فرينا دهقان نژاددرارانداش

Aseptic loosening is the primary cause of long-term failure in hip replacement prostheses, predominantly resulting from the stress-shielding phenomenon. Stress shielding originates from the stiffness mismatch between the bone an‎d the implant. Under daily loads such as walking, this phenomenon disrupts biological stimuli, leading to bone resorption an‎d eventual loosening. To address this issue without compromising other mechanical properties, a strategy of stiffness reduction through the porosification of the implant structure was investigated. A titanium alloy was selec‎ted for its optimal biocompatibility an‎d strength; however, structural modification was deemed essential for better compatibility with bone. Triply Periodic Minimal Surface (TPMS) porous structures, known for their tunable geometric properties, were chosen to achieve adequate stress dispersion while maintaining structural integrity. Given that specific patient medical conditions alter design requirements, a single, unique geometry is not universally applicable for all individuals with varying anatomies, physical health, an‎d activity levels. Consequently, instead of resorting to costly testing for all possible scenarios, a multi-objective Bayesian optimization algorithm based on machine learning was developed to target three key outputs: stress shielding, maximum implant stress, an‎d maximum bone stress. This approach enables the personalized geometric design of gradient porous implants tailored to individual patient conditions, facilitating the development of a superior, customized prosthesis for each patient. Following a mechanical performance comparison with four other porous geometries, the split-P implant was selec‎ted as the best option. It was simulated with a linear gradient in the design space, derived from mathematical computations, for implant porosity fabrication using the Finite Element Method. The formulated Bayesian optimization algorithm in this research, tasked with minimizing stress shielding an‎d maintaining maximum stresses within permissible limits for each component, successfully enhanced the femoral stem of the shelled linear-gradient split-P porous implant. The algorithm achieved an 80% improvement in stress shielding an‎d a 15% improvement in the maximum implant stress compared to a conventional solid-market prosthesis.

بهينه‌سازي چندهدفه‌ي كاشتني‌هاي متخلخل مفصل ران , تخلخل‌هاي TPMS گرادياني ايمپلنت‌هاي ران غيرسيماني , دسته‌ي راني متخلخل پروتزهاي جايگزيني مفصل ران

Porous femoral stem of hip replacement prostheses , Multi-objective optimization of porous hip implants

Farina Dehghan Nezhad Derarandash

Siavash Kazemirad