Validation of a functional method for the estimation of hip joint centre location

Abstract

The present study assesses the accuracy with which the subject specific coordinates of the hip joint centre (HJC) in a pelvic anatomical frame can be estimated using different methods. The functional method was applied by calculating the centre of the best sphere described by the trajectory of markers placed on the thigh during several trials of hip rotations. Different prediction methods, proposed in the literature and in the present investigation, which estimate the HJC of adult subjects using regression equations and anthropometric measurements, were also assessed. The accuracy of each of the above-mentioned methods was investigated by comparing their predictions with measurements obtained on a sample of 11 male adult able-bodied volunteers using roentgen stereophotogrammetric analysis (RSA), assumed to provide the true HJC locations. Prediction methods estimated the HJC location at an average rms distance of 25–30 mm. The functional method performed significantly better and estimated HJCs within a rms distance of 13 mm on average. This result may be confidently generalised if the photogrammetric experiment is carefully conducted and an optimal analytical approach used. The method is therefore suggested for use in motion analysis when the subject’s hip range of motion is not limited. In addition, the facts that it is not an invasive technique and that it has relatively small and un-biased errors, make it suitable for regression equations identification with no limit to sample size and population typology.

Introduction

In human movement analysis models the acetabulum and the femoral head are represented by spherical surfaces with a common centre (hip joint centre, HJC). The accuracy with which the HJC location may be estimated is crucial in terms of error propagation to the relevant kinematics and kinetics (Crowninshield et al., 1977; Cappozzo, 1986; Woltring and Fioretti, 1989; Kadaba et al., 1990; Pennock and Clark, 1990; Ramakrishnan and Kadaba, 1991; Cappozzo, 1991).

The subject specific HJC coordinates in a pelvic anatomical coordinate system (ACS) have been estimated previously using either a functional or a prediction approach. The former method (Cappozzo, 1984; Blankevoort et al., 1990; Cappozzo et al., 1995b; Shea et al., 1997) estimates the HJC as the pivot point of a 3-D relative movement between femur and pelvis recorded using photogrammetry. An adequate hip joint range of motion is required which may exclude some patients. The prediction approach uses regression equations (REs) with the independent variables describing the geometry of the pelvis. Regression coefficients have been obtained using imaging techniques on relatively small sample sizes of living adult males (Crowninshield et al., 1978; Andriacchi et al., 1980; Tylkowski et al., 1982; Bell et al., 1989, Bell et al., 1990; Davis et al., 1991) or direct measurements on a sample of adult male cadaveric specimens (Seidel et al., 1995). The REs proposed by Bell et al. (1990) were more accurate than any previously published equations and, according to these authors, were also more accurate than the functional method. Those provided by Davis et al. (1991) are currently the most widely used. More recently Seidel et al. (1995) have proposed more promising equations, however their use is not practical since the method entails the identification, through palpation, of an awkward pubic symphysis landmark.

The aim of the present study was to validate the functional method by assessing its accuracy. To this purpose, reference data were acquired using roentgen stereophotogrammetric analysis (RSA, Selvik, 1974). A comparative analysis with the REs proposed by Bell et al. (1990) and by Davis et al. (1991), was also carried out. Using the reference data and a correlation analysis, indications about effective RE form and relevant independent variables were pursued. It was also investigated whether the functional method would be appropriate for the estimation of reliable REs coefficients for which there is great demand. This method, is totally safe for the individuals involved and thus could allow studying able-bodied subjects with no limit to sample size or population typology.