[Best matching. Experimental comparison of different matching procedures for use in computer navigation]

Abstract

Introduction: Computer navigation systems are increasingly becoming an integral part of the surgical routine in orthopedic and trauma surgery due to improvements in intraoperative visualization procedures. The matching, i.e. data alignment between virtual and therapeutic objects, is however still a persisting problem. In recent years various matching procedures have been developed to attempt to solve the problem. In this study we compared three matching procedures using the VectorVision navigation system.

Material and methods: For each matching procedure three artificial models of the lumbar spine (TH10-Os sacrum, Synbone, Malans CH) were used with the VectorVision system as navigation system (BrainLAB, Feldkirchen/Siemens, Erlangen, Germany). The experimental setup was standardized for the different navigational methods. First a CT scan of the models was acquired and based on this data set a master planning for all pedicle drillings was performed. A total of 48 drillings with pairpoint, region or CT fluoromatching were accomplished and evaluated. The time periods needed for the matching procedures were documented and compared with each other. The precision of the drillings was evaluated within the postinterventional CT. Altogether 144 drillings for pedicle screws were performed on artificial models based on an identical planning for all test series.

Results: Within the experimental study of 144 drillings, only 2 perforations (1.3%) of the lateral cortical wall were detected. The time needed for the matching procedure was the shortest for region matching, followed by the pairpoint matching. The CT fluoromatching was the longest procedure. Variations in the distance of the individual drilling to the cortical wall were comparably large in all test series (on average 1.3 mm). Significant differences concerning the precision of the different matching procedures could only be shown for pairpoint matching.

Conclusion: In our study pairpoint matching was the only procedure without misplacement and on average had the largest safety margin for drillings. Thus this method was the most precise procedure. The region matching procedure offers the advantage of the fastest matching with comparable precision.