Our research focuses on developing an advanced deep learning model to simulate the deformation and force behavior of soft tissues using geometric 3D data. By leveraging graph neural networks (GNNs), we aim to replace traditional biomechanical methods and create more realistic simulations. These simulations will not only offer visual representations of tissue behavior but also enable haptic feedback, making it possible to interact with the simulated tissue in a visuo-haptic setup.

To complement our data-driven approach, we are collecting experimental data and using finite element method (FEM) simulations to further inform and validate our model. By combining these data sources, we enhance the model's ability to accurately predict how tissues respond to forces, which could lead to significant advancements in medical training, surgical planning, and therapeutic techniques.