Abstract
Introduction and Aims: Stem migration determined by Radiostereometry (RSA) typically requires markers attached to the stem. Previous studies with marker-less stems either detem1ined movement of the prosthesis head only, or approximate the stem to a simple geometric model. We compare the precision of this method to one utilising a surface model of the implan.
Method: The first method assumes axially symmetric distal stem (SDS) and hence calculates the stem tip and axis. The method was implemented as an addition to UMRSA (RSA Biomedical AB, Umea, Sweden). Applied originally to manual measurement of radiographs, it is now also implemented as automated digital image analysis for use with Digital UMRSA. The second method (ASRSA) matches a three-dimensional surface model of the stem to the digitised images. Exeter stem models were supplied by the manufacturer (Stryker Howmedica Osteonics, France). Clinical precision of each method was determined from repeat same day RSA examinations obtained from an ongoing study.
Results: Clinical precision, given as the 99% confidence interval, was calculated as 3.5 times the median absolute difference of the repeat measurement. Patient and radiographic equipment were repositioned for the repeat examination. All repeat pairs were from different patients. For analysis, 43 pairs were available for the manual SDS method and 23 for digital SDS. For ASRSA, only nine pairs of Exeter stems had 3D models available for analysis. ASRSA translational precision was 0.1mm transverse, 0.1mm axial, and 0.2mm sagittal. Rotational precision was 0.3 degrees transversely, 0.9 degrees axially and 0.2 degrees sagittally. Digital analysis improved the precision of the SDS method considerably compared to manual analysis, in particular axial rotation from 3.1 to 1.3 degrees and sagittal translation from 0.8 to 0.4mm, however it was still between 1.4 and five times worse than the ASRSA method.
The poor result of the SDS method in axial rotation is probably due to the breakdown of axial symmetry, whereas the excellent result of the ASRSA method is comparable to that published for marked stems. Current results are limited in number and type of implant; further work is now being carried out in larger numbers and with more stems.
Conclusion: Surface model-based digital RSA can give a precision comparable to that for implants with attached markers. This has important implications, since it can dramatically increase the power of migration results where beads have not been used and can be applied to RSA studies retrospectively.
These abstracts were prepared by Editorial Secretary, George Sikorski. Correspondence should be addressed to Australian Orthopaedic Association, Ground Floor, The William Bland Centre, 229 Macquarie Street, Sydney, NSW 2000, Australia.
At least one of the authors is receiving or has received material benefits or support from a commercial source.
