Multi-Objective Optimization Design of Synthetic Polymer Heart Valve Geometry

BACKGROUND: Polymeric heart valves (PHVs) face critical trade-offs between durability and hemodynamics. This study aimed to improve the flow, leakage, and stress distribution of a polyurethane-based valve by simultaneously optimizing leaflet thickness and height via a data-driven workflow. METHODS: A 3D fluid-structure interaction (FSI) model of a siloxane-modified polyurethane valve was developed. Using a central composite design, effective orifice area (EOA), regurgitant fraction (RF), and peak von Mises stress were analyzed. Response-surface surrogates were coupled with a multi-objective genetic algorithm (MOGA) to identify optimal geometries, which were verified against ISO 5840-3 standards. RESULTS: The optimized geometry (thickness 0.197 mm; height 15.49 mm) significantly outperformed the baseline. EOA increased from 140.5 mm2 to 164.5 mm2, RF decreased from 3.10% to 2.88%, and peak stress dropped from 1.88 MPa to 1.59 MPa. Surrogate model predictions showed high precision with = 6% deviation from simulation results. CONCLUSIONS: Coupled optimization of leaflet thickness and height yields a valve that opens wider and leaks less while experiencing lower mechanical stress. This workflow, combining FSI simulation and multi-objective search, provides a transferable template for the rational engineering of next-generation transcatheter polymeric valves.