What is computational biomechanics?
Computational biomechanics uses mathematical models and numerical methods to study the mechanics of biological systems. In soft-tissue applications, this often means combining finite-element simulation, nonlinear constitutive modeling, deformation measurements, and medical images to estimate stress, strain, material properties, or internal mechanical contrast.
The field connects physical laws with data. Forward models predict deformation from known geometry, loading, and material properties. Inverse models work in the opposite direction: they use measured deformation or reaction forces to infer the unknown constitutive parameters or spatial material fields that produced those observations.
Nonlinear soft-tissue mechanics
Biological tissues commonly undergo large deformations and cannot be described adequately by linear elasticity. Hyperelastic models represent the stress response through a strain-energy density function. Common examples include Neo-Hookean, Mooney–Rivlin, Ogden, Gent, and polynomial formulations.
A central challenge is that several models or parameter combinations can produce similar responses under limited loading. This creates non-uniqueness and makes experimental design, identifiability analysis, and uncertainty assessment essential. My research studies how loading protocols can be selected to maximize information and how nonlinear material fields can be reconstructed from volumetric data while respecting mechanical equilibrium.
Medical imaging and inverse mechanics
Volumetric imaging can provide anatomical geometry and motion fields, but converting those measurements into quantitative material properties requires a mechanical model. Inverse elastography and mechanics-based imaging integrate deformation data with constitutive laws and boundary information to estimate tissue properties.
My work focuses on spatially resolved nonlinear characterization rather than only effective stiffness. This includes stress-constrained learning, voxel-wise parameter recovery, and mechanics-informed features that may help distinguish heterogeneous or pathological tissue regions.
Ongoing research directions
My current work is organized as a connected research program rather than a single project. The shared objective is to recover reliable, spatially resolved mechanical information from deformation and imaging data while remaining explicit about identifiability, uncertainty, and model assumptions.
Active directions include inverse tissue mechanics, deformation-informed medical and ultrasound imaging, uncertainty-aware constitutive characterization, comparison and compatibility of nonlinear material models, experimental and robotic measurement workflows, and mechanics-informed machine learning. Individual projects may emphasize different data sources or computational tools, but they are linked by the same goal: converting observed deformation into interpretable information about tissue mechanics.
Research tools and methods
The research program combines finite-element analysis, continuum mechanics, optimal experimental design, sensitivity analysis, medical-image processing, three-dimensional convolutional neural networks, and GPU-accelerated scientific computing. The objective is to build methods that are physically interpretable, reproducible, and suitable for future experimental and imaging applications.