Abstract
Background/Objectives: Pulp stones are calcified masses within the dental pulp that may complicate endodontic procedures. Although cone beam computed tomography (CBCT) has been reported to detect pulp stones more frequently than two-dimensional radiography, direct comparisons between digital periapical radiography (DPR) and CBCT, particularly at different voxel sizes, remain limited. This study aimed to compare pulp stone detection rates between DPR and CBCT acquired at two voxel sizes and to evaluate inter-modality agreement using a location-based analysis for pulp stone identification in extracted teeth. Methods: Fifty-two extracted human teeth were examined using DPR and CBCT at voxel sizes of 0.2 mm and 0.1 mm under standardized ex vivo conditions. Pulp stones were evaluated in the coronal and radicular regions using a binary scoring system (presence/absence). Detection rates were compared across imaging modalities, and inter-modality agreement was evaluated using McNemar's test in the analysis stratified by pulp stone location. Associations between pulp stone detection and age, gender, tooth status, and stone location were explored using chi-square tests. Results: CBCT at 0.1 mm demonstrated the highest detection rate for pulp stones (63.5%), followed by CBCT at 0.2 mm (57.7%) and DPR (50%), with no statistically significant differences among modalities (p > 0.05). Agreement analysis showed that CBCT at 0.2 mm had higher agreement with CBCT at 0.1 mm than DPR, particularly in the coronal region, whereas DPR showed lower agreement in the radicular region. No significant associations were observed between pulp stone detection and age, gender, or tooth status. Conclusions: Under standardized ex vivo conditions, CBCT showed numerically higher pulp stone detection rates than DPR. Location-based agreement analysis indicated greater consistency between CBCT voxel sizes than between CBCT and DPR, particularly in the radicular region. These findings highlight differences in pulp stone detectability across imaging modalities and voxel resolutions and may help explain variability in radiographic detection under controlled conditions.