Reproducibility of BMD Measurements in the Prosthetic Knee Comparing Knee-Specific Software to Traditional DXA Software: A Clinical Validation

The aim of this study was to validate new knee-specific dual X-ray absorptiometry (DXA) software for cemented total knee arthroplasty (TKA) before initiation of a randomized controlled trial. Firstly, in a phantom study, we evaluated if cementation influenced the measured BMD (g/cm2), the scan reproducibility with the new knee-specific software, and the consequences of leg rotation around a vertical axis. Secondly, in a clinical study, we assessed the clinical reproducibility in repetitive scans performed with the new knee-specific software and with traditional spine-mode DXA software, and further compared the 2 softwares' ability to point type implant and bone edges correctly. Cementation increased the measured bone mineral density (BMD) (p < 0.01). For reproducibility, the coefficient of variation (CV) was 0.52-0.70% in vitro. Leg rotation around a vertical axis significantly changed the measured BMD in most scans. Automatic point typing of implant and bone edge was of varying quality with frequent need of manual correction for both softwares. CVs of clinical reproducibility ranged from 2.78% to 6.19% for knee-specific software and from 1.45% to 6.06% for spine-mode software. We found the new knee-specific software valid for BMD measurement of the bone in proximity of cemented TKA and with clinical reproducibility and corrections of point typing similar to traditional spine-mode software.