Objectives: Cam-type femoroacetabular impingement (FAI) is a three-dimensional (3D) deformity that is still difficult to assess using traditional two-dimensional (2D) radiographic views. While measurements of alpha angle and head-neck offset are used to estimate the likelihood of actual impingement, these 2D measurements do not account for z-axis variations in femoral version (FV) and neck-shaft angle (NSA). The purpose of this qualitative proof-of-concept study was to evaluate the potential variation in alpha angle and neck-shaft offset measurements with incremental changes in NSA and FV by simulating traditional radiographic views with software-generated Digitally-Reconstructed Radiographs (DRRs). We hypothesize that incremental changes in hip morphology will produce qualitative changes in alpha angle and neck-shaft offset. Methods: 3D-CT reconstruction images were obtained from one subject with symptomatic cam-type FAI. The 3D reconstruction was cleaned to include only the femoral head, neck and subtrochanteric region along with the ipsilateral hemipelvis. Using 3D medical image processing software (Mimics; Materialise, Inc.; Belgium), the pre-processed 3D model was manipulated in a standardized manner to simulate 5-degree incremental variations in FV and NSA (-15 degrees to +15 degrees for FV; -15 degrees to +10 degrees for NSA). Negative FV reflected external rotation of the femoral head-neck unit, whereas negative NSA reflected abduction of the femoral head-neck unit. Each modified 3D model was then used to generate DRRs corresponding to traditional 2D radiographic views used for assessment of cam-FAI (Anteroposterior [AP], False Profile [FP]), Cross Table Lateral [CTL], Frog Leg Lateral [FLL], 45- and 90-degree Dunn [45D and 90D, respectively]. Alpha angle and head-neck offset were measured on each radiographic view corresponding to each incremental change in FV and NSA. All measurements utilized the perfect circle technique and were made by two independent observers for assessment of inter-observer reliability. Two-way random effects ANOVA was used for statistical assessment of inter-observer reliability and reported as intra-class correlation coefficients (κ). Comparisons between groups were performed using two-tailed paired t-tests assuming unequal variance. P-values less than 0.05 were considered statistically significant. Results: Inter-observer reliability (κ) for head-neck offset and alpha angles were 0.46 (fair) and 0.88 (excellent), respectively. Variations in head-neck offset and alpha angle with incremental variations in FV and NSA are summarized in Figure 1. There were statistically significant changes in mean alpha angles when the NSA was adjusted from Neutral to -5 degrees (p=0.01) and from -5 degrees to -10 degrees (p<0.001). There were no statistically significant differences in alpha angles or head-neck offsets between each incremental change in FV (p<0.05). Alpha angle measurements were significantly more variable than head-neck offset measurements for all variations in FV (p<0.001) and NSA (p=0.02) (Figure 2). Conclusion: Two-dimensional evaluation of three-dimensional Cam morphology (alpha angle and head-neck offset) was found to be significantly affected by alterations in femoral version and head-neck offset. Head-neck offset measurements were significantly less variable than alpha angle measurements across all FVs and NSAs within each radiographic view. Future work should be done to develop standardized procedures for routine 3D radiographic assessment of cam-type FAI. [Figure: see text][Figure: see text]
Proximal femoral morphology after transtrochanteric posterior rotational osteotomy for osteonecrosis of the femoral head: A three-dimensional simulation study
