Accuracy of conventional motion capture in measuring hip joint center location and hip rotations during gait, squat, and step-up activities

Accurate measurements of hip joint kinematics are essential for improving our understanding of the effects of injury, disease, and surgical intervention on long-term hip joint health. This study assessed the accuracy of conventional motion capture (MoCap) for measuring hip joint center (HJC) location and hip joint angles during gait, squat, and step-up activities while using dynamic biplane radiography (DBR) as the reference standard. Twenty-four young adults performed six trials of treadmill walking, six body-weight squats, and six step-ups within a biplane radiography system. Synchronized biplane radiographs were collected at 50 images per second and MoCap was collected simultaneously at 100 images per second. Bone motion during each activity was determined by matching digitally reconstructed radiographs, created from subject-specific CT-based bone models, to the biplane radiographs using a validated registration process. Errors in estimating HJC location and hip angles using MoCap were quantified by the root mean squared error (RMSE) across all frames of available data. The MoCap error in estimating HJC location was larger during step-up (up to 89.3 mm) than during gait (up to 16.6 mm) or squat (up to 31.4 mm) in all three anatomic directions (all p < 0.001). RMSE in hip joint flexion (7.2°) and abduction (4.3°) during gait was less than during squat (23.8° and 8.9°) and step-up (20.1° and 10.6°) (all p < 0.01). Clinical analysis and computational models that rely on skin-mounted markers to estimate hip kinematics should be interpreted with caution, especially during activities that involve deeper hip flexion.

Predictive estimation of ovine hip joint centers: A regression approach

Estimation of the hip joint center in ovine biomechanical analysis is often overlooked or estimated using a marker on the greater trochanter which can result in large errors that propagate through subsequent analyses. The purpose of this study was to develop a novel method of estimating the hip joint centers in sheep to facilitate more accurate analysis of ovine biomechanics. CT scans from 16 sheep of varying ages, weight, sex, and phenotypes were acquired and the data was used to calculate the known hip joint center by sphere fitting the femoral head. Anatomical measurements and additional subject information were used to create a variety of regression models to estimate the hip joint centers in absence of CT data. The best regression equation created utilized markers placed on the tuber coxae and tuber ischii of the pelvis and resulted in a mean 3D Euclidean distance error of 6.43 ± 2.22 mm (mean ± standard deviation) between the known and estimated hip joint center. The regression models produced allow for more detailed, accurate and robust analysis of sheep biomechanics.

A regularized functional method to determine the hip joint center of rotation in subjects with limited range of motion

The symmetrical center of rotation estimation (SCoRE) is probably one of the most used functional method for estimating the hip join center (HJC). However, it requires of complex multi-plane movements to find accurate estimations of HJC. Thus, using SCoRE for people with limited hip range of motion will lead to poor HJC estimation. In this work, we propose an anisotropic regularized version of the SCoRE formulation (RSCoRE), which is able to estimate the HJC location by using only standard gait trials, avoiding the need of recording complex multi-plane movements. RSCoRE is evaluated in both accuracy and repeatability of the estimation as compared to functional and predictive methods on a self-recorded cohort of fifteen young healthy adults with no hip joint pathologies or other disorders that could affect their gait. Given that, no medical images were available for this study, to quantify the global error of HJC the SCoRE residual was used. RSCoRE presents a global error of about 12 mm, similarly to the best performance of SCoRE. The comparison of the coordinate's errors at each coordinate indicates that HJC estimations from SCoRE with complex multi-plane movements and RSCoRE are not statistical significantly different. Finally, we show that the repeatability of RSCoRE is similar to the rest of the tested methods, yielding to repeatability values between 0.72 and 0.79. In conclusion, not only the RSCoRE yields similar estimation performance than SCoRE, but it also avoids the need of complex multi-plane movements to be performed by the subject of analysis. For this reason, RSCoRE has the potential to be a valuable approach for estimating the HJC location in people with limited hip ROM.

Functional hip joint centre determination in children with cerebral palsy

BACKGROUND

Although functional methods determining the hip joint center (HJC) are becoming increasingly popular, no systematic investigation has been conducted yet to assess the reliability of functional hip joint calibration in patients with cerebral palsy (CP).

RESEARCH QUESTION

What is the most reliable way to conduct functional calibration motions for estimating HJC location in children with CP and movement disorders?

METHODS

Twenty-two patients with CP were included in the study. A marker set for Plug-in Gait with additional cluster markers was used. Two functional calibration movements, including a new movement, were proposed and tested with one and three repetitions each. Functional HJCs were determined using the SCoRE approach and compared to results obtained by applying the conventional regression method for assessing face validity.

RESULTS

The choice of calibration movement had significant impact on SCoRE residuals and HJC location. Increasingly repeating calibration movements did not improve results. A modified star movement by allowing the toes to tip the ground provided the most reliable data and is feasible for children with GMFCS level I-III. The feasibility of the method is further improved by analyzing hip motion in the contralateral stance limb and, among the calibration movements, gave the most precise HJC estimation.

SIGNIFICANCE

Type and performance of the functional calibration movement is one key factor for determining a robust HJC. Analyzing the data in the stance leg via the modified star motion yielded robust and reasonable results for the HJC location, which should be validated in further studies that include imaging methods. Using one repetition instead of three seems promising in terms of feasibility for patients with movement disorder.

Estimating soft tissue artifact of the thigh in high knee flexion tasks using optical motion Capture: Implications for marker cluster placement

Soft tissue artifact in motion capture is widely accepted as a significant source of error in kinematic and kinetic measurements. Non-invasive methods of estimating soft tissue artifact, those requiring only motion capture, provide a feasible method to evaluate marker placement on a segment and enable recommendations for marker configurations which can minimize soft tissue artifact. The purpose of this study was to investigate the effect of thigh marker cluster location on soft tissue artifact during high knee flexion (>120 deg) as unique deformation of soft tissue occurs in this range (e.g. thigh-calf contact). Motion of the pelvis and lower limbs were recorded during squatting and kneeling in fifty participants. Six rigid marker clusters were affixed to the skin on the anterior, lateral, and anterolateral aspect, at the distal and middle third of the thighs. To estimate soft tissue artifact, the functional hip joint center was reconstructed relative to the pelvis cluster and each of the six thigh clusters throughout motion. The difference in the position of these two points was input into Bland-Altman analyses and compared between the thigh clusters. Across the tasks, the total mean difference ranged from 2.81 to 8.95 cm while the lower and upper limits of agreement ranged from -0.79 to 2.54 cm and 5.04 to 17.65 cm, respectively. Using this non-invasive method, the mid-anterolateral cluster was least susceptible to soft tissue artifact and thus would be recommended, while the lateral clusters were most susceptible and should avoided in high knee flexion and similar tasks.

Verification of an improved hip joint center prediction method

In motion analysis, the hip joint center (HJC) is used to define the proximal location of the thigh segment and is also the point about which hip moments are calculated. The HJC cannot be palpated; its location must be calculated. Functional methods have been proposed but are difficult to perform by some clinical populations. Therefore, regression methods are utilized, but yield large errors in estimating the HJC location. These prediction methods typically utilize the anterior and posterior superior iliac spines, where excessive adipose tissue makes correctly locating difficult. A new regression method (Hara) utilizes leg length and has been shown to improve HJC location in cadavers and less error than previous pelvic based regression methods, such as those proposed by Harrington et al. This study compared the accuracy of the HJC location calculated with both of the Harrington methods and the Hara method. The coronal knee angle was calculated for each method using a static motion analysis trial, and compared to the tibiofemoral angle measured on a gold standard digital full-leg coronal radiograph. This study demonstrated that the Hara method was more accurate than either of the Harrington methods. The mean error between the gold standard x-ray measurement and the motion analysis calculation for the Harrington (stepwise and LOOCV), the Harrington (linear regression), and Hara regression methods, respectively were 6.0°, 4.0°, and 1.8°. Accurately modeling the HJC is critical for data interpretation and patient care. This study confirmed that the Hara HJC regression method is valid in an in-vivo setting.

Validation of hip joint center localization methods during gait analysis using 3D EOS imaging in typically developing and cerebral palsy children

Localization of the hip joint center (HJC) is essential in computation of gait data. EOS low dose biplanar X-rays have been shown to be a good reference in evaluating various methods of HJC localization in adults. The aim is to evaluate predictive and functional techniques for HJC localization in typically developing (TD) and cerebral palsy (CP) children, using EOS as an image based reference. Eleven TD and 17 CP children underwent 3D gait analysis. Six HJC localization methods were evaluated in each group bilaterally: 3 predictive (Plug in Gait, Bell and Harrington) and 3 functional methods based on the star arc technique (symmetrical center of rotation estimate, center transformation technique and geometrical sphere fitting). All children then underwent EOS low dose biplanar radiographs. Pelvis, lower limbs and their corresponding external markers were reconstructed in 3D. The center of the femoral head was considered as the reference (HJCEOS). Euclidean distances between HJCs estimated by each of the 6 methods and the HJCEOS were calculated; distances were shown to be lower in predictive compared to functional methods (p<0.0001). Contrarily to findings in adults, functional methods were shown to be less accurate than predictive methods in TD and CP children, which could be mainly due to the shorter thigh segment in children. Harrington method was shown to be the most accurate in the prediction of HJC (mean error≈18mm, SD=9mm) and quasi-equivalent to the Bell method. The bias for each method was quantified, allowing its correction for an improved HJC estimation.

Side-to-side variation in normal femoral morphology: 3D CT analysis of 122 femurs

BACKGROUND

The contralateral femur is often used as reference for reconstruction in unilateral hip joint pathology. The objective of this study was to quantify the side-to-side variation in proximal femur. We hypothesized that significant side-to-side differences exist between left and right femur with implications for preoperative planning and leg length discrepancy following hip arthroplasty.

MATERIALS AND METHODS

CT-based 3D femoral models were reconstructed for 122 paired femurs in 61 young healthy subjects (46.9±6.8 years) with no history of hip pathology. Side-to-side differences of several femoral morphologic parameters, including femoral head diameter, femoral anteversion, horizontal offset and femoral head center location, were compared and correlated with demographic factors using multiple linear regression.

RESULTS

Significant side-to-side differences (P<0.01) were found in femoral anteversion (4.3±3.8°; range: 0.2° to 17.3°), horizontal offset (2.5±2.1mm; range: 0.1 to 10.3mm), and femoral head center location (7.1±3.8mm; range: 0.5 to 19.4mm). The difference in femoral anteversion was strongly correlated with the difference in neck diameter (R(2)=0.79), whereas the difference in horizontal femoral offset was highly correlated with the head diameter difference (R(2)=0.72). Femoral head center difference was correlated with the femoral anteversion, horizontal offset and neck-shaft-angle difference (R(2)=0.82).

DISCUSSION

Relying on the anatomic landmarks of the contralateral femur during hip arthroplasty may not necessarily result in restoration of native anatomy and leg-length. Knowledge of the baseline side-to-side asymmetry could provide a range of error that would be tolerable following hip reconstruction.

LEVEL OF EVIDENCE

Level IV.

TYPE OF STUDY

Retrospective observational study.

Regression models to predict hip joint centers in pathological hip population

The purpose was to investigate the validity of Harrington's and Davis's hip joint center (HJC) regression equations on a population affected by a hip deformity, (i.e., femoroacetabular impingement). Sixty-seven participants (21 healthy controls, 46 with a cam-type deformity) underwent pelvic CT imaging. Relevant bony landmarks and geometric HJCs were digitized from the images, and skin thickness was measured for the anterior and posterior superior iliac spines. Non-parametric statistical and Bland-Altman tests analyzed differences between the predicted HJC (from regression equations) and the actual HJC (from CT images). The error from Davis's model (25.0 ± 6.7 mm) was larger than Harrington's (12.3 ± 5.9 mm, p<0.001). There were no differences between groups, thus, studies on femoroacetabular impingement can implement conventional regression models. Measured skin thickness was 9.7 ± 7.0mm and 19.6 ± 10.9 mm for the anterior and posterior bony landmarks, respectively, and correlated with body mass index. Skin thickness estimates can be considered to reduce the systematic error introduced by surface markers. New adult-specific regression equations were developed from the CT dataset, with the hypothesis that they could provide better estimates when tuned to a larger adult-specific dataset. The linear models were validated on external datasets and using leave-one-out cross-validation techniques; Prediction errors were comparable to those of Harrington's model, despite the adult-specific population and the larger sample size, thus, prediction accuracy obtained from these parameters could not be improved.

Accuracy of Functional and Predictive Methods to Calculate the Hip Joint Center in Young Non-pathologic Asymptomatic Adults with Dual Fluoroscopy as a Reference Standard

Predictions from biomechanical models of gait may be sensitive to joint center locations. Most often, the hip joint center (HJC) is derived from locations of reflective markers adhered to the skin. Here, predictive techniques use regression equations of pelvic anatomy to estimate the HJC, whereas functional methods track motion of markers placed at the pelvis and femur during a coordinated motion. Skin motion artifact may introduce errors in the estimate of HJC for both techniques. Quantifying the accuracy of these methods is an area of open investigation. In this study, we used dual fluoroscopy (DF) (a dynamic X-ray imaging technique) and three-dimensional reconstructions from computed tomography images, to measure HJC locations in vivo. Using dual fluoroscopy as the reference standard, we then assessed the accuracy of three predictive and two functional methods. Eleven non-pathologic subjects were imaged with DF and reflective skin marker motion capture. Additionally, DF-based solutions generated virtual markers placed on bony landmarks, which were input to the predictive and functional methods to determine if estimates of the HJC improved. Using skin markers, functional methods had better mean agreement with the HJC measured by DF (11.0 ± 3.3 mm) than predictive methods (18.1 ± 9.5 mm); estimates from functional and predictive methods improved when using the DF-based solutions (1.3 ± 0.9 and 17.5 ± 8.6 mm, respectively). The Harrington method was the best predictive technique using both skin markers (13.2 ± 6.5 mm) and DF-based solutions (10.6 ± 2.5 mm). The two functional methods had similar accuracy using skin makers (11.1 ± 3.6 and 10.8 ± 3.2 mm) and DF-based solutions (1.2 ± 0.8 and 1.4 ± 1.0 mm). Overall, functional methods were superior to predictive methods for HJC estimation. However, the improvements observed when using the DF-based solutions suggest that skin motion artifact is a large source of error for the functional methods.