Liver tissue characterization from uniaxial stress-strain data using probabilistic and inverse finite element methods.

Biological soft tissue is highly inhomogeneous with scattered stress-strain curves. Assuming that the instantaneous strain at a specific stress varies according to a normal distribution, a nondeterministic approach is proposed to model the scattered stress-strain relationship of the tissue samples under compression. Material parameters of the liver tissue modeled using Mooney-Rivlin hyperelastic constitutive equation were represented by a statistical function with normal distribution. Mean and standard deviation of the material parameters were determined using inverse finite element method and inverse mean-value first-order second-moment (IMVFOSM) method respectively. This method was verified using computer simulation based on direct Monte-Carlo (MC) method. The simulated cumulative distribution function (CDF) corresponded well with that of the experimental stress-strain data. The resultant nondeterministic material parameters were able to model the stress-strain curves from other separately conducted liver tissue compression tests. Stress-strain data from these new tests could be predicted using the nondeterministic material parameters.