Significance of Prediction Techniques in Spinal Fusion Surgery

Pedicle screw fixation (PSF) is a well-established therapy for a variety of spinal problems, including spinal disc degeneration, spondylolisthesis, tumors, and spinal deformities. Insertion techniques include freehand and imaging-guided techniques. PSF is a commonly used treatment for spinal instability and another type of spinal problems such as degenerative spinal illnesses, scoliosis, vertebral fractures, etc. On the other hand, loosening the pedicle screw and pullout due to weak bone-screw inter-facial strength is rather common, particularly in those with osteoporosis. This study utilized the 3D slicer tool to generate and segment the 3rd lumbar vertebra (L3) in 3-Dimension (3D). 104 Computed Tomography Scan (CT scan) images with a slice thickness of 1mm were used as input for the tool. The dimensions of the vertebra of interest are used as the input to the prediction algorithm. In this case machine learning modalities such as Regression Decision Tree, Support Vector Machines(SVM), and artificial neural networks can be used. The results of this study should be helpful to spine surgeons in selecting the preoperative, intraoperative, and postoperative management strategies that will reduce the rate of postoperative failure. Furthermore, medical costs could be reduced by selecting the most appropriate surgical treatment based on the new pedicle screw insertion angle range prediction method and calculation of bone mineral density (BMD) with more accuracy.