Acurácia da tomografia computadorizada de feixe cônico no diagnóstico da adaptação marginal de dentes reconstruídos com o sistema cerâmico à base de dissilicato de lítio

Abstract

This study investigated the accuracy of cone-beam computed tomography (CBCT) compared to digital interproximal radiography in evaluating the marginal adaptation of teeth reconstructed with full lithium disilicate crowns. The influence of artifacts on the identification of gaps was also assessed by quantifying the contrast-to-noise ratio (CNR) in CBCT examinations obtained from two equipment. Thirty molars restored with lithium disilicate full crowns with adapted proximal surfaces and gaps of 0.3 and 0.5 mm were subjected to CBCT and radiographic examination. Six examiners, three prosthodontists and three radiologists, evaluated the proximal surfaces of the teeth for the presence or absence of gaps. One ROI (region of interest) was selected in an axial slice for CNR quantification. To evaluate the detection of marginal gaps in different image types, the area under the ROC curve, accuracy, sensitivity and specificity were calculated, considering the gap size, for each examiner and imaging examination. The Mann-Whitney U test was used to compare the diagnostic test results considering the independent variables "gap size" and "examiner group," and the Kruskal-Wallis test was used for the variable "imaging examination". Contrast-to-noise ratio values were compared between the two CT scanners using the Mann-Whitney U test. The results showed significant differences in the area under the ROC curve, accuracy, and sensitivity for gap detection, as well as between imaging exams in relation to diagnostic tests and between the CNRs from different equipment. No significant differences were observed between radiologists and prosthodontists. It can be concluded that CBCT showed accuracy in evaluating the marginal adaptation of lithium disilicate crowns, performing better than intraoral digital radiography. The significant difference observed in relation to CNRs reflected the different acquisition protocols used. Regarding the detection of different gap sizes, significant differences were observed between the AUC ROC values, accuracy, and sensitivity. No significant differences were observed in the diagnosis when performed by prosthodontists or radiologists.