The interference of digital radiographic image acquisition and processing protocols in the diagnosis of incipient enamel carious lesions

Background: This study aims to evaluate the diagnosis of proximal carious lesions through different parameters of execution and visualization of the images. Material and Methods: Proximal carious lesions were artificially induced in human teeth three different times (10, 20, and 30 hours) by in vitro demineralization. The teeth were radiographed with the Dürr digital system using three exposure times (0.10, 0.16, and 0.20s) and three different resolutions in the scanning of the phosphor plates (20-, 25- and 40-line pairs). After acquiring the images, they were submitted to VistaScan Fine, Caries 1, and Caries 2 software filters. Three radiologists analyzed the images in JPEG format. Results: The Pearson’s chi-square test showed an association between longer demineralization time and the presence of injury according to the professional’s classification. There was no statistically significant association among the other parameters, exposure time, resolution, and use of filters and the presence of a lesion. All parameters showed a greater sensitivity than specificity. The exposure time of 0.2s showed better accuracy, the resolutions 25 and 40lp showed equal results and better than 20lp and the Fine filter obtained better values than the Caries 1 and 2 filters. Conclusions: Despite the variation of parameters, exposure time, resolution, and use of filters, there were no statistically significant differences. For better diagnostic accuracy, it is recommended to use 0.2s of exposure time and a Fine filter, and scan the phosphor plates with 25lp.