Design optimization of coronary stent based on finite element models

Area of Science:

• Biomedical Engineering
• Mechanical Engineering
• Computational Mechanics

Background:

• Stent implantation is crucial in treating vascular diseases.
• Stent dogboning, an undesirable expansion effect, can compromise treatment efficacy.
• Optimization methods are needed to improve stent deployment and reduce adverse effects.

Purpose of the Study:

• To develop and evaluate an effective optimization method for reducing stent dogboning.
• To investigate the influence of different finite element models on stent expansion.
• To assess the impact of optimized stent design on thrombosis risk.

Main Methods:

• Utilized a Kriging surrogate model combined with modified rectangular grid sampling for optimization.
• Employed an expected improvement (EI) criterion for balancing local and global search in optimization iterations.
• Validated the method using four finite element models of stent dilation and three thrombosis models.

Main Results:

• The proposed optimization method effectively reduced the stent dogboning effect.
• Finite element models including artery and plaque demonstrated superior stent expansion behavior.
• Models without artery and plaque offered computational efficiency for optimization.

Conclusions:

• The Kriging surrogate model with EI sampling is an effective approach for optimizing stent expansion and minimizing dogboning.
• Including anatomical structures like arteries and plaques in simulations provides more accurate insights into stent performance.
• Computational efficiency can be achieved by selecting appropriate finite element models for optimization studies.