Optimal screw orientation for the fixation of cervical degenerative disc disease using nonlinear C3-T2 multi-level spinal models and neuro-genetic algorithms.

Abstract

Anterior cervical discectomy and fusion is a common surgical procedure performed to remove a degenerative or herniated disc in cervical spine. Unfortunately, clinical complications of anterior cervical plate (ACP) systems still occur, such as weak fixation stability and implant loosening. Previous researchers have attempted to ameliorate these complications by varying screw orientations, but the screw orientations are mainly determined according to the investigator's experiences. Thus, the aim of this study was to discover the optimal screw orientations of ACP systems to achieve acceptable fixation stability using finite element simulations and engineering algorithms. Three-dimensional finite element models of C3-T2 multi-level segments with an ACP system were first developed to analyze the fixation stability using ANSYS Workbench 14.5. Then, artificial neural networks were applied to create one objective function, and the optimal screw orientations of an ACP system were discovered by genetic algorithms. Finally, the numerical models and the optimization study were validated using biomechanical tests. The results showed that the optimal design of the ACP system had highest fixation stability compared with other ACP designs. The neuro-genetic algorithm has effectively reduced the time and effort required for discovering for the optimal screw orientations of an ACP system. The optimum screw orientation of the ACP system could be successfully discovered, and it revealed excellent fixation stability for the treatment of cervical degenerative disc disease. This study could directly provide the biomechanical rationale and surgical suggestion to orthopedic surgeons.