Presented at NACOB 98: North American Congress on Biomechanics Canadian Society for Biomechanics - American Society of Biomechanics University of Waterloo Waterloo, Ontario, Canada August 14-18, 1998

---

RADIOGRAPHIC AND NON-INVASIVE DETERMINATION
OF THE HIP JOINT CENTER LOCATION: EFFECT ON HIP JOINT ANGLES

R. N. Kirkwood ^1,2 , E. G. Culham ¹ , P. Costigan ^2
1 Department of Anatomy and Cell Biology and ² Clinical Mechanics Group
Queen's University, Kingston, Ontario, Canada K7L 3N6

INTRODUCTION

The accurate quantification of the angles obtained at the hip not only depends on the accuracy of locating bony landmarks on body segments, but also on the estimation of the location of the hip center of rotation (HC) (1). The purpose of this study was to determine the error in hip joint angle measures when hip center location was estimated using non-invasive methods by comparing values to those obtained using a standardized x-ray technique (QPR). The sensitivity of the angles measured to errors in locating the HC was also investigated.

REVIEW AND THEORY

Previous researchers have estimated the location of the HC based on external non-invasive measurements of points on the pelvis (2,3,4). Andriacchi et al. (2,5) estimated that in the frontal plane the HC would lie 2 cm directly distal to the midpoint of a line connecting the antero-superior iliac spine (ASIS) and the symphysis pubic (SP). Bell et al. (4,6) estimated the location of the HC in all three planes using a fixed percentage of the pelvic width (PW) defined as the distance between the right and left ASIS. Seidel et al. (3) stated that estimating the HC location using fixed proportions of three pelvic parameters, pelvic width (PW), pelvic depth (PD) and pelvic height (PH), optimized hip center location (Figure 1). Radiographic techniques have also been used to determine the location of the HC (7). However, planar films are prone to errors of parallax and poor control of subject positioning (8). The QPR system was developed to control these sources of error (8,9). Although costly and involving x-ray exposure, the QPR is accurate in estimating the hip joint center of rotation.

METHODOLOGY

Kinematic data were obtained using an Optoelectronic system (10,11,12). Three-dimensional angular motion at the hip was calculated using the floating axis method described by Grood and Suntay (13). The fixed coordinate system of the thigh and lower back were determined based on external markers placed over the femur and lumbar spine. Invasive method (QPR): Lead beads placed over the anatomical landmarks are visible on the x-rays, and used to calculate the relationship between the surface marker and the internal structure of the segments. At the hip both the external marker over the greater trochanter and the hip joint center (centre of the femoral head) are digitized from the x-ray and the distance between them in the frontal and transverse directions calculated. Non-invasive methods: Five anatomical landmarks (right ASIS (P1), left ASIS (P2), SP (P4), greater trochanter (P6), PSIS (P5)) were palpated and 3D coordinates obtained using the Optotrak (Figure 1).

Figure 1

The distances between P1 and P2 (PW) and P1 and P5 (PD) were calculated using the distance formula:

Where x ² is equal to (P2x - P1x) ² thus PW = sum of the square root of (P2x -P1x) ² + (P2y - P1y) ² + (P2z - P1z) ²

The 3D coordinates of P3 and PA were first obtained by averaging the x, y, and z coordinates of P1 and P2, and P1 and P4, respectively. A distance formula was then used to obtain PH. Method 1 used the technique described by Seidel et al. (2). In the frontal plane, the right HC was located at 14% of PW medial to the right ASIS. In the sagittal plane, the HC was located at 34% of PD posterior to the right ASIS and in the transverse plane the HC was located at 79% of PH distal to the midpoint (P3) between the right and left ASIS. The correction vector was obtained by subtracting the 3D coordinates of the greater trochanter marker (P6) from the coordinate of the estimated hip center in each direction. Method 2 used the technique described by Bell et al. (3). The HC was located at 14% of PW medial to the right ASIS, 30% of PW distal to the right ASIS and 22% of PW posterior to the right ASIS. Method 3A used the technique described by Andriacchi et al. (1) correcting the HC only medially to PA. Method 3B is the same technique described by Andriacchi however, the HC was corrected medially to PA and inferiorly to a point located 2 cm below the point PA. The correction vector was obtained by subtracting the vertical external marker coordinate of P6 from the PA - 2 cm point location. The hip angles profile in all three planes obtained using the four non-invasive methods of HC location were compared descriptively with the angles obtained using the QPR correction vectors. The sensitivity analysis was obtained by displacing the HC in steps of +/- 5 and 10cm in each plane and the hip angles recomputed. A 95% confidence interval of the maximum error difference in angles measured between the non-invasive methods and the QPR was determined.

RESULTS AND DISCUSSION

Figure 2 below shows the average hip angle profiles of 10 healthy subjects during level walking obtained with the QPR and Methods 1, 2 3A and B of correcting the HC. Methods 1 and 2 showed poor agreement with the QPR in the frontal and transverse planes, whereas Methods 3A and B showed good agreement with the QPR in all three planes. The 95% confidence interval of the maximum difference in angles measured between Method 3A and the QPR ranged from 0.2° to 3.0° in the frontal plane, 1.4° to 2.9° in the sagittal plane and -2.4° to 0.5° in the transverse plane. The interval for Method 3B ranged from 0.1° to 2.7° in the frontal plane, 1.5° to 2.8° in the sagittal plane and -2.1° to 0.7° in the transverse plane. The sensitivity analysis demonstrated that errors in estimating the hip joint center location in the medial/lateral direction had a small effect on the transverse plane measures and a major effect on the frontal plane angles. Errors in estimating the hip joint center location in the anterior/posterior direction affected the frontal, sagittal and transverse angles measured. The sagittal plane angles were affected to the greatest extent. Errors in estimating the hip center location in the vertical direction caused no changes in the hip angles measured in any plane. The results suggested that Methods 3A and B are accurate in estimating the hip joint center of rotation, resulting in joint angles profiles similar in shape and magnitude to those obtained using the radiographic technique.

REFERENCES

1 Ramakrishnan et al. J Biomech 1991,24:969-977.

2 Andriacchi et al. J Bone J Surg 1980;62-A(5):749-57.

3 Seidel et al. J Biomech 1995;28(8):995-8.

4 Bell et al. J Biomech 1990;23(6):617-21.

5 Andriacchi et al. Biomechanics of Normal and Pathological Articulating Joints. p. 83-103.

6 Bell et al. Hum Mov Science 1989;8:3-16.

7 Crowninshield et al. Clin Orthop Rel Res 1978;132:140-4.

8 Siu et al. Invest Radiol 1991;26:71-7.

9 Wevers et al. J Biomed Eng 1982;4:319-24.

10 Costigan et al. Medical & Biological Enginnering & Computing 1992;30:343-50.

11 Li et al. J Biomed Eng 1993;15:392-400.

12 Deluzio et al. J Biomech 1993;26(6):753-9.

13 Grood et al. J. Biomed Eng 1983, 105:136-144.

ACKNOWLEDGMENTS

This work was supported by grants from MRC and CAPES (Brazil).