Abstract
PURPOSE: To investigate the reproducibility of 3 different tracing methods to determine a reliable method to define the proper anatomical position of the mandibular canal based on cone beam computed tomography (CBCT) data. MATERIALS AND METHODS: Five dentate and 5 edentate patients were selected at random from the CBCT database. Two independent observers traced both the left and the right mandibular canal using 3-dimensional image-based planning software (Procera System NobelGuide; Nobel Biocare, Goteborg, Sweden). All mandibular canals were traced using 3 different methods. Method I was based on coronal views, also known as cross-sections. Panorama-like reconstructions were the starting point for method II. The third method combined methods I and II. RESULTS: With respect to interobserver reliability, no significant difference (P = .34) for the various methods was observed. The reproducibility was better in edentate than in dentate jaws (P = .0015). The difference between 2 tracings was the lowest for the combined method: within a range of 1.3 mm in 95% of the course of the canal. The most obvious deviations were mainly seen in the anterior part of the canal. CONCLUSIONS: The best reproducible method for mandibular canal tracing is the combined method III. Between observers, still a mean 95th percentile deviation threshold of 1.3 mm (SD 0.384) is noted, indicating that a safety zone of 1.7 mm should be respected. When planning surgery on CBCT-based data, surgeons should be aware of the obvious deviations located in the region of the anterior loop of the canal.
