ABSTRACT
The aim of the study was to evaluate the accuracy of microcomputed tomography (mCT) to detect dentinal cracks when compared with scanning electron microscopy (SEM) and operating microscopy (OM). Different conditions of pixel size (10 or 17 µm), sample moisture (dry/moist), and transillumination (with/without) were evaluated. Additionally, the influence of the dentinal defect width on its detection was analyzed. The root canals of human mandibular incisors were prepared with the Reciproc R40 instrument (VDW, Munich, Germany). The roots were sectioned 5 and 10 mm from the apex, and mCT scans of middle and apical segments were performed at two pixel sizes: 10 µm and 17 µm, under dry and moist conditions (groups: 10dry, 10moist, 17dry, and 17moist). The operating microscope was used with and without transillumination (groups: OMTrans and OM). Findings showed that accuracy was moderate for the 10dry, 10moist, and OMTrans groups, poor for OM and very poor for 17dry and 17moist. The thickness of the dentin crack significantly influenced its detection by mCT using the resolution of 10 µm in both dry and wet conditions (P = .002), 17 µm in the dry condition (P = .002), and by the operating microscope using transillumination (P = .009). Some cracks visualized in SEM were not detected by mCT and an operating microscope. Not only the mCT resolution but also the sample moisture condition and the dentinal crack width can significantly influence its detection.