The influence of liquid crystal display monitors on observer performance for the detection of interstitial lung markings on both storage phosphor and flat-panel-detector chest radiography

Abstract

Purpose: To compare observer performance with a flat-panel liquid crystal display (LCD) monitor and with a high-resolution gray-scale cathode ray tube (CRT) monitor in the detection of interstitial lung markings using a silicon flat-panel-detector direct radiography (DR) and storage phosphor computed radiography (CR) in a clinical setting.

Materials and methods: We displayed 39 sets of posteroanterior chest radiographs from the patients who were suspected of interstitial lung disease. Each sets consisted of DR, CR and thin-section CT as the reference standard. Image identities were masked, randomly sorted, and displayed on both five mega pixel (2048x2560x8 bits) LCD and CRT monitors. Ten radiologists independently rated their confidence in detection for the presence of linear opacities in the four fields of the lungs; right upper, left upper, right lower, and left lower quadrant. Performance of a total 6240 (39 setsx2 detector systemsx2 monitor systemx4 fieldsx10 observers) observations was analyzed by multi-reader multi-case receiver operating characteristic (ROC) analysis. Differences between monitor systems in combinations of detector systems were compared using ANOVA and paired-samples t-test.

Results: Area under curves (AUC) for the presence of linear opacities measured by ROC analysis was higher on the LCDs than CRTs without statistical significance (p=0.082). AUC was significantly higher on the DR systems than CR systems (p=0.006). AUC was significantly higher on the LCDs than CRTs for DR systems (p=0.039) but not different for CR systems (p=0.301).

Conclusion: In clinical conditions, performance of the LCD monitor appears to be better for detecting interstitial lung markings when interfaced with DR systems.