Computational modeling and simulations in medicine: development of in silico approaches for personalized surgery

Computational modeling and simulation (CM&S) is emerging as a powerful tool in personalized medicine, enabling the investigation of disease mechanisms, virtual evaluation of medical devices, and optimization of therapeutic strategies. By developing patient-specific models, CM&S supports surgical planning and improves clinical outcomes. This thesis develops and applies advanced patient- specific computational frameworks, mainly based on finite element (FE) analysis, in Maxillofacial, Otologic, and Vascular Surgery to enhance mechanistic understanding and support individualized treatment planning. In Maxillofacial Surgery, an innovative and reproducible multi-level validation framework was developed to experimentally validate a patient-specific computational model for mandibular reconstruction using fibula free flaps. The analysis demonstrated the model’s reliability in predicting load distribution and deformation of fixation plates under physiological loading. Additionally, a mechanobiological FE model was implemented to simulate the ossification process following mandibular reconstruction. By integrating tissue differentiation algorithms, the model predicted the spatiotemporal evolution of bone healing and identified the fixation plate configuration that generates the most favorable mechanical environment for ossification, thereby informing the optimization of fixation strategies. In Otologic Surgery, a validated finite element model of the healthy middle ear was developed to compare strategies for autologous incus reconstruction. While all configurations restored mechanical continuity, one design achieved superior functional performance at low and mid frequencies, underscoring the importance of prosthesis geometry over material properties. In Vascular Surgery, a FE framework simulated embolization coil deployment in abdominal aortic aneurysms after endovascular repair. Simulations revealed progressive, non-uniform sac filling with predominant peripheral occupation, providing quantitative insights for optimizing endovascular planning. Overall, this work establishes a foundation for patient-specific, physics-based in silico strategies. The findings support optimization of fixation methods in mandibular reconstruction, refinement of prosthesis geometry for ossicular repair, and improved planning of endovascular embolization, contributing to the advancement of personalized surgical practice.