The anterior center edge angle has limited ability to predict three-dimensional coverage of the femoral head in patients with developmental dysplasia of the hip undergoing curved periacetabular osteotomy

Abstract

Introduction

Femoral head coverage in patients with hip dysplasia (DDH) is typically quantified using 2D measurements of the lateral center edge angle (LCEA) and anterior center edge angle (ACEA). However, as the morphology of DDH is complex and varies between patients, 2D measurements may not predict the true 3D femoral head coverage. Herein, 2D and 3D coverage were quantified before and after curved periacetabular osteotomy (CPO) and their relationships were assessed.

Materials and Methods

Forty-three hips that underwent CPO for DDH were analyzed. For 2D evaluation, LCEA was quantified from X-rays and CT images. The ACEA was measured from CT images (CT-ACEA) and digitally reconstructed radiographs generated from CT images (DRR-ACEA). Three-dimensional coverage was quantified from CT reconstructions of the hip and evaluated in the anterior, superior, posterior, and inferior regions of the femoral head. Two-dimensional measurements were correlated to 3D coverage to assess their relationships.

Results

The median preoperative 3D percent coverage was 17.7%, 36.1%, 56.1%, and 14.6% for the anterior, superior, posterior, and inferior region, respectively. After CPO, all LCEAs and ACEAs increased significantly (all p<0.001). For the 3D coverage, anterior and superior coverage significantly increased while the posterior and inferior coverage decreased (all p<0.001). Moderate to strong correlations were detected between the two LCEAs and the 3D superior coverage in both the preoperative and postoperative period. For the correlation between 3D anterior coverage, no significant correlation was found between the CT-ACEA while a moderate correlation was found between the DRR-ACEA (r_s=0.41, p=0.023).

Conclusions

Our results indicate that the LCEA can be used to predict 3D coverage in the superior region of the femoral head. However, as the CT-ACEA or DRR-ACEA had no or only moderate correlation between the 3D anterior coverage, these measurements are not recommended for evaluating/estimating the 3D anterior coverage in patients with DDH.

Introduction

Periacetabular osteotomy is a common surgery to treat developmental dysplasia of the hip (DDH) [1–4]. Surgeons aim to improve symptoms and lower the likelihood of future osteoarthritis by increasing femoral head coverage through rotation of the acetabular fragment. Studies have reported good long-term clinical results of periacetabular osteotomy when femoral head coverage increases, but surgical outcomes are poor when the osteotomy fails to produce adequate containment of the femoral head [1–3]. Most often, clinical diagnosis and surgical planning is guided by evaluation of two-dimensional (2D) measurements of femoral head coverage, including the lateral center edge angle (LCEA [5]) and anterior center edge angle (ACEA [6]). Usually, an LCEA less than 20° is considered indicative of DDH, LCEA values between 20° and 25° are considered borderline DDH, and an LCEA between 25° and 45° is considered ‘normal’ [7].

The LCEA and ACEA provide a general sense of the severity of DDH. However, recent evidence suggests that patients with DDH present with complex, three-dimensional (3D) deformities, including acetabular and femoral version deformities [8–12] and an abnormally shaped femoral head [13,14]. Research has also demonstrated considerable variability with regards to the type of acetabular deficiency present, including global coverage loss as well as localized reductions in coverage in the anterosuperior and posterosuperior regions [15]. Such deficiencies in hip joint coverage may not be adequately evaluated through measurements of the LCEA and ACEA on X-ray alone [11].

Volumetric images provided by computed tomography (CT) or magnetic resonance imaging (MRI) can be segmented to create 3D reconstructions suitable for coverage analysis. However, segmentation is time-consuming and coverage analysis requires specialized software and domain expertise. These limitations may explain why few studies have reported changes in 3D femoral head coverage following periacetabular osteotomy [4,16]. Results from these prior studies suggest that periacetabular osteotomy improves global as well as regional coverage, notably in the lateral and anterior regions. Still, given the lack of ready-to-use tools, 3D coverage analysis has yet to become a clinical standard, leaving the 2D measurements of femoral head coverage to be used. To our knowledge, no studies have assessed relationships between 2D measurements with 3D coverage in both the pre- and postoperative status, and thus, it remains uncertain if 2D measurements can be used to estimate the true, 3D femoral head coverage. The objectives of this study were to: 1) quantify the changes in 2D and 3D head coverage after periacetabular osteotomy, and 2) analyze the relationship between 2D and 3D head coverage preoperatively and postoperatively.

Materials and Methods

This retrospective study was approved by the Institutional Review Board. Fifty-seven dysplastic hips of fifty Japanese patients who underwent periacetabular osteotomy between June 2009 and November 2017 by a single surgeon were considered for this study. For these consecutive 57 hips, periacetabular osteotomy was performed by curved periacetabular osteotomy (CPO), a modification of the Bernese periacetabular osteotomy wherein a curved osteotomy is created to facilitate a congruent interface between the fragment and pelvis once fixed [2,3]. The acetabulum was rotated laterally to achieve an LCEA of 30–40°. From a total of 57 hips, males (3 hips), hips with Kellgren-Lawrence classification [17] grade 3 (5 hips), and hips that had a change in the postoperative lunate surface (6 hips) were excluded, leaving 43 hips of 39 females for analysis (age: 37±10 years, height: 158±4cm, weight: 54±8cm, BMI: 21.5±2.9). Preoperative and postoperative anteroposterior radiographs acquired in a supine position, preoperative CT images acquired for surgical planning, and postoperative CT images acquired at a median of 8 weeks after surgery were utilized. CT images were acquired with a slice thickness of 1–1.25mm with the pelvis set at a neutral position and the knees and toes facing directly upward. Of note, the differences in hip angles between the preoperative CT images and postoperative CT images were 2.7°, 0.2°, and 5.0° for flexion, abduction, and internal rotation, respectively.

Measurement of 2D head coverage

Femoral head coverage was evaluated in 2D with the LCEA and ACEA defining superior and anterior coverage, respectively [18,19]. Two different LCEA measurements were obtained. First, the LCEA was measured from the anteroposterior plain film in Image J (v1.51, National Institutes of Health, Bethesda, MD, USA) and defined as the ‘Xray-LCEA’ (Fig. 1a). Next, LCEA measurements were obtained from the CT images after the pelvis was axially rotated until the bilateral anterior iliac spines were aligned on the same coronal plane. This step was necessary as positional differences in the orientation of the hip during CT imaging could affect measurements of coverage across patients and between pre- and postoperative states. After the axial rotation of the pelvis was established, the LCEA was measured on a coronal CT image slice that included the head center and referred to as ‘CT-LCEA’ (Fig. 1b) [18,19].

Fig. 1.

Measurements of Xray-LCEA (a), CT-LCEA (b), CT-ACEA (c), and DRR-ACEA (d) shown for a left hip. For the measurement of DRR-ACEA, only the hemi-pelvis was used to avoid overlap of the contralateral hip. For each figure, the dotted circle (cyan) expresses the best fit circle of the femoral head, the orange line indicates the horizontal axis of the pelvis that connects the most inferior point of the both ischium, the blue line indicates the vertical axis of the pelvis, and the magenta and yellow lines indicate the line connecting the head center and the edge of the acetabulum. Angles measured on each view are expressed in white.

As lateral X-rays were not available for the patients included herein, two different ACEA measurements were obtained from the CT images to evaluate anterior coverage. First, the ACEA was measured on the sagittal slice that included the head center and was referred to as ‘CT-ACEA’ [18–20] (Fig. 1c). The CT-ACEA was not recorded in cases where the CT-LCEA was less than 0°, as there was no coverage of the anterior region. Further, to simulate the ACEA usually measured on lateral X-rays, digitally reconstructed radiographs (DRRs) were created by sagittal projection of the CT data; anterior coverage was measured on the resulting DRR to define the ‘DRR-ACEA’ (Fig. 1d). All measurements from CT images were performed using software that allowed for multiplanar reconstruction and generation of DRR (3D-template; Kyocera, Kyoto, Japan).

Quantification of 3D head coverage

For the quantification of 3D coverage, surface models of the pelvis and femur were first semi-automatically generated from CT images using Amira (v.6.0.1, Visage Imaging, San Diego, CA, USA). From these surface models, the lunate surface (Fig. 2a) and the head surface excluding the fovea capitis (Fig. 2b) were selected using principal curvature in Postview (v.2.0, University of Utah, Salt Lake City, UT, USA), as described elsewhere [21]. The 3D coverage of the femoral head was then quantified using the Coverage Tool in Postview [22,23]. Surface elements of the femoral head were considered ‘covered’ if they were intersected by the normal projection of any element of the lunate surface. To define regional coverage, the femoral head was divided into four regions as done elsewhere (Fig. 2c) [22,23]. Briefly, the head region was divided around the neck axis in 90° increments using the neck axis and the knee center to define the regions of anterior, superior, posterior, and inferior in Matlab (v.7.10, The MathWorks, Natick, MA, USA). Coverage was then calculated in each quadrant and represented as a percent of the total available surface area for that region. Of note, errors in quantifying regional 3D coverage between observers using this method are reported as 0.2–1.1% [22]. Quantification of 3D coverage was performed both pre- and postoperatively and was correlated to the measurements of 2D coverage to study interactions. Further, the change in 2D coverage after CPO and the change in regional 3D coverage after CPO were correlated to evaluate the predictive capabilities of these 2D measurements.

Fig. 2.

Generation of the three-dimensional surface models of the lunate surface (a), femoral head surface (b), and the quadrants of the femoral head surface (c) shown for a left hip. (a) The lunate surface (dark brown) was selected from the pelvis surface model using second principal curvature and manual selection. (b) The femoral head excluding the area of fovea capitis (gray) was selected from the femur model using first principle curvature. (c) Using the femoral head center, femoral neck axis, and knee center, the femoral head surface was divided into four regions around the neck axis in 90° increments. Area of fovea capitis (white area) was excluded from the analysis. Ant, anterior; Sup, superior; Post, posterior; Inf, inferior.

Postoperative management and clinical assessment

Postoperative management was performed as has been reported in the literature [2,3]. Briefly, patients were set non-weight bearing for two days and started 10 kg of partial weight-bearing from day three using crutches. The amount of weight-bearing was increased every two weeks, and full weight-bearing was permitted at two months following CPO.

For clinical assessment, hip function scores were evaluated using Harris hip score (HHS, [24]) and Japanese Orthopaedic Association Hip Score (JOA H-S, [25]) preoperatively and at the most recent follow-up. Briefly, the JOA H-S consists of 4 subcategories: pain (0 – 40 points), range of motion (0 – 20 points), ability of walking (0 – 20 points), and activities of daily living (0 – 20 points), and has a full score of 100 points. Preoperative HHS and JOA H-A were compared with those collected at the most recent follow-up.

Statistical analysis

Data were represented as median (interquartile range) after confirming a non-normal distribution of data with a Shapiro-Wilk test. Pre- and postoperative 2D and 3D coverage was compared using Wilcoxon signed-rank test. Relationships between 2D and 3D measurements of coverage were assessed for each region using Spearman’s rank correlation. Additionally, the correlation coefficient (r_s) was calculated. The Benjamini-Hochberg procedure was used to correct for multiple comparisons. Further, as the measurements of 2D coverage was performed manually, the level of agreement was assessed by two Orthopaedic surgeons using the intra-class correlation coefficient (ICC). Specifically, all 2D measurements were obtained twice for the first 20 consecutive cases by a single surgeon to assess intra-observer agreement while the same 20 cases were measured by another surgeon to assess inter-observer agreement. For the purpose of evaluating correlation coefficients (r_s) and ICC, values of 0.00–0.19 were regarded as very weak, 0.20–0.39 as weak, 0.40–0.59 as moderate, 0.60–0.79 as strong, and 0.80–1.00 as very strong [26]. Values of p<0.05 were considered to represent statistical significance; all statistical analyses were performed using JMP (JMP Pro Version 14, SAS Institute Inc., Cary, NC, USA).

Results

Change in 2D measurements after CPO

In the preoperative assessment, the median Xray-LCEA, CT-LCEA, CT-ACEA, and DRR-ACEA was 12.1°, 11.8°, 43.3°, and 52.7°, respectively (Table 1). There were three cases in which the CT-ACEA could not be measured as the CT-LCEA was <0°. After CPO, all measurements increased significantly to achieve a median of 36.0° for Xray-LCEA, 34.3° for CT-LCEA, 59.2° for CT-ACEA, and 60.0° for DRR-ACEA (all p<0.001) (Table 1).

Table 1.

Change in 2D measurements and regional 3D coverage following curved periacetabular osteotomy.

Dimension	Measurement		Preoperative	Postoperative	Change
2D	LCEA	Xray-LCEA (°)	12.1 (7.2–16.3)	36.0 (31.1–42.4)	24.1 (18.6–29.9)
		CT-LCEA (°)	11.8 (6.0–17.8)	34.3 (28.7–39.1)	20.8 (14.7–30.3)
	ACEA	CT-ACEA (°)	43.3 (35.3–48.2)	59.2 (50.7–64.2)	16.0 (9.3–22.2)
		DRR-ACEA (°)	52.7 (47.3–55.7)	60.0 (52.0–64.1)	6.9 (−0.5–12.8)
3D	Anterior coverage (%)		17.7 (13.3–21.9)	22.6 (18.9–27.7)	5.3 (−0.3–10.1)
	Superior coverage (%)		36.1 (31.9–41.0)	55.1 (45.2–59.3)	15.8 (10.2–23.7)
	Posterior coverage (%)		56.1 (49.9–63.5)	44.5 (34.3–52.4)	−11.5 (−18.5–−5.6)
	Inferior coverage (%)		14.6 (11.6–16.9)	6.7 (4.0–9.9)	−6.0 (−10.1–−3.9)

Values expressed as median (interquartile range)

Change in 3D coverage in the regions of anterior, superior, posterior, and inferior after CPO

In the preoperative analysis, median 3D percent coverage relative to the available surface area in each head region was 17.7%, 36.1%, 56.1%, and 14.6% for the anterior, superior, posterior, and inferior region, respectively (Table 1). In the postoperative analysis, 3D anterior and superior coverage increased to 22.6% and 55.1%, respectively, while the posterior and inferior coverage decreased to 44.5% and 6.7%, respectively (all p<0.001) (Table 1).

Relationship between 2D and 3D head coverage

In the preoperative analysis, significant strong correlation was found between the LCEAs and 3D superior coverage (Xray-LCEA: r_s=0.71, p<0.001, and CT-LCEA: r_s=0.67, p<0.001) (Table 2). For the CT-ACEA and DRR-ACEA, significant moderate correlation was found between the DRR-ACEA and 3D anterior coverage (r_s=0.41, p=0.023), with a weak correlation between DRR-ACEA and 3D superior coverage (r_s=−0.36, p=0.033). No significant correlation was found between the CT-ACEA and each regional 3D coverage (Table 2).

Table 2.

Correlation between 2D measurements and regional 3D coverage preoperatively and following curved periacetabular osteotomy.

Measurement			Anterior	Superior	Posterior	Inferior
LCEA	Xray-LCEA	Pre-op	rs=0.03, p=0.839	rs=0.71, p<0.001*	rs=−0.18, p=0.418	rs=−0.16, p=0.419
		Post-op	rs=−0.08, p=0.604	rs=0.53, p=0.001*	rs=−0.30, p=0.083	rs=−0.28, p=0.083
	CT-LCEA	Pre-op	rs=0.15, p=0.342	rs=0.67, p<0.001*	rs=−0.38, p=0.021*	rs=−0.26, p=0.128
		Post-op	rs=−0.19, p=0.218	rs=0.62, p<0.001*	rs=−0.40, p=0.010*	rs=−0.41, p=0.010*
ACEA	CT-ACEA	Pre-op	rs=0.29, p=0.171	rs=0.25, p=0.171	rs=−0.24, p=0.171	rs=−0.12, p=0.434
		Post-op	rs=0.61, p<0.001*	rs=−0.57, p<0.001*	rs=−0.08, p=0.627	rs=0.18, p=0.338
	DRR-ACEA	Pre-op	rs=0.41, p=0.023*	rs=−0.36, p=0.033*	rs=−0.00, p=0.991	rs=0.17, p=0.364
		Post-op	rs=0.57, p<0.001*	rs=−0.52, p<0.001*	rs=0.01, p=0.969	rs=0.23, p=0.191

Pre-op: Preoperative, Post-op: postoperative

Statistical significance indicated by *

In the postoperative analysis, significant moderate to strong correlations were found between the LCEAs and 3D superior coverage (Xray-LCEA: r_s=0.53, p=0.001, and CT-LCEA: r_s=0.62, p<0.001) (Table 2). Further, significant moderate correlation was found between the DRR-ACEA and 3D anterior coverage and between DRR-ACEA and 3D superior coverage (r_s=0.57, p<0.001, and r_s=−0.52, p<0.01, respectively). For the CT-ACEA, correlation between 3D anterior coverage and 3D superior coverage was significant (r_s=0.61, p<0.001, and r_s=−0.57, p<0.001, respectively).

When changes in 2D parameters after CPO were correlated to the change in each regional 3D coverage, the change in the LCEAs significantly correlated to changes in 3D superior coverage (Xray-LCEA: r_s=0.67, and CT-LCEA: r_s=0.81) and 3D posterior coverage (Xray-LCEA: r_s=−0.56, and CT-LCEA, all p<0.001) (Table 3). For the change in CT-ACEA and DRR-ACEA, both were significantly related to the change in 3D anterior coverage (CT-ACEA: r_s=0.45, p=0.012 and DRR-ACEA: r_s=0.40, p=0.010).

Table 3.

Correlation coefficients between the change in 2D measurements and change in regional 3D coverage following curved periacetabular osteotomy.

Measurement		Anterior	Superior	Posterior	Inferior
LCEA	Xray-LCEA	rs=−0.16, p=0.312	rs=0.67, p<0.001*	rs=−0.56, p<0.001*	rs=−0.37, p=0.021*
	CT-LCEA	rs=−0.22, p=0.161	rs=0.81, p<0.001*	rs=−0.62, p<0.001*	rs=−0.36, p=0.024*
ACEA	CT-ACEA	rs=0.45, p=0.012*	rs=−0.33, p=0.077	rs=−0.09, p=0.698	rs=0.06, p=0.698
	DRR-ACEA	rs=0.40, p=0.010*	rs=−0.55, p<0.001*	rs=−0.02, p=0.898	rs=0.42, p=0.010*

Statistical significance indicated by *

Clinical assessment

The mean preoperative HHS and JOA H-S were 75±13 and 75±14, respectively. With a median follow-up of 3.9 years (3.0 – 8.0 years) after CPO, HHS and JOA H-S significantly improved to 97±5 and 97±6, respectively (both p<0.001). Progression of OA was not observed in any patient and no patient reported pain that would likely be due to impingement. All hips except for 1 hip (85°) were able to flex for more than 90°.

Intra- and inter-observer agreement in measuring the 2D parameters

The ICC ranged from 0.875 to 0.982 for intra-observer agreement and ranged from 0.721 to 0.962 for inter-observer agreement (Table 4). All ICCs indicated strong to very strong agreement.

Table 4.

Intra- and inter-observer agreement of 2D measurements as quantified by the intra class coefficient.

	Xray-LCEA	CT-LCEA	CT-ACEA	DRR-ACEA
Preoperative
Intra-observer	0.944 (0.868–0.977)	0.875 (0.716–0.948)	0.928 (0.830–0.971)	0.924 (0.822–0.969)
Inter-observer	0.906 (0.792–0.959)	0.913 (0.798–0.964)	0.727 (0.429–0.882)	0.741 (0.480–0.881)
Postoperative
Intra-observer	0.913 (0.797–0.964)	0.982 (0.957–0.993)	0.904 (0.777–0.960)	0.886 (0.738–0.953)
Inter-observer	0.774 (0.513–0.904)	0.962 (0.907–0.985)	0.721 (0.420–0.880)	0.748 (0.466–0.892)

Discussion

In this study, we found that all 2D measurements of head coverage increased significantly after CPO and that 3D coverage in the anterior and superior regions increased significantly, with a corresponding decrease in posterior and inferior regional coverage. Significant correlations were observed between the two LCEAs and 3D superior coverage and between the DRR-ACEA and 3D anterior coverage both pre- and postoperatively. However, the correlation between DRR-ACEA and 3D anterior coverage was only moderate. Further, no significant correlation was found between the CT-ACEA and 3D anterior coverage in the preoperative analysis. Collectively, these results indicate that both the DRR-ACEA and CT-ACEA should not be used to predict 3D preoperative anterior coverage for patients with DDH since large errors are likely to occur.

3D coverage in the anterior and superior regions increased after CPO to reach a postoperative value of 22.6% and 55.1%, respectively. These results are strikingly similar to a previous study that analyzed asymptomatic, morphologically-screened control subjects, where 24.4% of the anterior and 53.1% of the superior region of the femoral head was covered [22]. Therefore, CPO in our sample effectively normalized anterior and superior coverage, which lends confidence that the results presented herein can be interpreted with respect to cases of DDH where good surgical correction is achieved. This is further supported by the clinical outcomes evaluated by HHS and JOA score at the final follow-up, which were both 97 and indicated excellent clinical results.

When the two LCEAs were correlated to regional 3D coverage, a significant correlation was found between the 3D superior coverage in both the pre- and postoperative analysis. Furthermore, when the changes in the two LCEAs were correlated to the changes in regional 3D coverage, strong positive correlations were found for the superior region (Table 3). These results indicate that the LCEA based on X-ray or CT can be used to predict 3D coverage in the superior region. On the other hand, when the changes in the two LCEAs were correlated to the change in 3D posterior coverage, significant negative correlations were found. Such results indicate that lateral rotation of the acetabulum produces an increase to superior coverage, but does so at the expense of a reduction in posterior coverage; these findings are consistent with a previous study that reported deficiencies in posterior coverage after rotational acetabular osteotomy [4]. Collectively, our results suggest surgeons should exercise caution when basing the osteotomy correction off of an improvement in LCEA as it could result in a loss of posterior coverage, which is associated with a decrease in joint survivorship [1].

There were no significant correlations between the preoperative CT-ACEA measurements and 3D anterior coverage. This finding is important as it is a common practice to evaluate anterior coverage using the CT-ACEA in patients with a suspected diagnosis of DDH [18,19]. As the head center may have little lateral coverage in patients with DDH and because the CT-ACEA cannot be measured in all cases (in three cases with CT-LCEA <0°), we recommend that surgeons do not use the CT-ACEA to quantify anterior coverage in patients with DDH. On the other hand, a significant correlation was found between the DRR-ACEA and 3D anterior coverage in the preoperative analysis. Further, when the changes in the two ACEAs were correlated to the changes in regional 3D coverage, a positive correlation was found for the change in the anterior region (Table 3). Overall, analysis of the correlation between the two ACEAs and 3D anterior coverage indicates the possibility of using the ACEAs to quantify anterior coverage. However, it should be noted that each correlation was only moderate, indicating that both the CT-ACEA and DRR-ACEA are not an ideal parameter to evaluate anterior coverage.

In this study, the relationships between 2D coverage and 3D coverage were analyzed in both the preoperative and postoperative period. While the postoperative clinical results were excellent in a median follow-up of 3.9 years, it remains uncertain if surgical planning using information of preoperative 3D coverage is superior to that made using 2D measurements or if assessment of postoperative 3D coverage better predicts the clinical outcomes than the 2D measurements. In the future, we plan to analyze the relationship between postoperative 3D coverage and joint survivorship in a long term to clarify factors that lead to progression of osteoarthritis, which could then be used to improve clinical outcomes following periacetabular osteotomy.

Limitations

There are some limitations to this study. First, only female patients with DDH were included. Our study conclusions could change if men were evaluated since the morphology of the pelvis and femur differs between genders [8]. However, since the majority of DDH cases in Japan are female [27], we believe that our study provides relevant information. Second, variation in the positioning of the patient when acquiring images and variation in hip angles across postures may have affected the results. However, a previous study reported that an 1° change in hip flexion only leads to 0.5% change in 3D anterior coverage [23]. As the difference in hip angles between the preoperative CT images and postoperative CT images were no more than 5°, it is unlikely that intra-subject differences in patient positioning affected the major findings and conclusion of our study. Finally, while our measurement of the two ACEAs is based on the lateral view of the pelvis and is often used for clinical investigation [18–20,28], it is different from the original ACEA [6], which utilizes the false-profile view. In fact, previous research has found that the ACEA in the lateral view is 13.2° larger than that measured in the false-profile view [29]. However, the tight 95% confidence intervals (11.2–15.3°) indicate a consistent difference between the ACEAs in the two positions, which likely maintains the relationship between the ACEA and the 3D anterior coverage found in this study.

Conclusions

In this population of Japanese females with a diagnosis of DDH, CPO improved femoral head coverage as evaluated with both 2D and 3D measurements. Notably, anterior and superior coverage increased while posterior and inferior coverage decreased. There was a moderate to strong correlation between the LCEA and the 3D superior coverage in both the preoperative and the postoperative period. Further the correlations between the changes in LCEA and change in 3D superior coverage were strong to very strong. Collectively, results suggest that the LCEA may be used for the evaluation of superior coverage. On the other hand, no significant correlation was found between the CT-ACEA and 3D anterior coverage in the preoperative state. Further, the correlation between DRR-ACEA and 3D anterior coverage was only moderate. Thus, we suggest that the ACEA not be used to evaluate/estimate anterior coverage in patients with DDH.