SUMMARY

In the field of regenerative medicine cells, exogenous factors, and scaffolds are applied to support cell and tissue growth to restore physiological tissue function. Supportive scaffolds play a key role in this process, where enhancing performance and controlling the rate of scaffold degradation are important. Modeling and simulation is an efficient tool to investigate the effect of scaffold degradation and enable design optimization compared to costly in vivo biodegradation studies. In this work, multi-scale modeling techniques are developed to design biodegradable polymeric scaffolds. A multi-objective Bayesian optimization approach is introduced to develop new molecular dynamic force fields for more accurate predictions of water diffusion in polycaprolactone (PCL) and its mechanical properties. A reduced-order kinetic Monte Carlo modeling scheme is created to reduce the computational time for water diffusion simulations and to simulate the hydrolysis reaction and subsequent biodegradation of PCL scaffolds under physiological conditions. Assisted by simulations, structure-property (SP) relationships are established to predict the degraded properties throughout the lifecycle from the initial scaffold geometry. A new extended surfacelet transformation method is developed as the structural descriptor to capture the geometric information of the scaffold. A surrogate of the SP relationship is constructed to predict mechanical property changes during biodegradation, which provide an efficient tool for rapid exploration of design variations for target scaffold lifetime.