No abstract

We evaluated the effect of the displayed image sizes on observers' ability to detect nodular ground-glass opacity (n-GGO) on CT and investigated the optimal viewing size for soft-copy reading at CT screening for lung cancer. A total of 46 patients' high-resolution computed tomography (HRCT) images (22 patients with one GGO; 24 without GGO) were displayed on a monochromatic liquid crystal display monitor at a resolution of 1,200 × 1,600. HRCT was presented on the screen with cine-mode display. We compared two viewing sizes (original size, i.e., the image displayed with a zoom factor of 1 in which each pixel value in the image is displayed as one pixel on the display: 13 cm × 13 cm; fit size, i.e., by zooming the captured image until it occupies the entire screen: 30 cm × 30 cm) in terms of radiologists' performance for detecting n-GGO on HRCT and the viewing times required for soft-copy reading decisions. Observer performance was analyzed in terms of the receiver operating characteristic (ROC) curve. A statistically significant improvement was found with the original size in the average area-under-the-ROC curve values for the accuracy of diagnosis and the viewing times compared to the fit size (P < 0.05). The original size with cine-mode display leads to increased lung GGO detection at CT screening for lung cancer, and the reduced time spent performing the diagnosis offers cost savings.

No abstract

No abstract

In our conventional visual inspection for setup verification, we have routinely used a console monitor of a CR system and a monitor of a treatment-planning system (i.e., the separate-monitor method) in order to avoid the need for CR portal-film generation. However, the separate-monitor method provided insufficient precision in detecting setup errors. We devised a setup verification procedure that uses a high-resolution liquid-crystal display monitor (i.e., the single-monitor method). Our objective in the present study was to evaluate the precision of the single-monitor method. These two methods were compared in terms of the precision of visual inspection. The single-monitor method was significantly superior to the separate-monitor method in sensitivity and in the magnitude of the discrepancy that could not be detected. The single-monitor method provides higher precision in visual inspection than does the separate-monitor method, and is a useful verification procedure.