Abstract Archives of the RSNA, 2005
SSA04-06
Optimization of Image Processing for Chest Radiographs at Low Tube Voltage for Digital Selenium Radiography in a Phantom Study
Scientific Papers
Presented on November 27, 2005
Presented as part of SSA04: Chest (Digital Chest Imaging)
Thomas M. Bernhardt MD, Abstract Co-Author: Nothing to Disclose
Ulrike Rapp-Bernhardt MD, Presenter: Nothing to Disclose
Horst Lenzen MS, Abstract Co-Author: Nothing to Disclose
Rainer Esseling MS, Abstract Co-Author: Nothing to Disclose
Friedrich W. Roehl PHD, Abstract Co-Author: Nothing to Disclose
Walter Leonhard Heindel MD, Abstract Co-Author: Nothing to Disclose
To optimize image processing parameters for chest radiographs at different tube voltages for digital selenium radiography.
All patterns of simulated interstitial lung disease (ground-glass, linear, reticular, miliary), nodules ( 1cm) and catheters were superimposed over an anthropomorphic chest phantom and images obtained at 150 kVp, 90 kVp and 70kVp. The same effective dose was used for each tube voltage. Three different image processings were applied and images printed on a laser imager. Four board-certified radiologists participated in this ROC-analysis.
216,000 observations were analyzed. Best detection performance was observed for almost all patterns for 150 kVp tube voltage (state-of-art): lung density 1.8, abdomen density 0.65, gamma lower limit 1.89, gamma upper limit 4.0, detail contrast enhancement 1.0, noise compensation 0.75 (processing I). However, detection performance with highest Az values of the whole study were found at 90 kVp for nodular ( 1cm and catheters were at least as high using processing II at 90 kVp compared with 150 kVp (processing I and II). Lowest Az values for all tube voltages and almost all patterns were found using image processing: lung density 1.5, abdomen density 0.3, gamma lower limit 1.5, gamma upper limit 4.0, detail contrast enhancement 1.0, noise compensation 1.0 (processing III). All three images processings did not have a significant impact for the detection of catheters and nodules (>1cm) over low attenuation areas (p>.05).
Different images processings must be used for chest radiographs obtained at lower tube voltages compared with high tube voltage in order to obtain overall optimal images quality at 90 kVp.
Bernhardt, T,
Rapp-Bernhardt, U,
Lenzen, H,
Esseling, R,
Roehl, F,
Heindel, W,
Optimization of Image Processing for Chest Radiographs at Low Tube Voltage for Digital Selenium Radiography in a Phantom Study. Radiological Society of North America 2005 Scientific Assembly and Annual Meeting, November 27 - December 2, 2005 ,Chicago IL.
http://archive.rsna.org/2005/4411106.html