About this Abstract |
| Meeting |
2026 TMS Annual Meeting & Exhibition
|
| Symposium
|
Advanced Biomaterials and Implants
|
| Presentation Title |
A-5: Patient-Specific Implant Evaluation in Osteosarcoma Using Machine Learning |
| Author(s) |
Behzad Amirzade, Tareq Zobaer, Janis Lapsley, Laura Selmic, Ali Nassiri |
| On-Site Speaker (Planned) |
Behzad Amirzade |
| Abstract Scope |
Canine bones affected by osteosarcoma (OSA) are highly susceptible to pathological fractures following Stereotactic Body Radiation Therapy (SBRT). The IlluminOss System (IS), a minimally invasive stabilization technique using a light-curable monomer within a balloon catheter, has shown promise in humans but demonstrated high failure rates in canine clinical trials. To investigate IS performance, spherical nanoindentation at high penetration depths was used to determine the yield properties of trabecular and cortical bone. Nonlinear finite element analysis (FEA) was then conducted on CT-reconstructed models of healthy and cancerous canine bones implanted with optimally designed IS devices to estimate fracture initiation loads under various loading conditions. A machine learning (ML) model trained on patient-specific clinical features and FEA results showed strong predictive performance and offered a rapid, viable alternative to computationally expensive simulations. This physics-informed framework can be extended to evaluate alternative implant materials and configurations, supporting efficient preclinical decision-making in veterinary orthopedics. |
| Proceedings Inclusion? |
Planned: |
| Keywords |
Modeling and Simulation, Computational Materials Science & Engineering, Characterization |