Radiography of the mandible prior to endosseous implant treatment. Localization of the mandibular canal and assessment of trabecular bone

Sammanfattning: Mandibular autopsy specimens were examined with different radiographic techniques in order to evaluate the visibility of the mandibular canal and the measurement accuracy of distances related to the mandibular canal. Hypocycloidal, spiral and computed tomography (CT) were superior to periapical and panoramic radiography in visualising the mandibular canal. The tomographic techniques were more accurate when measurements of distances related to the mandibular canal were performed. No difference in measurement accuracy was found between the tomographic techniques. Concerning visibility of the mandibular canal, interobserver agreement was lowest for periapical radiography and highest for CT. Intraobserver agreement was moderate or good for all techniques. A high interobserver variation was found for measurability of distances related to the mandibular canal. The trabecular bone tissue in mandibular autopsy specimens was studied concerning different characteristics. Quantitative measurements of bone mass were performed and expressed as bone mineral density and trabecular bone volume. There was a significant correlation between these two measures. Significant inter- and intra- mandibular variations were found for bone mineral density as well as for trabecular bone volume. A classification system to be used prior to implant treatment, based on the trabecular pattern in periapical radiographs, was proposed. The proposed classification system was compared with the classification system proposed by Lekholm and Zarb (1985) and evaluated in terms of diagnostic accuracy and observer performance. The highest accuracy was found when the proposed classification system was used together with reference images. The diagnostic accuracy of the classification system devised by Lekholm and Zarb (1985) was not possible to evaluate. Interobserver agreement was similar with and, without reference images but higher than for the Lekholm and Zarb (1985) classification. Intraobserver agreement was comparable for the classification systems.