Abstract

To study observers' ability to detect mesiobuccal (MB) canals in maxillary molars using iCAT cone-beam computed tomography (CBCT) at different voxel dimensions and to assess the impact of clinical experience on accuracy of detection. Using 12 experimental models with two molars each, CBCT scans were acquired at four different voxel dimensions. From the cross-section view of these scans, 96 videos were generated. Five endodontic postgraduate students and two endodontic staff watched the videos and counted the MB canals in each root. Horizontal sections of the roots were evaluated under magnification to determine the true canal numbers. The detection of MB canals within the four resolutions was compared by odds ratio, and the weighted χ(2) test compared detection accuracy to raters' clinical experience. Rater agreement was measured by kappa statistics. Overall, 92% of the maxillary molars had two MB canals upon analysis of horizontal cross-sections. The CBCT detection increased from 60.1% at 0.4 mm voxel size to 93.3% at 0.125 mm voxel size. Significant differences (P < 0.01) were observed between the different resolutions except for the 0.2 and the 0.125 voxel scans. Second-year trainees were significantly (CI = 0.2929-0.712) more accurate than first-year trainees and endodontic staff at MB canal detection (87.9% against 77.1% and 76.8%). Intra-rater reliability increased with higher-resolution scans (41.1% to 96.4%). The reliability of detection of maxillary molar MB2 canals in CBCT scans increased as the resolution improved. Accuracy of MB2 canal detection among observer groups could not be correlated with the observers' level of clinical experience.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call