Interstitial lung disease: impact of postprocessing in digital storage phosphor imaging.

Abstract

The ability to resolve the fine linear structures of interstitial lung disease is one measure of the limiting performance characteristics of an imaging system. Conventional screen-film radiography was compared with six algorithms of isodose storage phosphor digital radiography (0.2-mm x 10-bit pixel matrix) in the detection of interstitial lung abnormality documented by means of computed tomography in 40 patients with abnormalities and 25 healthy control subjects. Performance was evaluated with an analysis of variance (the Fisher paired comparison test; P less than .05) of the average receiver operating characteristic area of 2,730 observations by six readers. The moderately and the more markedly high-frequency edge-enhanced algorithms of storage phosphor digital radiographs were equivalent in performance to screen-film radiography. The default mode, low- and medium-frequency edge-enhanced algorithms, and gray scale reversed mode of storage phosphor digital radiography were inferior to screen-film radiography. The authors conclude that high-frequency edge-enhanced algorithms can perform as well as screen-film radiography in the detection of interstitial disease.

Effects of Varying Filter Kernel Sizes on the Image Quality of Interstitial Lung Diseases

“…In their studies with simulated lung disease, PROKOP et al (20) state that the usable illustration of linear and reticular patterns can also be attained by using larger filter kernels. Whereas prevailing research suggests that high-frequencyenhancing filtering is essential in the detection of fine linear structures in the lungs (19,22), our results indicate that the representation of linear, honeycomb and fine reticular structures can be visualized in hi& image quality by using medium-and low-frequency-enhancing postprocessing (Figs. 7, 8 and 9).…”

“…TAKASHIMA (26) explains the unfavorable results of the edge-enhancing mode by the masking of nodular structures due to the emphasis of physiological structures with minimally uniform differences in density. Linear details, however, can be seen more clearly in high-frequency emphasis filterings using a small filter kernel and moderate edge enhancement (6,7,9,12,19,22). Concurrent with this, the present study managed to clearly detect linear details with high-fre- quency emphasizing filtering (S 5 ) and edge enhancement (DLR 11) (Fig.…”

“…Concerning image quality in interstitial lung diseases, opinions are divided on the storage phosphor technique, which depends on the performance of an individually adapted and time-consuming postprocessing (6,9,10,12,22,23). The influence of various digital image processing procedures on image quality was therefore examined under clinical conditions, and compared with conventional screenfilm radiographs.…”

Diagnostic usefulness of chest computed radiography—Film versus cathode-ray tube images

“…Although analyses of the arca under the ROC curves showed no significant differences in detection of pulmonary abnormalities between film and CRT, lower sensitivity in detecting interstitial abnormalities on CRT than on film shown in this study may have been affected by the lower resolution of the CRT workstation in comparison to that with a 2 K-x 2 K-pixel matrix, as well as observers' unfamiliarity with the CRT workstation 11 and uneffective utilization image processing including zooming procedure and high-frequency edge-enhanced processing. 24 Further consideration with regard to spatial resolution requirements and image processing is needed. Regarding image compression, the proportion of subjective judgments indicating that even 10:1 compressed images were poor as chest radiographs was significantly higher than that of uncompressed images.…”

Fractal-feature distance analysis of contrast-detail phantom image and meaning of pseudo fractal dimension and complexity

“…In general, evaluation on medical image quality is performed by measurements of physical indices such as signal-to-noise ratio (SNR) 5,6 and detective quantum efficiency (DQE) 7,8 , and confirmed by observer performance studies such as receiver operating characteristics (ROC) 9,10 and contrast-detail (C-D) analyses 11,12 . In order to perform these observer performance studies accurately, many observers must interpret many images repeatedly, and these processes are tedious and expensive.…”