Grey-scale reversed radiographic display in the detection of approximal caries

Purpose of this study was to determine the influence of grey-scale reversal of digital radiographic images on the detection of proximal caries.

Five observers assessed digital bitewing radiographs (Sidexis((R)), Sirona) of 320 unrestored surfaces of extracted posterior teeth embedded in 20 models, simulating pairing of maxillary and mandibular arches, using a six-category caries rating scale. Images were displayed in normal [N] and inverse [IN] mode at different image sizes (display ratio: 1:1, 1:2, 1:7) on a cathode-ray tube monitor (Nokia 446 XS) and a TFT display (Panasonic LC 50 S). Validation standard was defined as histological lesion depth. Validity was expressed as areas under receiver operating characteristic curves (AUC) calculated for two levels of histological caries penetration: presence of caries and presence of a dentine lesion. The factors 'inverse display', 'image size on-screen' 'display type' and 'disease cut-off' were analysed by repeated measures ANOVA.

Inverse image display significantly influenced the diagnostic validity (P=0.014), but a reduced accuracy was only seen at the lowest image size (AUC (SE): [N] 0.64 (0.02); [IN] 0.62 (0.02)). At the validation threshold 'dentine caries' approximal caries detection deteriorated when using grey-scale reversal (AUC (SE): [N] 0.71 (0.02); [IN] 0.69 (0.02)).

In conclusion, grey-scale reversal of digital radiographs did not optimise approximal caries detection and aggravated the detectability of dentinal lesions.