Abstract
Background:
The orientation of the acetabulum in the axial and coronal planes is well studied in the pathogenesis of impingement and instability of the hip. In contrast, the sagittal orientation of the acetabulum (SOA) is not well understood.
Purpose:
To determine (1) the SOA in a large cohort of mature hips and (2) to assess the relationship between the SOA and acetabular version, acetabular center-edge angles (CEAs), and spinopelvic alignment.
Study Design:
Descriptive laboratory study.
Methods:
A total of 3695 patients (7390 mature hips) who underwent computed tomography (CT) scans for assessment of nonorthopaedic abdominal and pelvic conditions were studied. An automated measurement software was utilized to reconstruct 3-dimensional models from CT scans and to measure the SOA, functional SOA (not neutralizing pelvic position on sagittal plane), acetabular version, as well as acetabular CEAs and spinopelvic alignment, including the pelvic tilt (PT), sacral slope (SS), and pelvic incidence (PI).
Results:
The SOA was on average (± SD) 19.6°± 7.5°. The functional SOA (not neutralizing pelvic position on sagittal plane) was on average (± SD) 20.5°± 5.7°. The functional SOA had a statistically significant but negligible correlation with PI (r = 0.13; P < .001) and SS (r = −0.06; P < .001), and a weak positive correlation with PT (r = 0.23; P < .001). The SOA had a positive moderate correlation with the cranial (r = 0.41; P < .001) and central acetabular version (r = 0.39; P < .001) and a strong correlation (r = 0.63; P < .001) with caudal acetabular version. A 10° increase in SOA was associated with a 6.6° increase on the caudal acetabular version. The SOA had a moderate negative correlation (r = −0.48; P < .001) with the CEA at 3 o'clock (anterior for left and right hips). A 10° increase in SOA was associated with a 4.9° decrease in CEA at 3 o'clock.
Conclusion:
The acetabulum is on average 19.5° cephalically oriented in the sagittal plane in asymptomatic individuals. The SOA correlates with acetabular version and cannot be presumed based on spinopelvic alignment.
Clinical Relevance:
The assessment of the SOA may aid in the diagnosis of hip impingement and instability, allowing a more precise correction of the acetabulum in hip arthroscopy and osteotomies.
Keywords
The orientation of the acetabulum in the axial and coronal planes is fundamental in the pathogenesis of impingement and instability of the hip. The acetabular version represents the acetabular orientation in the axial plane, while the center-edge angle (CEA) and acetabular inclination are the most used parameters to assess the orientation of the acetabulum in the coronal plane.13,25,27 The measurement and applications of those parameters are widely reported and utilized in clinical practice.1,18 Conversely, the positioning of the acetabulum in the sagittal plane is a sporadic subject of research.5,8,11,17 Even reports on redirectional acetabular osteotomies usually ignore their effects in the sagittal orientation of the acetabulum (SOA).3,7,15,23 A few cadaveric or 3-dimensional (3D) computed tomography (CT) studies are reported to determine the SOA, utilizing the acetabular notch or the distal ends of the anterior and posterior acetabular horns as the distal reference for measurement.5,9,11,17
The influence of spinopelvic parameters and native hip biomechanics has been increasingly studied in the past decade.2,20 The SOA is often assumed to be predominantly determined by the lumbosacral positioning—that is, spinopelvic alignment including the pelvic tilt (PT), sacral slope (SS), and pelvic incidence (PI). This assumption may be inappropriate for patients without severe lumbosacral deformity, for whom the SOA may be an intrinsic feature, similarly to the orientation in the axial (acetabular version) and coronal (lateral coverage) planes.
The main purpose of this study was to determine the SOA in a large cohort of asymptomatic adult hips. A secondary purpose was to assess the relationship between the SOA and the acetabular version, acetabular lateral CEAs, and spinopelvic alignment as assessed by the PT, SS, and PI.
Methods
Study Population
The present study was performed at a tertiary hospital after institutional review board approval. We first identified all the patients who underwent abdominal, hip, or pelvic CT scans from 2004 to 2022, resulting in 52,360 scans. Using a validated natural language processing pipeline, 24 we systematically surveyed medical records, including clinical notes and radiology reports, to exclude patients with any documented musculoskeletal pathology involving the pelvis, the proximal femur, or low-quality images, which resulted in 12,055 CT scans. Most of the CT scans included in the study were performed as part of the diagnostic workup for abdominal pain, with suspected appendicitis being the most common diagnosis. We then narrowed our analysis to those aged ≥16 years at the time of the CT scan (3695 patients and 7390 hips).
CT Segmentation, Automated Landmark Detection, and Measurement
Selected CT scans were then imported to a custom-developed validated program14,16,22 for 3D assessment of acetabular morphology. The program uses a hierarchal UNet-based convolutional neural network (CNN) to segment the pelvic and femoral bones to identify major anatomical landmarks including the femoral head, femoral neck, and acetabulum. The CNN was trained and tested on manually segmented clinical CT scans (n = 424; 80% training; 20% testing) with varying degrees of bone deformity (eg, femoroacetabular impingement, developmental dysplasia of the hip, slipped capital femoral epiphysis, Perthes disease, and cerebral palsy) from a wide range of males and females at different stages of skeletal maturity (mean ± SD age, 17 ± 9 [range, 2-61]; 48% female; 22% without hip pathology). The CT images for validation were randomly selected from our institutional radiology database, including images obtained between 2000 and 2022.
The CNNs were then supplemented by a custom-developed postprocessing step to refine the segmentation by removing random components. The final models were exported as fine surface mesh (ie, each model was defined by a set of faces and vertices) to be used for anatomic measurements. The dice coefficients for automatic segmentation and landmark detection on the test set (which included a broad spectrum of cases, both in terms of patient demographics and clinical conditions) were 0.98 (pelvis) and 0.96 (acetabulum). The manual segmentations were performed by professionals with knowledge of human anatomy familiar with various hip conditions and pathologies.
The models were rotated around the z-axis such that the line connected pubic tubercles, and anterosuperior iliac spines centers aligned with the y-axis (Figure 1). All measurements, except for functional SOA, were calculated after the models were aligned. For the functional SOA, the last rotation around the z-axis was excluded, mimicking the hip position during CT acquisition. The developed 3D models were then used to automatically measure the SOA, acetabular CEAs, acetabular version, PI, PT, and SS.
(A) The anterior pelvic plane (APP) is represented by the dashed shadow, defined by the anterosuperior iliac spines (ASIS; blue dots) and most anterior pubis (red dots). The nondashed shadow represents a plane perpendicular to the APP. (B) The sagittal orientation of the acetabulum (SOA) was defined as the angle between the plane perpendicular (blue line) to the APP (black line), and a line running from the distal limit of the anterior acetabular horn to the distal limit of the posterior acetabular horn (red line). (C) The function SOA was defined as the angle between the plane perpendicular (blue line) to the CT acquisition table (black line), and a line running from the distal limit of the anterior acetabular horn to the distal limit of the posterior acetabular horn (red line).
SOA and Functional SOA
The SOA was defined as the angle between 2 references: (1) a line running from the distal limit of the anterior acetabular horn to the posterior acetabular horn and (2) the plane perpendicular to the anterior pelvic plane (APP), defined by the right and left anterosuperior iliac spines and most anterior pubis) (Figure 1, A and B). The utilization of the APP as reference for the SOA could hinder the detection of the spinopelvic alignment influence in the acetabulum. Therefore, the functional SOA was also measured (Figure 1C). The functional SOA was defined as the angle between 2 references: (1) a line running from the distal limit of the anterior acetabular horn to the posterior acetabular horn and (2) the plane perpendicular to the CT acquisition table. The cephalic orientation of the acetabulum in the sagittal plane was considered a positive SOA. Considering that measurement of SOA has not been reported in 3D-reconstructed images in previous studies, additional validation was performed for the SOA. The reference points to measure the SOA were manually selected in segmented 3D models by an orthopaedic hip surgeon (M.A.H.) with 12 years of experience in 76 hips.
Acetabular Version and CEAs
A clockface reference 19 was utilized for automated measurements of the acetabular version and acetabular CEAs. To define 12 o'clock, the acetabular axis was defined based on the fitted cylinder to the side surface of the pelvic acetabular model. A plane was then defined using the axis and a point above (on the superior in the positive y direction) of the axis. The intersection of this plane and the pelvic acetabular model resulted in a closed curve. The most lateral point on the superior side of the curve was 12 o'clock. Other clocks from 1 to 11 o'clock were defined by 30°-interval rotations starting from 12 o'clock. The direction of rotation was such that 3’oclock was at the anterior regardless of the hip side. At each clock, the CEA was defined as the angle between the line connecting the center of the femoral head to the acetabular clock face and its projection on the xy plane starting from the femoral head center. The angle was calculated on the plane of the corresponding clock, which contained both lines (Figure 2). The acetabular version was defined as the angle between a line connecting the posteroanterior acetabular clock face and the line perpendicular to the coronal plane passing through the femoral head center in the axial view (Figure 2). The acetabular version was measured at 3 levels: cranial (1-11 o’clock), central (3-9 o'clock), and caudal (5-7 o'clock). The acetabular CEA was measured at 3 (anterior), 9 (posterior), and 12 o'clock (superior).
Clockface-based measurement of acetabular center-edge angle and version.
Spinopelvic Parameters
The SS, PI, and PT were measured in the sagittal plane after the alignment based on the APP. The SS was represented by the angle between the anteroposterior direction (x) and a line tangent to the superior endplate of S1. The PI was the angle between the perpendicular to the sacral plate at its midpoint and the line connecting this point to the bicoxofemoral axis. The PT was the angle between the inferosuperior (y) direction and the line joining the middle of the sacral plate and the bicoxofemoral axis (Figure 3).
Measurements of spinopelvic parameters. PI, pelvic incidence; PT, pelvic tilt; SS, sacral slope.
Statistical and Reliability Analysis
To assess the bivariate correlations between SOA as well as functional SOA and radiographic parameters of hip morphology or spinopelvic alignment, Pearson correlation coefficients were calculated. Scatterplots were generated to visually inspect the relationships. For variables demonstrating moderate correlations (r≥ 0.4), univariate and multivariable linear regression models were constructed to quantify the associations, adjusting for potential confounders. Covariates in the models included age, sex, and body mass index (BMI) group. Regression coefficients, 95% CIs, and P values are reported to quantify the strength of the associations.
To consider the effects of the spinopelvic parameters on the acetabular positioning, the functional SOA was used for correlation with the spinopelvic parameters instead of SOA. Different from SOA, the functional SOA was determined, not neutralizing pelvic position on sagittal plane.
Interrater reliability between the orthopaedic surgeon and the software-measured SOA and APP measurements was assessed using a 2-way mixed-effects model, absolute agreement, single-measurement model. Intraclass correlation coefficients (ICCs) were calculated along with 95% CIs to quantify the level of agreement between raters. Additionally, Bland-Altman plots were constructed to visually inspect variability between the orthopaedic surgeon and the software for both SOA and APP measurements. The 95% limits of agreement were calculated as the mean difference between raters ± 1.96 times the standard deviation of the differences. The ICC demonstrated excellent interrater reliability between the orthopaedic surgeon and the software for measuring SOA (ICC, 0.87; 95% CI, 0.80-0.92) and APP (ICC, 0.98; 95% CI, 0.96-0.98). Bland-Altman analysis further showed strong agreement between raters for both SOA and APP, with nearly all data points falling within the limits of agreement. The analysis was conducted using R Version 4.3.2.
Results
In total, 3695 patients (7390 hips) with a mean (± SD) age of 18.4 ± 2.4 years were included (1862 [50.4%] female and 1833 male). Except for 1 patient (2 hips), all hips had a cephalic orientation in the sagittal plane (distal end of the anterior acetabular horn more proximal than the distal end of the posterior acetabular horn). The SOA was on average (± SD) 19.6°± 7.5°.
The functional SOA (not neutralizing pelvic position on sagittal plane) was a mean (± SD) of 20.5°± 5.7°. Table 1 summarizes the morphologic parameters for the entire sample. The SOA was a mean of 18.8°± 7.8° in females and 20.3°± 6.5° in males (P < .001). The functional SOA was a mean of 20.9°± 5.5° in females and 20.1°± 5.5° in males.
Summary of Morphologic Parameters for Entire Sample of 3695 Patients (7390 hips) a
Data are presented in degrees. CEA, acetabular center-edge angle; SOA, sagittal orientation of acetabulum.
Positive value for the anterior pelvic plane indicates that the pubic anterior surface is anterior to a line between right and left anterosuperior iliac spines. A negative value for the anterior pelvic plane indicates that the pubic anterior surface is posterior to a line between right and left anterosuperior iliac spines.
The 3 o'clock position corresponded to the anterior acetabulum in right and left hips.
The 9 o'clock position corresponded to the posterior acetabulum in right and left hips.
The functional SOA had negligible correlation with PI (r = 0.13; P < .001) and SS (r = −0.06; P < .001), and weak positive correlation with PT (r = 0.23; P < .001) (Figure 4).
Scatterplot illustrates the relationship between (A) functional SOA and pelvic incidence, (B) functional SOA and sacral slope, and (C) functional SOA and spinopelvic tilt. Pearson correlation coefficients and P values are displayed. Solid red lines show the fitted linear model.
The SOA had a positive moderate correlation with the cranial (1-11 o’clock) acetabular version (r = 0.41; P < .001) and central (3-9 o’clock) acetabular version (r = 0.39; P < .001), and a strong correlation (r = 0.63; P < .001) with caudal (5-7 o’clock) acetabular version, meaning that as the SOA increased, the acetabular version increased, particularly more distally in the acetabulum (Figure 5). A 10° increase in SOA was associated with a 4.1° (95% CI, 3.7°-4.4°; P < .001) increase in acetabular version at 1 to 11 o'clock, a 3° (95% CI, 2.8°-3.3°; P < .001) increase at 3 to 9 o'clock, and a 6.6° (95% CI, 6.3°-6.9°; P < .001) increase at 5 to 7 o'clock after adjusting for age, sex, and BMI.
Scatterplot illustrates the relationship between sagittal orientation of the acetabulum and acetabular version at 1-11 o'clock, 3-9 o'clock, and 5-7 o'clock. Pearson correlation coefficients and P values are displayed. Solid red lines show the fitted linear model.
The SOA had a moderate negative correlation (r = −0.48; P < .001) with the CEA at 3 o'clock (anterior for left and right hip), meaning that as the SOA increased, the anterior acetabular coverage of the femoral head decreased. A 10° increase in SOA was associated with a 4.9° (95% CI, –5.2° to −4.6°; P < .001) decrease in CEA at 3 o'clock after controlling for age, sex, and BMI. The SOA had a negligible correlation (r = 0.11; P < .001) with the CEA at 9 o'clock (posterior for left and right hip) and a negligible correlation with the CEA at 12 o'clock (r = −0.15; P < .001) (Figure 6).
Scatterplot illustrates the relationship between sagittal orientation of the acetabulum and center-edge angle (CEA) at 3 o'clock, 9 o'clock, and 12 o'clock. Pearson correlation coefficients and P values are displayed. Solid red lines show the fitted linear model.
Discussion
The current study, including 7390 mature hips (3695 patients), has demonstrated the acetabulum to be on average 19.6° cephalically oriented in the sagittal plane. The functional SOA had negligible correlation with PI (r = 0.13; P < .0001) and SS (r = −0.06; P < .0001) and weak positive correlation with PT (r = 0.23; P < .001). These findings indicate that the SOA, when comparing different individuals in the same body posture, is essentially an intrinsic feature of the native hip.
The orientation of the acetabulum in the sagittal plane has been the subject of a limited number of publications. Oberländer et al 17 and Köhnlein et al 11 reported the acetabular tilt to assess the sagittal plane orientation of the acetabulum in 30 and 104 cadaveric hip specimens, respectively. The acetabular tilt represents the angle between the acetabulum and the frontal plane and is traditionally determined utilizing a meridional line from 12 o'clock to the middle of the acetabular notch (Figure 7A). The mean SOA observed in the present study (19.5° cephalic) was similar to the acetabular tilt described by Oberländer et al (18.3°) and Köhnlein et al (18.9°). In another study, Fuji et al 5 utilized the center of the acetabulum as proximal reference in 3D-reconstructed CT images and reported a mean acetabular tilt of 21° in 40 nondysplastic hips and 25° in 72 dysplastic hips (Figure 7B). In contrast to the studies by Oberländer et al, Köhnlein et al, and Fuji et al, the distal end of the articular surface of the acetabular horns was used to assess the SOA in the present study (Figure 7C). We consider it important to include the distal ends of the acetabular horns as reference, as the horns are articulating to the femoral head providing stability, in contrast with the acetabular notch. Hatem et al 9 assessed the SOA in 300 cadavers using the distal end of the acetabular horns and reported a mean cephalic angle of 25°, compared to 20° in the present study. This difference could be explained by limitations of measurement in photographs in which the posterior and anterior horns are not at the same level or because of changes in pelvic position after reassembling pelvises in the cadaveric study. 9 Another study has utilized the transverse ligament of the acetabulum, which spans the acetabular notch and articulates with the femoral head, to assess the SOA in 94 magnetic resonance arthrograms of symptomatic hips and reported a mean cephalic orientation of 18°. 8 Despite the differences in methods to assess the SOA, this study's use of a large cohort of asymptomatic patients corroborates the findings of previous studies that were limited by the number of patients showing the acetabulum to be around 20° cephalically oriented.
Methods to assess the orientation of the acetabulum in the sagittal plane. (A) Angle between the frontal plane (blue line) and a line from the 12 o'clock position to the middle of the acetabular notch (red line). 17 (B) Angle between the frontal plane (blue line) and a line from the center of the acetabulum to the middle of the acetabular notch (red line). 5 (C) Angle between the axial plane (blue line) and the distal limit of the anterior and posterior acetabular horns (arrows, red line). 9
The present study was based on automated measurement of 3D CT images, which allowed the utilization of the large sample of 7390 mature hips. Considering that most clinicians do not have access to automated measurements yet, manual measurement of the SOA is possible utilizing 3D-reconstructed images of CT scans. Measurement of SOA would be possible in a lateral view of the 3D-reconstructed CT in which the acetabular image is only rotated in the axial plane (not in the sagittal or coronal plane) to provide an image such as that in Figure 7. However, the process of rotating the 3D image axially would possibly be subject to the human error of changing the sagittal orientation, leading to an unreliable measurement. Therefore, further studies are needed to assess the reliability of manually measured SOA when compared with the automated method presented in the present study. Another method described to assess the acetabular orientation on the sagittal plane involves MRI, utilizing the transverse ligament of the acetabulum as reference since it runs from the distal limit of the anterior acetabular horn to the distal limit of the posterior acetabular horn 8 . However, utilizing the transverse ligament as reference in MRI is particularly challenging in acetabula with increased acetabular version or moderate to severe dysplasia due to the difficulty in obtaining a sagittal magnetic resonance imaging (MRI) scan along the axis of the transverse ligament.
The correlation observed in the present study between the SOA and the acetabular version indicated that changes in the SOA affected the orientation of the acetabulum in the axial plane. As the SOA increased, the acetabular version increased, particularly more distally in the acetabulum. A 10° increase in SOA was associated with a 6.6° (95% CI, 6.3°-6.9°; P < .001) increase in the acetabular version at 5 to 7 o'clock after adjusting for age, sex, and BMI.
The SOA was also correlated with the CEA. A 10° increase in SOA was associated with a 4.9° (95% CI, –5.2° to −4.6°; P < .001) decrease in CEA at 3 o'clock (anterior for left and right hips) after controlling for age, sex, and BMI, meaning that as the SOA increased, the anterior acetabular coverage of the femoral head decreased. Further studies on the SOA in symptomatic hips with impingement and instability will be helpful in understanding the implications of the SOA when planning for hip preservation surgeries.
The SOA was measured in a neutralized pelvic position with the right and left anterosuperior iliac spines and pubic tubercles on the same horizontal plane. In contrast, the functional SOA was measured in a nonneutralized pelvis to assess potential effects of the lumbosacral positioning in the acetabulum. However, in patients with lumbosacral deformity or arthrodesis, the spinopelvic parameters could affect SOA more. The values for mean functional SOA (20.5°) were very close to SOA (19.6°), but that may not be reflected in individual patients, particularly older patients and those with spinal deformities. The position in which the patients underwent the CT studies should also be considered when interpreting the results of the current study. Our findings were based on CT obtained with the patient in a supine position, and the correlation between the spinopelvic parameters and the SOA could be different in a standing position. However, the mean SS (38.1°± 7.7°) and PT (10.7°± 6.3°) in the present study differed <3° from those described in publications in which the measurements were performed in standing radiographs.4,6,12,21,26 In addition, a previous study has shown that spinopelvic sagittal parameters could be measured with high reliability on supine CT when compared with standing radiographs. 10 However, when patients with spinal or pelvic pathology are considered individually, changing from supine to standing position could have implications on the SOA as a consequence of pelvic change in positioning. For example, if lumbar lordosis and SS increase significantly from supine to standing position in a patient with neurologic disease, the SOA would proportionally decrease (the hip would be more stable anteroinferiorly and less stable posteroinferiorly in the standing position compared with supine). The results of the present study cannot be extrapolated to the sitting position, considering that changing from a supine or standing position to a sitting position modifies the SS and PT in about 50% and 200%, respectively.4,12
The SOA may have significant clinical and surgical repercussions in hip pathology. A more cephalic SOA results in reduced anteroinferior femoral head coverage and increased posteroinferior coverage, predisposing patients to anteroinferior hip instability. Conversely, a less cephalic SOA, associated with reduced femoral torsion, increases the risk of posteroinferior instability. Additionally, a decreased SOA predisposes to anterosuperior femoroacetabular impingement. Acetabula with less cephalic orientation may contribute to the crossover sign observed in pincer-type femoroacetabular impingement. Periacetabular osteotomy (PAO) has the potential to alter the acetabular orientation in the coronal, axial, and sagittal planes. However, the SOA is not currently controlled on PAOs, and preoperative assessment of SOA using CT or MRI could enhance surgical precision and achievement of a more normal anatomy. In patients with significant difference between the SOA and functional SOA, planning the PAO based on the functional SOA would provide a more reliable acetabular coverage in terms of treating hip instability and preventing secondary impingement. Evaluating SOA may also be valuable prior to lumbar spinal surgeries involving fusion or deformity correction. Notably, we have observed patients with acetabula exhibiting cephalic orientation exceeding 30° and adopting a trunk flexion posture to enhance anterior femoral head coverage, particularly when associated with increased femoral torsion.
Limitations
This study presents limitations. First, the CT scans were conducted in the supine position, which may result in a SOA that differs from that in a standing position, as discussed previously. Second, the imaging studies were not specifically performed for the hip joint, and CT scans focused on the hip joint with thinner slices could potentially yield more accurate images for automated measurements. Third, the manual measurement of the SOA is possible, but would require adequate 3D reconstruction for sagittal view with controlled pelvic positioning, which is not routinely performed in clinical practice.
Conclusion
The acetabulum is on average 19.5° cephalically oriented in the sagittal plane in asymptomatic individuals. The SOA correlates with acetabular version and cannot be presumed based on spinopelvic alignment.
Footnotes
Final revision submitted September 12, 2025; accepted September 15, 2025.
The authors declared that there are no conflicts of interest in the authorship and publication of this contribution. AOSSM checks author disclosures against the Open Payments Database (OPD). AOSSM has not conducted an independent investigation on the OPD and disclaims any liability or responsibility relating thereto.
Ethical approval was provided by the Institutional Review Board at Boston Children's Hospital, IRB-P00046914.
