Detecting Residual Root Canal Filling Material After Retreatment: Cone-Beam Computed Tomography and Digital Microscopy Compared with Micro-Computed Tomography

Aim: Accurate detection of residual root canal filling after retreatment is essential for comparing protocols. This in vitro study compares cone beam computed tomography (CBCT) and digital microscopy (DGM) with micro computed tomography (micro CT) as the reference standard. Fifteen extracted single rooted mandibular premolars were instrumented and obturated with gutta percha and a calcium silicate based sealer. ProTaper Universal Retreatment files were used for retreatment. Residual material in the coronal, middle, and apical thirds was assessed using CBCT (voxel size 0.10 mm), micro CT (60 µm), and DGM on the split-root surface. Volumetric percentages (CBCT, micro CT) and surface percentages (DGM) were analyzed separately. Diagnostic performance was evaluated using binary detection (presence/absence) based on an ordinal scoring system. DGM had a low median residual surface percentage (0.34–1.52%). CBCT yielded higher median residual volume percentages than did micro CT for all thirds. For example, after Bonferroni correction, 13.10% (CBCT) and 3.27% (micro CT) of the middle thirds presented discordance, with *p* = 0.002. In binary detection, CBCT showed high sensitivity (57.1–100%) but low specificity (25.0–50%), leading to frequent false positives. DGM had 100% specificity in all thirds, but its sensitivity was very low in the coronal (14.3%) and middle (0.0%) thirds. Its perfect apical score (100%) may be spurious due to the small sample size. Within study limitations, micro CT remains the most reliable laboratory reference. CBCT produced higher volume estimates than micro CT, which is consistent with blooming and partial volume artifacts. DGM is highly specific but cannot detect subsurface remnants. Neither method should be used interchangeably with micro CT for volumetric quantification.