2103 Comparative Analysis of Artificial Intelligence and Traditional X-Ray Parameter Measurements in Spinal Surgical Planning

INTRODUCTION: Accurate measurement of spinal parameters is essential for diagnosing and treating spinal disorders. Traditional manual measurement methods are time-consuming and prone to variability. METHODS: We conducted a comparative analysis of spinal parameter measurements obtained through an AI model and the same measurements manually taken by a neurosurgeon in 100 patients who underwent full-body X-rays. Parameters examined included thoracic kyphosis, lumbar lordosis, pelvic incidence (PI), sacral slope (SS), pelvic tilt (PT), sagittal vertical axis (SVA), C7 coronal alignment (C7 CSL), thoracic Cobb angle, and lumbar Cobb angle. Data analysis utilized Pearson coefficients, p-values, and interrater reliability. RESULTS: The results demonstrated a high correlation and significant agreement between measurements obtained through AI and Operator 1. Sacral slope (SS) and pelvic tilt (PT) demonstrated Pearson coefficients of 0.96 and 0.97 (both with p-values<.001) and interrater reliabilities of 0.94 and 0.92, respectively. Sagittal vertical axis (SVA) exhibited a Pearson coefficient of 0.96 (p-value <.001) with an interrater reliability of 0.96. PI-LL and PI had Pearson coefficients of 0.89 and 0.92, respectively (both with p-values<.001) and interrater reliabilities of 0.92 and 0.91. Kyphosis T1-T12 demonstrated a Pearson coefficient of 0.87 (p-value<.001) with an interrater reliability of 0.86. C7 CSL had a lower agreement with a Pearson coefficient of 0.82 (p-value<.001) and an interrater reliability of 0.80. Thoracic and lumbar Cobb angles showed Pearson coefficients of 0.84 (p-value<.001) and interrater reliabilities of 0.85 and 0.81, respectively. CONCLUSIONS: Artificial intelligence facilitates thoracolumbar spinal parameter measurements, demonstrating accuracy and reliability with good correlation and interrater reliability (=0.8) in all assessed thoracolumbar parameters.