Summary

Proximal femoral version plays a key role in torsional deformities. There is no consensus on the best method to evaluate it. The aim of this study is to evaluate the reproducibility of a new method to measure the proximal femoral version using the Maximum Intensity Projection (MIP) imaging on CT scan

Abstract

Introduction

Femoral version is a key parameter in different pathologies at both the hip and knee, particularly in torsional deformity evaluation. Different measurement methods are described in the literature, with no consensus on the most reproducible and reliable one. The aim of the study is to evaluate a new measurement method for proximal femoral version evaluation, simple and using the maximum intensity projection (MIP) algorithm applied to standard Computer Tomography (CT) scans and to compare it with the existing methods.

Methods

116 consecutive bilateral lower limb CT angiographies were screened. 50 patients (50 right femurs) were included (15 females and 35 males, mean age 72.4±9.7 years). At time 0 and after 30 days, 2 observers independently measured proximal femoral version with a standard digital software using different techniques, including: Murphy, Hernandex, Weidelich, Yoshioka and our original method (MIP). Intra observer reliability was tested with the Pearson correlation index. The inter-observer reliability was calculated with intraclass-correlation-coefficient (ICC) and Kappa coefficient. Differences between the five measurement techniques were evaluated with the Bonferroni post hoc tests.

Results

Significant differences between the average values for femoral torsion were observed (Murphy:18.4° ± 7.2°; Waidelich: 24.6° ± 7.4°; Yoshioka: 11.5° ± 7.8°; Hernandez: 10.7° ± 7.0°; MIP: 13.7° ± 6.2°). Intraobserver (observer 1: 0.821-0.97, P<0.0001; observer 2: 0.920-0.991, P<0.0001) and interobserver agreement (ICC 0.604–0.843; kappa value 0,318 - 0,711) showed moderate to high level of reproducibility for all techniques, with the highest results being observed with MIP technique (ICC: 0.8426, IC95% 0.7369 – 0.9080; kappa value: 0.711, IC95% 0.604 - 0.817).

Conclusions

Based on the results obtained, this new measurement method has demonstrated greater intra- and inter-observer reproducibility compared to the methods currently used in the literature, thus representing a valid alternative for use in future studies for torsional deformity evaluation.