Assessing the accuracy and precision of manual registration of both femur and tibia using EOS imaging system with multiple views

Abstract

1. Introduction The EOS imaging system produces two simultaneous X-rays of a subject. Both views are used for the generation of a sub- ject-specific 3D skeletal model. This model then allows to obtain quantitative parameters concerning the morphology and the positioning of bones. Furthermore, thanks to the low dose of the EOS imaging, it is possible to take several stereoradiographies of the same subjects in order to study different positions of the fem- oro-patellar joint (Dagneaux et al. 2014) or to analyze different positions of the subjects. Also, it can be a way to compare tibia and femur relative positions without and with an orthotic device such as a brace for example. For that, the 3D model, obtained from the reference standing position, is registered across the different pairs of X-rays for the other positions. The goals of this study are 1/ to evaluate the reproducibility of the registration of both in vitro and in vivo and 2/ to a o assess the accuracy of this manual method using in vitro data.