Surrogate-based multi-objective design optimization of a coronary stent: Altering geometry toward improved...

Area of Science:

• Biomedical Engineering
• Computational Mechanics
• Medical Device Design

Background:

• Coronary stents require optimal biomechanical performance to ensure patient safety and treatment efficacy.
• Multi-objective optimization is crucial for balancing competing design criteria in medical devices.
• Finite element analysis (FEA) is essential for evaluating complex biomechanical behaviors of medical implants.

Purpose of the Study:

• To optimize coronary stent geometry for improved biomechanical performance using multi-objective optimization.
• To compare the efficacy of four surrogate-based optimization algorithms (EI_hv-EGO, P_hv-EGO, ParEGO, SMS-EGO).
• To identify optimal stent designs balancing multiple performance metrics like vessel injury and radial strength.

Main Methods:

• Utilized finite element models to calculate key performance metrics: vessel injury, radial recoil, bending resistance, longitudinal resistance, radial strength, and prolapse index.
• Employed surrogate-based multi-objective optimization algorithms with a limited sample budget.
• Assessed algorithm performance using hypervolume, R2, epsilon, and generational distance quality indicators.

Main Results:

• P_hv-EGO emerged as the top-performing algorithm overall.
• Analysis of the Pareto front revealed correlations and conflicts between objective functions.
• Cluster analysis identified solution families and trade-offs between design parameters and biomechanical performance.

Conclusions:

• The study successfully identified optimal coronary stent designs with enhanced biomechanical properties.
• P_hv-EGO is recommended for similar multi-objective optimization problems in biomechanical engineering.
• A constrained-based selection identified designs superior to the baseline across all objectives.