Ching-Chi Hsu and Yu-Rui Huang
Vertebral body cage, Loosening, Finite element methods, Artificial neural networks, Genetic algorithms
Vertebral body cages (VBCs) have been applied to different vertebral morbidities. However, interface loosening and subsidence of a VBC are the complications after a surgical operation. Generally, the interface pullout strength of VBCs was discussed by exerting a single direction loading (normally forward and backward). However, the VBC would take the loading from different directions, such as flexion, extension, lateral bending, and axial rotation. Therefore, the goal of this research was to search the VBC design with excellent interface pullout strength under different situations. Three-dimensional nonlinear finite element model was created to simulate the interface pullout strength of VBCs. To decrease the computational time, the artificial neural network (ANN) was used to substitute for the finite element model. The optimum VBC design would be obtained by using genetic algorithms (GAs). Two kinds of situations are discussed including a VBC with bone fusion condition and a VBC without bone fusion condition. The results showed that the spike rows should be as large as possible for both kinds of situations. This would produce the largest contact area between the spikes and vertebral body. In this study, ANNs and GAs could accurately and effectively predict the interface pullout strength of VBCs under different conditions. The optimum design might be useful information to engineers to develop a new VBC. In addition, this can help surgeons to select a suitable VBC to their patients.
Important Links:
Go Back
