The establishment and biomechanical analysis of finite element model for halo-pelvic traction in scoliosis correction.

Background: Halo-pelvic traction (HPT) is increasingly used for severe spinal deformity correction, but its biomechanical mechanisms remain poorly understood, particularly lacking of a comprehensive finite element model incorporating the complete Halo-pelvic-spine-pelvis construct. This study aimed to establish a comprehensive halo-pelvic traction model and assess its biomechanical reliability. Methods: A severe kyphoscoliosis (SS) model was created using patient CT data, reconstructed in Mimics, optimized in Geomagic Wrap, and assembled in UG12 for finite element analysis in Ansys. Material properties and appropriate boundary conditions were defined. Model validity was verified by measuring geometric parameters and the stress loading tests including T1-T4 range of motion, T12-L2 stiffness and L4-L5 displacement and comparing results with published data. An equivalent adolescent idiopathic scoliosis model was also developed (MS). Both models were analyzed under traction displacements (50, 80, 100, 125 and 150 mm). Results: The SS model closely matched clinical measurements with differences of less than 0.5° in Cobb and less than 2 mm in apical vertebral translation and spinal balance parameters. The T1-T4 mobility was lower than literature values, due to the involvement of T1-T4 segments in compensatory curvature formation. The T12-L2 stiffness and L4-5 displacement were consistent with published data, confirming the model's validity. Using patient weight-based maximum traction force, the MS model results demonstrated that at 150 mm distraction distance, the pelvic pins reached their maximum von Mises stress of 956.99 MPa with 2.29% pelvic pin tract strain, while at 125 mm, the cranial pins showed 2.39% strain and the support rod reaction force exceeded the maximum traction force. In the SS model, as traction increased, the stresses in both pelvic and cranial pins, pelvic pin tract strain, and support rod reaction forces all showed increasing trends, peaking at 150 mm without reaching critical thresholds. The cranial pin tract strain followed the same trend, reaching 2.26% at 150 mm. Conclusion: The validated finite element models demonstrate high anatomical and biomechanical accuracy. HPT may serve as an interim treatment for MS models with strict traction force and displacement control. While rigid SS models require dynamic force adjustment and distraction distances maintained below 150 mm for optimal safety.