Is cone beam computed tomography accurate for post-operative evaluation of implants? an in vitro study
Available online 3 August 2017
To evaluate the accuracy of CBCT images for bone/implant interface diagnosis in comparison to periapical radiographs.
Titanium implants were inserted in 74 bovine rib blocks in intimate contact to the bone walls, and with a gap of 0.125 mm (simulating a failure in the osseointegration process). Periapical radiographs were taken with conventional film, and CBCT scans were acquired with i-CAT (0.2mm and 0.125mm voxel) and Kodak (0.2mm and 0.076mm voxel) units. Three examiners evaluated the images using a 5-point scale. Diagnostic accuracy was analyzed through sensitivity, specificity, and the area under the ROC curve (AUC) with 95% confidence intervals (CI). Intra- and inter-examiner agreements were analyzed through Kendall's concordance test.
Intra- and inter-examiner agreements showed satisfactory results. The greatest accuracy was observed with conventional radiographs (AUC = 0.963 / CI 95% = 0.891-0.993). I-CAT 0.125mm images showed good accuracy (AUC = 0.885 / CI 95% = 0.790-0.947), with no significant difference compared to conventional radiography. Kodak images had high specificity and low sensitivity, presenting more false-negative results.
Conventional radiographs showed the highest accuracy for bone/implant interface diagnosis. However, CBCT scans (i-CAT 0.125 mm voxel), if available or if performed for pre-surgical assessment of another implant site, may provide similar accuracy.