Accuracy of computed radiography in osteometry: a comparison of digital imaging techniques and the effect of magnification

Abstract

Introduction Osteometric data may be obtained using digital imaging techniques, such as post-mortem computed tomography (PMCT) and digital radiography, non-invasively and without ethical objection. Osteometric data from PMCT is known to be as accurate as taking direct measurements. Digital radiography is more accessible and affordable than PMCT but is limited due to the magnification of its subjects.

Objectives To investigate the viability of implementing digital radiographic techniques for measurement of long bones, and to establish whether magnification can be accurately corrected for.

Materials and Methods Twenty hind pig (Sus scrofa) legs were imaged using computed radiography (CR) and PMCT, and osteometric data obtained from the digital images and 3D CT volume reconstructions. Direct measurements were taken following maceration. A calibration object was imaged using CR, to provide magnification correction factors.

Results Accuracy was determined by mean absolute error (AE), giving values of 3.3 ± 2.5 mm for PMCT (MPR), 2.4 ± 1.3 mm for PMCT (3D), 11.1 ± 7.4 mm for CR (PA), and 18.3 ± 14.5 mm for CR (LAT). PMCT data was more accurate than CR data. Through applying correction for magnification, CR data became closer to the direct measurement data, and stature estimation was substantially altered.

Conclusion Magnification in computed radiography affects osteometric data and resulting stature estimations. Applying correction factors may be a viable option for improving accuracy. For digital radiography to be used reliably in forensic anthropology, further empirical research is needed to validate a magnification correction method.