The influence of viewing conditions on observer performance in dental radiology.

Observer performance under different viewing conditions was monitored by means of radiographs, showing teeth with approximal caries, teeth with simulated approximal caries, and Plexiglas phantoms. Series a and c were exposed at two different exposure times, thus obtaining a dark and a light set of radiographs. The series a, b, and c were read by 11, 2, and 5 observers, respectively. All radiographs were examined under two viewing conditions on different occasions as follows: ceiling light in a room without a window, and X-ray viewer (X 2 magnification) with two alternative light sources. Diagnostic quality, measured as the ROC area, showed small variations in accordance with viewing conditions. For dark radiographs the X-ray viewer improved diagnostic quality compared with the viewing against ceiling illumination; however, room illumination provided the best viewing conditions when light radiographs were examined. Most of these differences were not statistically significant (p greater than 0.05). Only small differences in diagnostic strategies (cut-off points) were recorded between viewing conditions.