Conventional film-screen versus computed storage phosphor radiography. Simulated miliary lung disease in an anthropomorphic phantom

Abstract

Rationale and objectives: In this experimental study, the authors examined whether laser printed hardcopies from digital storage phosphor radiographs yield diagnostic performance equal to conventional film-screen radiographs in the detection of simulated miliary disease, using a standardized object.

Methods: A commercially available anthropomorphic chest phantom was used for radiographic evaluation. Miliary disease was simulated by superimposing one to four sheets of millet seeds on the lungs, resembling a miliary disease pattern with varying degrees of detectability. An observer study (receiver-operating characteristic) with eight radiologists was conducted to compare the reader performance using hardcopies of computed storage phosphor radiography versus the conventional film-screen system, optimized for chest x-rays. The digitally generated images were presented as a double-image hardcopy, with a conventionally adopted version and an edge enhanced image version.

Results: When analyzed separately, one out of the eight observers performed slightly better using the conventional films. When treated as a group, analysis of the areas under the receiver-operating characteristic curves demonstrated no significant difference in reader performance for each of the systems under investigation (t = 0.286). The Wilcoxon test could not prove a statistical difference.

Conclusion: Storage phosphor technology is a method that yields equal diagnostic performance as conventional film in evaluating miliary disease of the chest.