Accuracy of Planar Anteversion Measurements in Anteroposterior Radiographs

Introduction Number of methods has been suggested to measure anteversion in the simple AP radiograph and a few studies have validated their accuracy and reliability. In the current study, we compared six widely used anteversion measurement methods using two different radiographs (conventional pelvis AP vs hip centered AP) in order to determine which measurement method and which radiograph provides highest accuracy and reliability. Material and methods We developed custom made two planar anteversion measurement models for the validation. Each models are designed for Pelvis AP and hip AP. The radiographs were taken with the both models’ inclination changing from 10° to 70° at 10° increments and for each inclination angles, anteversion was changed from 0° to 30° at 5° increments. Therefore, the x-ray of the two models were taken in 49 scenarios. The measurements were done independently by two orthopaedic surgeons blinded from each other using 6 methods : 1) Pradhan et al, 2) Lewinak et al, 3) Widmer et al, and 4) Liaw et al. 5) Hassan et al, 6) Ackland et al. Result The highest accuracy was found when Liaw method was used in hip centered AP radiograph which showed difference of 1.37° ± 1.73. Also, regardless of the type of the radiographs, Pradhan, Lewinnek, Liaw all showed relatively high accuracy. However, substantial difference was found when Widmer, Hassan, Ackland method were utilized regardless of the type of radiograph used. When the anteversion were measured in the inclination between 30° and 50°, Pradhan’s method in pelvis AP showed the highest accuracy (1.23° ± 0.92°). Also, we found no significant difference when measured anteversions were compared between Pelvis AP and hip centered AP. Conclusion The study indicates that the Pradhan, Liaw and Lewinnek method may provide the relatively accurate anteversion measurement regardless of the type of the radiographs. Also, it would be unnecessary to take the hip centered AP radiographs in addition to the pelvis AP radiographs for the purpose of measuring anteversion.

changing from 10° to 70° at 10° increments and for each inclination angles, anteversion was changed from 0° to 30° at 5° increments. Therefore, the x-ray of the two models were taken in 49 scenarios.
The measurements were done independently by two orthopaedic surgeons blinded from each other centered AP radiograph which showed difference of 1.37° ± 1.73. Also, regardless of the type of the radiographs, Pradhan, Lewinnek, Liaw all showed relatively high accuracy. However, substantial difference was found when Widmer, Hassan, Ackland method were utilized regardless of the type of radiograph used. When the anteversion were measured in the inclination between 30° and 50°, Pradhan's method in pelvis AP showed the highest accuracy (1.23° ± 0.92°). Also, we found no significant difference when measured anteversions were compared between Pelvis AP and hip centered AP. Conclusion The study indicates that the Pradhan, Liaw and Lewinnek method may provide the relatively accurate anteversion measurement regardless of the type of the radiographs.
Also, it would be unnecessary to take the hip centered AP radiographs in addition to the pelvis AP radiographs for the purpose of measuring anteversion.

Background
Accurate positioning of the implanted prosthesis after total hip replacement (THR) is critical to achieving an optimal outcome as the postoperative complications including polyethylene liner wear, impingement, and instability are largely attributed to malposition of the acetabular cup (1)(2)(3)(4)(5). The 3 appropriateness of the acetabular cup position is determined by measuring inclination and anteversion. While measurement of inclination is relatively straightforward and can be done using a simple pelvis anteroposterior (AP) radiographs, controversies remain regarding the measurement of the anteversion of the acetabular component.
Measurement of anteversion using cross-table hip lateral view is one of the most commonly used methods (2,6,7). However, this method could provide inaccurate measurement in patients with joint contracture or lumbar stiffness or if the hip lateral radiograph is inadequately taken (8-10). As such, a number of methods have been suggested to measure anteversion in the simple AP radiograph and several studies have validated their accuracy and reliability (11)(12)(13). However, which anteversion measuring formula provides the most accurate anteversion measurement remains controversial.
The current study compared six widely used anteversion measurement methods using two different radiographs (conventional pelvis AP vs. hip-centered AP) to determine which measurement method and which radiograph provided the highest accuracy and reliability. We developed a custom-made planar anteversion measurement model for validation. to the plexiglass plate 9.9 cm lateral and 4.9 cm distal to the first model. The distance between the two models represented the distance from the center of the triangle formed by the anterior superior iliac spine (ASIS) and symphysis pubis, which is typically used for conventional pelvis AP radiographs, and the hip joint. Thus, an X-ray beam directed toward Model A represented the simple X-ray of hipcentered AP radiograph, while the image in Model B represented the acetabular cup in conventional pelvis AP radiographs. The radiographs were taken with both models' inclinations changing from 10° 4 to 70° at 10° increments; for each inclination angle, the anteversion was changed from 0° to 30° at 5°i ncrements. Therefore, X-rays of the two models were taken in 49 scenarios.
All images were digitally acquired using a Picture Archiving and Communication System (INFINITT PACS system, Seoul, South Korea) and all measurements on radiographs were subsequently made using PACS software. The measurements were done independently by two orthopedic surgeons blinded from each other using the six methods described by Pradhan et al (14) In which D is the distance of the long axis of the acetabular component and x is the distance along the line AB. An arbitrary tangent is drawn at a right angle to the diameter, and y is the distance from the two-cup rims along this tangent.

Statistics
Reliability was defined as the consistency of the measurements, while accuracy was defined as the proximity to the reference anteversion angle. The reference anteversion was defined as the anteversion measured by the protractor of the manufactured model. Reliability tests were performed on interobserver and intraobserver measurements and intraclass correlations (ICCs) were calculated using JMP software (SAS Institute, Cary, NC). For accuracy determination, the differences in anteversion measurements from each method from the reference anteversion measurement were analyzed using paired t-tests with statistical significance set at p < 0.05.
To prevent bias from acquiring outlier data, subset analysis for accuracy of anteversion was performed for the anteversions within a safe inclination zone (30 to 50°).

Results
Interobserver and intraobserver correlations (ICC) were high in all measurements, ranging from 0.886 to 0.938 and 0.899 to 0.934, respectively ( Table 1). The highest accuracy was observed for the Liaw method in hip-centered AP radiographs, which showed a difference of 1.37° ± 1.73 from the reference. Also, regardless of the radiograph type, the Pradhan, Lewinnek, and Liaw methods showed relatively high accuracies. However, substantial differences were observed for the Widmer, Hassan, and Ackland methods regardless of the radiograph type (Table 2). When the anteversions were measured in the inclination between 30 and 50°, Pradhan's method in pelvis AP showed the highest 6 accuracy (1.23° ± 0.92°) ( Table 3). We observed no significant differences in the measured anteversions between pelvis and hip-centered AP radiographs.

Discussion
Previous studies had limitations in their validations of anteversion measurement methods. First, many of the previous studies used computed tomography (CT) axial scans to validate the anteversion measured in plain anteroposterior radiograph (19)(20)(21)(22)(23). However, this may cause potential bias as the anteversion measured in CT scans reflects anatomical anteversion while the reference anteversion utilized in previous anteversion measurement methods using simple radiograph varied. Also, depending on how the simple AP radiograph is taken, the measured anteversion value may be significantly different. In simple pelvis AP radiographs, the radiation beam is projected toward the center of the triangle formed by the ASIS and the symphysis pubis. Thus, the radiation beam received by the hip joint in simple pelvis AP radiograph is deviated by about 6°. In contrast, the hip joint receives a perpendicular radiation beam in hip-centered AP radiographs.
The results of our study indicated that the anteversion measurement methods described by Pradhan, Lewinnek, and Liaw showed anteversion that differed by less than 4° from the real anteversion, while the methods proposed by Widmer, Hassan, and Ackland showed discrepancies of up to 29°. Moreover, measurement using hip-centered AP did not show superior accuracy compared to conventional AP; therefore, additional hip-centered AP radiographs are not recommended to improve the accuracy of anteversion measurement.
The correct positioning of the acetabular cup plays a significant role in minimizing wear and maintaining stability. The recommended inclination and version of the acetabular cup are controversial but the literature often refers to inclinations between 30 and 50° and anteversion between 5 and 30° as "safe zones" (24)(25)(26)(27).
The inclination angle of the acetabular cup is defined as the angle formed by the tear drop line and the elliptical long axis of the entrance of the acetabular cup; thus, the inclination angle can be directly measured by simple radiography (15). However, anteversion of the acetabular component may be less accurate and difficult to obtain with two-dimensional simple radiographs (11,13). Therefore, a number of studies have attempted to accurately measure the position of the acetabular components in plain AP radiographs using complex mathematics and trigonometric functions to describe the ellipses of the acetabular cup boundaries (1,(14)(15)(16)(17)(18). However, most of these studies did not specify where the center of the beam was directed or the reference plane that was utilized (Table 4). Several studies have tried to validate the accuracy and reliability of these methods but have reported inconsistent results (11)(12)(13). Marx  Pradhan, Woo, and Morrey), concluding that the values from Widmer's method were most similar to those measured using CT (12). It should be noted that study by Nomura et al is the only study to utilize the functional coronal plane as a reference while the other two studies used CT axial scans to measure anteversion (11). Compared to the previous studies, our results show that, while all measurements had high reliability, the accuracy was high only in the Pradhan, Lewinnek, and Liaw methods. We observed significant differences from the reference anteversion values for the Widmer, Hassan, and Ackland methods. We were unable to identify the reasons why the result of our study differed from those of the previous studies; however, we believe that we added precision to the reference anteversion values by adding a goniometer to the model.
We also found no significant difference in comparisons of anteversion measurements between conventional pelvis and hip-centered AP views. Depending on the measurement method, one radiograph measurement tended to be closer in value to the reference; however, the difference was minimal. We believe that this difference is based on which reference radiograph was used when the formula was first developed.
Another finding of our study was that the anteversion measurements tended to be closer to the reference anteversion in the inclination between 30° to 50°. As most of the cup during total hip 8 arthroplasty procedure is targeted in this range, the Liaw, Pradhan, and Lewinnek methods can be used with relatively high accuracy if the inclination is not excessively malpositioned.
Compared to the previous literature, our study showed a very high ICC. We believe this is because we did not utilize an in vivo model. As our model did not include soft tissue or a metal femoral head, we were able to more accurately identify the boundaries of the reference variables. Also, unlike taking Xrays in real humans, it can be hypothesized that the radiograph was taken with more accuracy.
However, we also acknowledge that this may also be a potential limitation of the current study since the accuracy may be lower in real total hip arthroplasty scenarios due to the interference of the soft tissue of the metallic head, which may result in haziness in the radiograph. Another limitation of the current study was that we used only one type of cup. Thus, our results may only be applied to a cup which is perfectly hemispherical.

Conclusions
The results of the present study indicated that the Pradhan, Liaw, and Lewinnek methods may provide relatively accurate anteversion measurement regardless of radiograph type. Furthermore, it may be unnecessary to obtain hip-centered AP radiographs in addition to pelvis AP radiographs to measure anteversion. SD standard deviation Table 3. Differences between measured and reference anteversions in pelvis and hip-centered Two custom-made anteversion-measuring models fixed to plexiglass. It represents the distance from the center of the triangle formed by both anterior superior iliac spine (ASIS) and symphysis pubis and the hip joint 16 Figure 3 Methods for measuring anteversion on plane anteroposterior radiographs. A) Pradhan, B)