SciML Inverse problem in hemodynamic models Addressing ill-posed inverse problems due to the lack of identifiability in stiff hemodynamic models, using both synthetic and real-world EHR data. Emulation, probabilistic modeling and identifiability analysis using generative models InVAErt: A neural network-based, data-driven framework for enhanced digital twin analysis. Surrogate modeling for parallel finite element simulations Increasing the computational efficiency of finite element simulations by building a data-driven surrogate to avoid synchronizations in parallel computation. Computational mechanics
