620, 2024..

Data

All data are in the folder data. The first word in the file name indicates the example name, and the last word before ".npy" indicates the constitutive model name. For example, "GRF_equi_disp0.4_neo.npy" contains data for the Gaussian random field example generated using the Neo-Hookean material model.

Code

All code are in the folder src. The code depends on the deep learning package DeepXDE v1.10.2.

• Neo-Hookean material model
• Mooney Rivlin material model
• Gent material model

To run the code, specify name of reference data file and network architecture type (defaults shown):

python NeoHookean_elasticity_map.py --data ../data/GRF_equi_disp0.4_neo.npy --network 2B

Cite this work

If you use this data or code for academic research, you are encouraged to cite the following paper:

@article{wu2024heterogeneousmaterial,
  author  = {Wensi Wu and Mitchell Daneker and Kevin T. Turner and Matthew A. Jolley and Lu Lu},
  title   = {Identifying heterogeneous micromechanical properties of biological tissues via physics-informed neural networks}, 
  journal = {Small Methods},
  pages   = {2400620},
  year    = {2004},
  doi     = {https://doi.org/10.1002/smtd.202400620}
}

Questions

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