A semi-automated bone mineral density measurement tool, phantom-less QCT: Technique optimization and clinical validation

Abstract Introduction: As the global population ages, osteoporosis poses a significant public health challenge. This study aimed to evaluate the accuracy and precision of a new phantom-less quantitative computed tomography (PL-QCT) technique for measuring T12 to L1 bone mineral density using thoracolumbar CT scans. Materials and methods: Waist fat and muscle tissue were used as calibration references for bone mineral density measurements in PL-QCT. Automatic bone mineral density measurements in PL-QCT were conducted using crude segmentation, tissue-specific thresholds, and a two-dimensional convolution algorithm. Region-of-interest selection in PL-QCT measurement was cross-referenced with phantom-based QCT (PB-QCT) through asynchronous calibration of body film. Measurement consistency between PB-QCT and PL-QCT was compared. Results: A total of 568 participants, aged 20 to 85 years (mean age, 49.74 ± 16.80 years), were included. Consistency analysis showed a mean deviation of 0.31 mg/cm³ in T12 to L1 BMD measurements between PB-QCT and PL-QCT, with the correlation coefficient (r) and Cronbach’s alpha values both ranging between 0.998 and 1.000. Bland–Altman analysis revealed that most T12 to L2 bone mineral density data points were within the 95% limits of agreement, demonstrating high accuracy. Conclusion: The PL-QCT system demonstrated excellent consistency in T12 to L2 vertebral BMD measurements when compared with the PB-QCT system. Overall, PL-QCT exhibited superior accuracy and precision, confirming its reliability as a tool to evaluate spinal bone mineral density.