Abstract
OBJECTIVES: The objective of this study was to compare three methods for localization of impacted maxillary canines using only conventional panoramic radiographs. METHODS: The panoramic radiographs of 94 patients (102 impacted maxillary canines) were reviewed and evaluated using the methods magnification, angulation and superimposition. The actual positions of them were decided with cone beam CT images. The predicted positions of impacted canines from the magnification and angulation methods were compared using the McNemar χ(2) test. Sensitivity, specificity, accuracy, positive-likelihood ratio and negative-likelihood ratio were calculated. The canine-incisor index values and α angles of palatally and bucally non-rotated impacted canines were compared using the Mann-Whitney U test. RESULTS: The statistical analysis revealed that there was a significant difference between the magnification and angulation methods (p < 0.01). Using the magnification method, 68.00% of buccal canines and 69.57% of palatal canines could be localized correctly. The results of the angulation method were 28.57% and 84.91%, respectively. The sensitivity of the angulation method for buccal canines was very low. In the superimposition method, 82.98% of the superimposing samples were palatal. CONCLUSIONS: The magnification and angulation methods were not reliable methods for locating the impacted canine with a single panoramic radiograph. Magnification was more successful than the angulation method. Further research is needed on the magnification method. The image superimposition method could be used as an adjunct to others.