AI-planned and robotic-assisted pedicle screw placement improves accuracy and reduces loosening in lumbar fusion for patients with low bone mass: a multicenter study.

This study aimed to evaluate the effectiveness of artificial intelligence (AI)-based pedicle screw trajectory planning combined with robotic-assisted screw placement in patients with lumbar degenerative disease (LDD) and low bone mass, based on a multicenter study. This retrospective study included patients with LDD and low bone mass, defined as a bone mineral density of 40-120 mg/cm3 measured by quantitative computed tomography, a range previously shown to benefit from AI-assisted screw trajectory planning. All patients underwent open posterior lumbar fusion between October 2022 and February 2023 and were categorized according to surgical technique into the AI-assisted (AI) group, which used AI-based planning with robotic-assisted screw placement, or the free-hand (FH) group, in which screws were placed manually. Clinical and radiographic outcomes were compared between the two groups. A total of 90 patients were included, with 42 in the AI group and 48 in the FH group. The mean follow-up duration was 29.7 months. Compared with the FH group, the AI group achieved a higher proportion of Grade A screw position and a lower rate of screw loosening. The AI group showed better preservation of intervertebral space height at final follow-up and shorter bone graft fusion time. Both groups showed significant postoperative improvements in Visual Analog Scale (VAS) and Oswestry Disability Index (ODI) scores, and the AI group exhibited significantly lower VAS scores for low back pain and ODI scores at final follow-up. AI-assisted pedicle screw trajectory planning combined with robotic guidance improves screw placement accuracy, preserves intervertebral space height, accelerates bone graft fusion, and enhances clinical outcomes in patients with LDD and low bone mass.