Development of Patient-Specific Lattice Structured Femoral Stems Based on Finite Element Analysis and Machine Learning
Hip implant optimization is increasingly receiving attention due to the development of manufacturing technology and artificial intelligence interaction in the current research. This study investigates the development of hip implant stem design with the application of lattice structures, and the util...
Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2025-07-01
|
| Series: | Crystals |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2073-4352/15/7/650 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Hip implant optimization is increasingly receiving attention due to the development of manufacturing technology and artificial intelligence interaction in the current research. This study investigates the development of hip implant stem design with the application of lattice structures, and the utilization of the MATLAB regression learner app in finding the best predictive regression model to calculate the mechanical behavior of the implant’s stem based on some of the design parameters. Many cases of latticed hip implants (using 3D lattice infill type) were designed in the ANSYS software, and then 3D printed to undergo simulations and lab experiments. A surrogate model of the implant was used in the finite element analysis (FEA) instead of the geometrically latticed model to save computation time. The model was then generalized and used to calculate the mechanical behavior of new variables of hip implant stem and a database was generated for surgeon so they can choose the lattice parameters for desirable mechanical behavior. This study shows that neural networks algorithms showed the highest accuracy with predicting the mechanical behavior reaching a percentage above 90%. Patients’ weight and shell thickness were proven to be the most affecting factors on the implant’s mechanical behavior. |
|---|---|
| ISSN: | 2073-4352 |