Introduction: In computer assisted orthopaedic surgery, rigid fixation of the Reference Marker (RM) system is essential for reliable computer guidance. A minimum shift of the RM can lead to substantial registration errors and inaccuracies in the navigation process. Various types of RM systems are available but there is little information regarding the relative stabilities of these systems. The aim of this study was to test the rotational stability of three commonly used RM systems.
Materials and methods: One hundred and thirty Synbones and 15 cadavers were used to test the rotational stability of three different RM systems (Schanz' screw, RM B and RM C adjustment systems). Using a specially developed testing device, the peak torque sustained by each RM system was assessed in various anatomical sites.
Results: Comparison of means for Synbone showed that the RM C was the most stable (mean peak torque 5.60 +/- 1.21 Nm) followed by the RM B system (2.53 +/- 0.53 Nm) and the RM A (0.77 +/- 0.39 Nm) (p<0.01). The order of stability in relation to anatomical site was femoral shaft, distal femur, tibial shaft, proximal tibia, anterior superior iliac spine, iliac crest and talus. Results from the cadaver experiments showed similar results. Bi-cortical fixation was superior to mono-cortical fixation in the femur (p<0.01) but not the tibia (p=0.22).
Conclusion: The RM system is the vital link between bone and computer and as such the stability of the RM is paramount to the accuracy of the navigation process. In choosing RM systems for computer navigated surgery surgeons should be aware of their relative stability. Anatomical site of RM placement also affect the stability. Mono-cortical fixation is generally less stable than bi-cortical.