Abstract
BACKGROUND: Quantitative computed tomography (QCT) is essential for assessing osteoporosis and monitoring spinal deformities. "Clari-QCT," a software that uses artificial intelligence to analyze conventional computed tomography (CT) scans and produce QCTequivalent reports. This study aims to evaluate the effectiveness of Clari-QCT by comparing its results with traditional QCT, with the goal of validating new diagnostic tools for spinal deformities. METHODS: The study analyzed dual energy X-ray absorptiometry, CT, and QCT data from 18 patients at Inha University Hospital. Clari-QCT software was evaluated for its ability to generate QCT-equivalent reports from CT images. The software processes abdomen CT images, calculates bone density in designated slices, and provides bone mineral density (BMD), T-score, and Z-score values. Patients were classified into normal, mild, and severe spinal deformity groups. Intraclass correlation coefficient (ICC) analysis was used to measure the agreement between actual and predicted BMD values. RESULTS: The study included participants with an average age of 64 and a mean body mass index of 24.88. The average BMD was 94.7 g/cm³ by QCT and 122.5 g/cm³ by Clari- QCT, with individual differences ranging from 4.9 to 61.8. T-score discrepancies ranged from 0.16 to 6.86. ICC analysis showed moderate to high agreement between methods, with ICC1 values of 0.597, ICC2 of 0.64, ICC3 of 0.81, and ICC1k, ICC2k, ICC3k values ranging from 0.748 to 0.895. CONCLUSIONS: Clari-QCT demonstrates good agreement with actual QCT measurements in normal and severe spinal deformity groups but shows reduced accuracy in patients with mild deformities. If the limitations are addressed, it could become a useful tool for monitoring bone health in patients with spinal deformities.