Michael Sacks, University of Texas-Austin
The ability to fully characterize and simulate the three-dimensional (3D) mechanical behavior soft biological structures, such as cells, soft tissues, and organs has been of great interest in health and disease. Their complex 3D hierarchical structures results in such features as anisotropy and structural and mechanical heterogeneity. Yet, the computational demands in performing such simulations remains challenging. This is particularly challenging in complex clinical problems, such as surgical simulations, where computational demands often preclude time efficient simulations. For example, in-silico implementation of complex 3D continuum soft tissue constitutive models to obtain the responses of varying boundary conditions and fibrous structures requires the solution of the associated hyper-elasticity problem, which remains impractical in translational clinical time frames. Machine Learning techniques have in recent years become an attractive new approach to address these issues, due to the prevalence of power computational tools and inexpensive GPU architectures. The goal of this mini-symposium is to bring together the state of the art in computational methods applied to the soft biological structures that utilize machine learning methods. Both fundamental and applied (e.g. medical devices) are equally encouraged. Topics at all length scales are encouraged, and include (but are not limited to):
1. New machine-learning methods as applied to soft biological structures.
2. Image-to-model approaches.
3. Organ level simulations.
4. Whole body simulations.
5. Uncertainty Quantification.
6. Gaussian Process and Neural Network approaches.
7. Translational applications
8. Numerical approaches to training.