Simulations of shear wave propagation in tissue based on kidney transplant histology

Abstract

End-stage kidney disease is the result of progressive declines in function and frequently requires a transplant. Maintaining kidney transplant health with regularly scheduled function tests or biopsies is necessary to adjust treatments. Ultrasound shear wave elastography (SWE) could provide an inexpensive, noninvasive modality to monitor allograft health. In this project, we built models of the kidney cortex from segmented kidney transplant histology images. We assigned viscoelastic properties, based on a Kelvin-Voigt model, to the major constituents, glomeruli, interstitia, tubules, and fluid and simulated the wave propagation through tissue with no disease, inflammation, interstitial fibrosis, and tubular atrophy. The resulting in silico wave motion in the time- and frequency-domains was characterized to compare the in silico and in vivo mechanical properties. We also attempted modeling the effects of perfusion by modifying the constituents’ viscoelastic properties. We found the most differentiation for shear elasticity, shear viscosity, and phase velocity dispersion in the inflammation patients. When accounting for perfusion we observed increases in shear wave group velocity, shear viscosity, and dispersion, bringing the results from in silico models in closer agreement with in vivo results. These results are encouraging and demonstrate first steps of simulating the shear wave propagation based on histological models.