Diagnostic accuracy of panoramic radiography, stereo-scanography and cone beam CT for assessment of mandibular third molars before surgery

Abstract

Objective. To compare the diagnostic accuracy of panoramic imaging, stereo-scanography and cone beam computed tomography (CBCT) for assessment of mandibular third molars. Material and methods. One hundred and twelve patients (147 third molars) underwent radiographic examination by panoramic imaging, stereo-scanography and CBCT. Tooth angulation, root morphology, number of roots and relation to the mandibular canal were assessed. The same variables were assessed intra- and post-operatively and served as reference for the radiographic assessments. The diagnostic accuracy for each variable was compared between the three modalities and accuracy was further expressed as sensitivity and specificity and tested between the modalities for identifying the relation to the mandibular canal. Results. There were no significant differences between the modalities regarding tooth angulation, root morphology and number of roots. However, CBCT was more accurate than stereo-scanography for determining root bending in the bucco-lingual plane (p = 0.02). Moreover, sensitivity for direct contact to the mandibular canal (panoramic imaging: 0.29, stereo-scanography: 0.57, CBCT: 0.67) was higher for CBCT than for panoramic images (p = 0.05) and specificity for no direct contact to the mandibular canal (panoramic imaging: 0.78, stereo-scanography: 0.53, CBCT: 0.68) was higher for panoramic images and CBCT than for scanograms (p < 0.001). Conclusion. Panoramic imaging, stereo-scanography and CBCT seem equally valuable for examination of tooth angulation, number and morphology of roots of mandibular third molars. However, CBCT was more accurate for assessment of root bending in the bucco-lingual plane and more accurate than panoramic images to identify direct contact to the mandibular canal.

Key Words::

• CT scanning
• radiography
• wisdom teeth

Acknowledgments

The authors would like to thank ‘Calcinfonden' for financial support. The staff at the Section of Oral Radiology and Section of Oral and Maxillofacial Surgery and Oral Pathology, Department of Dentistry, Health, Aarhus University are thanked for their help during the period of data sampling. Especially Bjarne Simonsen, Jens Hartlev and Birgit Kenrad are thanked for support with the surgical procedures.

Declaration of interest: The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper.