PURPOSE: The purpose of this receiver operating characteristic study was to compare diagnostic efficacy with images displayed on monitors of varying spatial resolutions and maximum luminance levels to that with storage phosphor radiographic hard copies. MATERIALS AND METHODS: Seven types of simulated lesions were superimposed onto an anthropomorphic chest phantom. Images were viewed by five radiologists on a 2,560 x 2,048 pixel monitor (maximum luminance, 75 foot-lamberts), on two 1,024 x 1,024 monitors with maximum luminance levels of 25 foot-lamberts and 75 foot-lamberts, respectively, as well as on hard copies. Monitor images were viewed both without and with systematic magnification. RESULTS: Overall visualization of the lesions was best on hard copies, but visualization on the 2,560 x 2,048 monitor was not found to be substantially different. Lines, reticular opacifications, and catheters were found to be particularly poorly visualized with the 1,024 x 1,024 monitor. These results could be statistically significantly improved only with a systematic magnification; however, this involved a considerable increase in viewing time. Observer performance was markedly inferior with the 1,024 x 1,024 monitor with the lower luminance. CONCLUSION: Diagnostic performance with a 1,024 x 1,024 monitor is statistically significantly inferior to that with hard copies. A statistically significant improvement can be achieved with a high-resolution 2,560 x 2,048 monitor with a maximum luminance of 75 foot-lamberts.
PURPOSE: The purpose of this receiver operating characteristic study was to compare diagnostic efficacy with images displayed on monitors of varying spatial resolutions and maximum luminance levels to that with storage phosphor radiographic hard copies. MATERIALS AND METHODS: Seven types of simulated lesions were superimposed onto an anthropomorphic chest phantom. Images were viewed by five radiologists on a 2,560 x 2,048 pixel monitor (maximum luminance, 75 foot-lamberts), on two 1,024 x 1,024 monitors with maximum luminance levels of 25 foot-lamberts and 75 foot-lamberts, respectively, as well as on hard copies. Monitor images were viewed both without and with systematic magnification. RESULTS: Overall visualization of the lesions was best on hard copies, but visualization on the 2,560 x 2,048 monitor was not found to be substantially different. Lines, reticular opacifications, and catheters were found to be particularly poorly visualized with the 1,024 x 1,024 monitor. These results could be statistically significantly improved only with a systematic magnification; however, this involved a considerable increase in viewing time. Observer performance was markedly inferior with the 1,024 x 1,024 monitor with the lower luminance. CONCLUSION: Diagnostic performance with a 1,024 x 1,024 monitor is statistically significantly inferior to that with hard copies. A statistically significant improvement can be achieved with a high-resolution 2,560 x 2,048 monitor with a maximum luminance of 75 foot-lamberts.
Authors: Uta Zaspel; David W Denning; Arne J Lemke; Reginald Greene; Dirk Schürmann; Georg Maschmeyer; Markus Ruhnke; Raoul Herbrecht; Patricia Ribaud; Olivier Lortholary; Harmien Zonderland; Klaus F Rabe; Rainer Röttgen; Roland Bittner; Klaus Neumann; Joerg W Oestmann Journal: Eur Radiol Date: 2004-08-11 Impact factor: 5.315