Liver physics model based on Deep Learning and SOFA

This is a code repository for training deep learning neural networks in physics using the SOFA environment. Deep Learning is applied to a liver model as an example. The repository is divided into four parts:

1. Definition of the SOFA environment liver model with appropriate material properties
2. Python code for generating data (batches) based on node displacement in SOFA simulations.
3. A neural network design for learning and predicting strain.
4. Visualization of model strain driven by the neural network.

Minimal workloop:

MAIN_Generate_scenarios.py --> PyTorch_learning.py --> network_PyTorch.pkl --> Liver_on_pygames.py