Roentgen Stereophotogrammetric Analysis

Model-based RSA software

In 1993, we started our research on improving the Roentgen stereophotogrammetry – or Radiostereometry. The basic RSA technology was originally developed by Selvik in Lund, Sweden, as a highly accurate technique for the assessment of three-dimensional migration of prostheses (Selvik, 1989). RSA used to be a complicated and time-consuming technique in which several steps needed to be set to come to its phenomenal accuracy which ranges between 0.05 and 0.5 mm for translations and between 0.15º and 1.15 º for rotations (95% confidence interval; Kärrholm, 1989).

The preparations for RSA start with the introduction small roentgen opaque markers in the bone and attachment of markers to the prosthesis to serve as well-defined artificial landmarks. Two synchronised roentgen foci are used to obtain a stereo image of the bone and the prosthesis. Using a calibration object that holds tantalum markers at accurately known positions, the positions of the roentgen foci are assessed. The coordinates of the bone and prosthesis markers are accurately measured and the three-dimensional micromotion is reconstructed using special RSA software.

In full collaboration with the Division of Image Processing of LUMC, our group has played an important role in the optimisation and accessibility of RSA by developing automatic marker detection algorithms and model-based RSA, an technique that no longer requires attaching markers to implants to facilitate RSA measurements, but uses three-dimensional surface models of prostheses, models that are either computer drawings provided by the prosthesis manufacturer or laser scanned surface models.