Comparative Accuracy of Lower Limb Osteometry from CT Scans Versus Direct Bone Measurements

Abstract Radiographic imaging is increasingly used for osteometric data collection in clinical and forensic settings, although measurement accuracy may be affected by imaging protocols and scanning parameters. This study evaluated the agreement between CT-derived lower-limb skeletal measurements and direct dry-bone measurements. Maximum (MLF) and bicondylar (BLF) femoral lengths and maximum (LMT) and malleolar-trochlear (MTL) tibial lengths were recorded from 68 femora and 49 tibiae from anatomical collections using an osteometric board, and through CT images of the same bones. Agreement was assessed using paired comparisons, absolute-agreement intraclass correlation coefficients (ICCs), Bland–Altman analysis, technical error of measurement (TEM), and relative TEM (rTEM). There were statistically significant differences between CT-derived measurements and direct measurements for MLF, BLF, LMT, and MTL; however, mean biases were small, ranging from − 0.46 to 0.76 mm. No significant proportional bias was detected. Intra-observer rTEM ranged from 0.10% to 0.41%, with ICCs of 0.9936–0.9993, while inter-observer rTEM ranged from 0.11% to 0.23%, with ICCs of 0.9975–0.9992. CT-derived femoral and tibial measurements demonstrated excellent agreement with direct dry-bone osteometry. Although small statistically significant method differences were present, their magnitude was unlikely to be practically meaningful. In summary, measuring from CT images provided precise, repeatable, and reproducible dry skeletal measurements and represents a reliable non-invasive alternative for tibia and femora at autopsy, when direct skeletal measurement is unavailable or impractical.