Brightness and contrast adjustments influence proximal caries detection on handheld X-ray radiographs.

Abstract

This study evaluated the influence of brightness and contrast adjustments on the diagnosis of proximal caries lesions in radiographs acquired with a handheld X-ray device. A complementary metal oxide semiconductor sensor (SnapShot, Instrumentarium Imaging, Milwaukee, WI, USA) and a handheld Eagle X-ray device (Alliage, São Paulo, Brazil) were used to acquire radiographs of 20 mandibular molars and 20 mandibular premolars randomly arranged in 20 phantoms. The device settings were 60 kVp, 2.5 mA, and 0.45 seconds of exposure. Brightness and contrast of the resulting radiographs were modified in four combinations: a) -30% brightness and +30% contrast; b) -15% brightness and +15% contrast; c) +15% brightness and -15% contrast; and d) +30% brightness and -30% contrast. After randomization, the original radiographs and the four modified versions were individually assessed by five oral and maxillofacial radiologists to detect proximal caries lesions. The area under the receiver operating characteristic curve (AUC), sensitivity, and specificity were calculated from the examiners' responses and compared with one-way analysis of variance (p<0.05). Intra- and inter-examiner agreement for radiographic diagnosis was assessed using the weighted kappa index. Sensitivity values for detecting proximal caries lesions on radiographs with increased brightness and decreased contrast (+ 30% brightness and -30% contrast) were significantly lower compared with the other combinations (p < 0.05), whereas AUC and specificity values were not influenced by the adjustments tested (p > 0.05). Therefore, increasing brightness and decreasing contrast on radiographs acquired with a handheld X-ray device is not recommended, since it may impair diagnostic accuracy for proximal caries lesions.