A new automated way to measure polyethylene wear in THA using a high resolution CT scanner: method and analysis

Abstract

As the most advantageous total hip arthroplasty (THA) operation is the first, timely replacement of only the liner is socially and economically important because the utilization of THA is increasing as younger and more active patients are receiving implants and they are living longer. Automatic algorithms were developed to infer liner wear by estimating the separation between the acetabular cup and femoral component head given a computed tomography (CT) volume. Two series of CT volumes of a hip phantom were acquired with the femoral component head placed at 14 different positions relative to the acetabular cup. The mean and standard deviation (SD) of the diameter of the acetabular cup and femoral component head, in addition to the range of error in the expected wear values and the repeatability of all the measurements, were calculated. The algorithms resulted in a mean (± SD) for the diameter of the acetabular cup of 54.21 (± 0.011) mm and for the femoral component head of 22.09 (± 0.02) mm. The wear error was ± 0.1 mm and the repeatability was 0.077 mm. This approach is applicable clinically as it utilizes readily available computed tomography imaging systems and requires only five minutes of human interaction.

Figure 1

Three 2D orthogonal projections showing the 3D spherical landmark located in the apical dome hole. (a) shows the axial projection, (b) the coronal, and (c) the sagittal.

Figure 2

The five landmarks chosen on the acetabular cup are shown in 3D. Only landmarks 1, 2, 3, and 5 are used in the surface extractor calculations. Landmark 4 is used only by the operator to locate the apical dome hole.

Figure 3

The top row contains the three 2D orthogonal projections showing the 3D spherical landmark located at the junction between the femoral component head and stem and the bottom row contains the three 2D orthogonal projections showing the 3D spherical landmark surrounding the femoral component head. (a) shows the axial projection, (b) the coronal, and (c) the sagittal.

Figure 4

The top row illustrates the initial and valid surface landmarks produced on the acetabular cup by the surface extraction algorithm and the bottom row shows those for the femoral component head. The set of initial landmarks produced is shown in (a). The set of valid landmarks is shown in (b). The overlapping colored spherical glyphs represent the generated landmarks.

Figure 5

The values of the acetabular cup diameters obtained from each individual calculation are shown with the median, mean, and two SD lines drawn through the data. The median value is quite close to the mean value.

Figure 6

The values of the femoral component head diameters obtained from each individual calculation are shown with the median, mean, and two SD lines drawn through the data. The median value is quite close to the mean value.

Figure 7

The calculated CupHeadSeparation plotted against the expected CupHeadSeparation from Table 1, for scan series one and two (both trials). Lines are drawn at the exact expected value and at ±0.1 mm demonstrating the data were within 0.1 mm of the expected distance.