Wear and loosening are the major causes for long tem failure in Total Hip Replacement (THR). Accurate three dimensional wear analysis of radiographs has its own limitations. We report the results of our clinical study of three dimensional volumetric wear measurements using our custom low radiation risk CT based algorithm and special software

Twenty four patients (32 hips) agreed to take part in our study. The male: female ratio was 1:4. The mean age was 75 years and the mean follow up was 5.4 years. All patients had 28 mm diameter ceramic heads. Of the 32 hips, 17 hips had polyethylene inserts and 15 hips had ceramic inserts. The maximum follow up for the polyethylene and ceramic groups were 12 years and 5.5 years respectively. All the patients were scanned using Somatom Sensation 4 scanner. Using custom software, 3D reconstruction of the components was done and landmark acquisition done on the femoral head, acetabular metal component and the insert. From these landmarks, a dedicated program was used to calculate the centre of the femoral head in relation to the centre of the acetabular component in all three axes and an indirect measurement of wear obtained. Using the axes measurements graphical 3D models of migration of the femoral head component into the acetabular liner were created and volume of wear measured using special software. Accuracy of the method was assessed by measuring the radius of the femoral head since all patients had 28mm diameter heads implanted in them. Assessment of precision of method was done by calculating the level of agreement between two independent observers.

In the polyethylene group, there was no significant (< 1mm) wear in x and y axis with time. However there was significant evidence of wear in relation to time in the z axis (max wear = −2.5 mm). In the ceramic group with relatively shorter follow up, there was no evidence of significant wear in all three axes. The mean volume measured in the polyethylene group was 685 mm3 (max = 1629 mm3, min = 132mm3 ). The mean volume measured in the ceramic group was 350mm3 (max = 1045 mm3, min = 139mm3 ). The mean radius of the femoral head measured in both groups was 14.02mm (range =13.8 to 14.4 mm). Accuracy was limited by artifacts particularly in bilateral hip arthroplasties and further in the ceramic group because of the restricted access to the ceramic head for placement of markers. Measurements obtained by two independent observers showed a strong correlation (0.99, p value = 0.001) for the polyethylene group. In the ceramic group the correlation (0.69, p value=0.0126) was not as strong as the polyethylene group.

This study has produced a method for three dimensional estimation of wear that can be obtained from low dose CT scans with better accuracy and repeatability (< 0.5 mm) even than to ex vivo studies particularly in polyethylene bearings(wear rate 0.14mm/yr). Noise reduction with appropriate artefact reduction software may further improve the accuracy of this simple and repeatable method.