Reliability of a new method for lower-extremity measurements based on stereoradiographic three-dimensional reconstruction.

INTRODUCTION Several clinical and radiological techniques have been described to assess lower limb length and angle measurements. None of them has yet met the ideal criteria for a reliable, reproducible, safe, and inexpensive system. In this context, a new biplanar X-ray system (EOS™, EOS imaging, Paris, France) makes it possible to obtain a 3D reconstruction of the lower extremities from two 2D orthogonal radiographic images, with associated calculation of 3D measurements. The reliability of this technique has never been documented on adults. HYPOTHESIS Lower limb measurements produced by the 3D EOS™ reconstruction system are reproducible regarding inter- and intraobserver assessment and more reliable with this 3D technique than when they are obtained from 2D measurements. MATERIALS AND METHODS This study included 25 patients awaiting total hip arthroplasty (50 lower limbs). Two independent observers made all measurements twice, both on the 2D frontal radiograph and using 3D reconstructions (femoral measurements of length, offset, neck shaft angle, neck length, and head diameter, as well as the tibia length, limb length, HKA and HKS). Reproducibility was estimated by intraclass correlation coefficients. RESULTS Both the inter- and intraobserver reproducibility of the EOS™ measurements was excellent; more specifically inter- and intraobserver reproducibility was 0.997 and 0.997 for femoral length, 0.996 and 0.995 for tibial length, 0.999 and 0.999 for limb length, 0.894 and 0.891 for HKS, 0.993 and 0.994 for HKA, 0.870 and 0.845 for femoral offset, and 0.765 and 0.851 for neck shaft angle. For most of the variables, the interobserver correlations were statistically better with the EOS™ 3D reconstruction. DISCUSSION Our results show that the EOS™ systems allow reproducible lower limb measurements. Furthermore, 3D EOS™ reconstructions offer better reproducible measures for most of the parameters than radiographic 2D projection. Its use before deciding on surgery and during planning for lower limb arthroplasty appears essential to us. LEVEL OF EVIDENCE Level III: diagnostic prospective study on consecutive patients.