AIMS: To investigate the effect of defect morphology, bone thickness and examiner experience on the accuracy to detect the presence, the type and the depth of peri-implant defects with digital periapical radiographs (PAs) and cone beam computed tomography (CBCT) in an in vitro model. METHODS: Thirty six implants were placed in fresh porcine rib bone with different types of standardized defects while sites with no defect served as control. Fourteen masked examiners evaluated 324 PAs and 108 CBCT images. The presence and type of defect, the location of the bottom of the defect and the location of first bone-to-implant contact were recorded. The sensitivity, specificity, accuracy, positive and negative predictive values were calculated for each type of defect using actual measurements as the gold standard. RESULTS: The diagnostic accuracy for PAs was affected by defect morphology, exposure time, thickness of bone walls and the level of experience of the examiner. The overall diagnostic accuracy of CBCT was high (>96%) for all types of defects. CONCLUSION: Cone beam computed tomography showed better diagnostic accuracy in the detection of peri-implant defects, this can be attributed to the fact that CBCT seems to be less affected by variables that contribute to the poorer performance of PAs.
AIMS: To investigate the effect of defect morphology, bone thickness and examiner experience on the accuracy to detect the presence, the type and the depth of peri-implant defects with digital periapical radiographs (PAs) and cone beam computed tomography (CBCT) in an in vitro model. METHODS: Thirty six implants were placed in fresh porcine rib bone with different types of standardized defects while sites with no defect served as control. Fourteen masked examiners evaluated 324 PAs and 108 CBCT images. The presence and type of defect, the location of the bottom of the defect and the location of first bone-to-implant contact were recorded. The sensitivity, specificity, accuracy, positive and negative predictive values were calculated for each type of defect using actual measurements as the gold standard. RESULTS: The diagnostic accuracy for PAs was affected by defect morphology, exposure time, thickness of bone walls and the level of experience of the examiner. The overall diagnostic accuracy of CBCT was high (>96%) for all types of defects. CONCLUSION: Cone beam computed tomography showed better diagnostic accuracy in the detection of peri-implant defects, this can be attributed to the fact that CBCT seems to be less affected by variables that contribute to the poorer performance of PAs.