Bayesian Inference For Anisotropic Soft Tissue Characterization Using Full-Field Optical Data
Please login to view abstract download link
Employing a non-deterministic approach, Bayesian inference proves instrumental in determining the mechanical parameters of solid materials. It has the ability to secure global optimization and quantify uncertainty on each identified parameter, conversely to the deterministic inverse problems. This study focuses on applying the Bayesian method to characterize skin anisotropy using full-field displacement data. The data, obtained from a multi-axial stretch induced by a ring suction in vivo test on human skin through the CutiScan ® CS100 device, was processed using the Digital Image Correlation technique. The targeted parameters correspond to mechanical constants of a transverse isotropic linear model that could be interpreted as Langer’s line and elastic moduli along and across it. In the computational problem, we employed a Markov chain Monte Carlo method to draw parameter samples and investigate posterior distributions.
