Purpose: A functional centre of rotation (CoR) is often required in biomechanical analysis of the hip or as a landmark in computer guided surgery. It was previously shown that circumduction motions predict a CoR that is inferior and lateral to the geometric centre of the hip bearing surfaces. It is therefore necessary to establish the best method for determining the CoR to improve surgical planning. The objective of this study was to compare the predicted CoR from circumduction and star motions, and to compare these to the geometric centre of the joint.
Method: Eight cadaveric hips from four cadavers were tested. Prior to testing, CT scans of the cadavers were made from the iliac crest to the tibial plateau; the alpha angle for all hips was less than 50° so all hips were considered ‘normal’. Reflective marker arrays were rigidly mounted on the femoral diaphysis and iliac spine using 4mm Steinman pins. A five-camera Vicon system (Oxford, UK) was used to track the motions of the arrays during manipulation of the lower limb. To determine the functional hip centre, trials consisting of five cycles each of circumduction, flexion-extension and abduction-adduction were performed on each lower limb; three trials of each motion were performed. The range of motion was approximately 45° in the coronal and sagittal planes. For the ‘star’ motion, the flexion-extension and abduction-adduction trial data were combined. Following the trials the hip was dissected to expose the articular surfaces of the femoral head and acetabulum. These surfaces were traced using a pointer equipped with reflective markers to determine the geometric centre. To calculate the functional centre, the 3D coordinates of the markers were used to construct a local-to-global 3D transform for each frame throughout the trial. The geometric centre was calculated using a least-squares sphere fit (Gauss-Newton) of the trace data, calculated in the respective local coordinate systems. The coordinates of the functional centres were then transformed to an anatomic coordinate system, using the geometric centre as the origin. All calculations were performed using Matlab (Mathworks, Inc, MA, USA). A t-test was performed in each anatomic direction to detect differences in CoR predicted by the two motions.
Results: Both the circumduction and star motions resulted in a similar CoR. Differences were 0.41±2.25mm in the anterior-posterior direction; 0.09±0.72mm in the superior-inferior direction; and 0.21±0.82mm in the medial-lateral direction, none of which were significant (p>
0.5). The overall mean distance between the CoR predicted by the two motions was 2.0±1.3mm. The functional centre was also found to be lateral and inferior to the geometric centre, and was consistent for each motion. Results for the acetabulum showed similar trends.
Conclusion: This study has shown that circumduction and star motions are equivalent in predicting the hip functional CoR; differences were small compared to the dimensions involved in studies such as gait analyses. However, both motions predicted a CoR that was inferior and lateral to the spherical centre of the femoral head, suggesting that the hip does not act as a true ball-and-socket joint with congruent spherical bearing surfaces. This may have important consequences in studies at the scale of the hip joint, especially for pathological conditions such as femoroacetabular impingement.