Abstract
To predict bone strength in the case of osteoporosis, it could be a real benefit to assess the three-dimensional (3D) geometry and the bone mineral density (BMD) with a single low-dose X-ray device, such as the EOS system (Biospace Med, Paris, France). EOS 3D reconstructions of the spine have already been validated. Thus, this study aims at evaluating the accuracy of this low-dose system as a densitometer first ex vivo. The European Spine Phantom (ESP) (number 129) was scanned ten times using both the EOS and a Hologic device (Hologic, Inc., Massachusetts, USA). Accuracy was given by the sum of the systematic error (difference between BMDs assessed and true values given by the phantom manufacturer) and the random error (coefficient of variation). EOS BMDs and Hologic BMDs of 41 ex-vivo vertebrae were calculated and compared. The reproducibility of the method evaluating the EOS BMD was assessed giving the coefficient of variation of three measurements of the 41 vertebrae. The accuracy of the EOS system is below 5.2 per cent, versus 7.2 per cent for the Hologic system in the same conditions. EOS BMDs are significantly higher than Hologic BMDs, but they are strongly correlated. The reproducibility of the method of assessment is equal to 0.95 per cent. The EOS system is accurate for ex-vivo BMD assessments, which is promising regarding the use of this new system to predict vertebral strength.
Get full access to this article
View all access options for this article.