Radiographic detection of occlusal caries in noncavitated teeth. A comparison of conventional film radiographs, digitized film radiographs, and RadioVisioGraphy.

The aim of this study was to compare the accuracy of conventional film radiographs, digitized radiographs, and RadioVisioGraphy (RVG) for the detection of dentinal caries in occlusal surfaces of noncavitated extracted teeth. Eighty-one fully erupted extracted third molars were assessed by four observers using five radiographic methods: conventional film radiographs, digitized radiographs with contrast enhancement and with a filtering procedure, and RVG with contrast enhancement and with x function. Scoring criteria were dichotomous. Histologic sections (500 to 600 microns thick) served as the validation criterion. The two digital methods with contrast enhancement tended to perform more accurately than, although not significantly different from, the other three methods (p greater than 0.05). Average likelihood ratios (true positive/false positive) were 4.3 (conventional film radiographs), 4.4 (RVG with contrast enhancement), 3.6 (RVG with x function), 4.9 (digitized radiographs with contrast enhancement), and 3.7 (digitized radiographs with filtering procedure). Observers 1 and 2 (oral radiologists) performed better than observers 3 and 4 with all five methods (p less than 0.005).