Share Email Print

Proceedings Paper

Exploring the relationship between SDNR and detectability in dual-energy breast x-ray imaging
Author(s): Roshan Karunamuni; Swathiu Kanamaluru; Kristen Lau; Sara Gavenonis; Predrag R. Bakic; Andrew D. A. Maidment
Format Member Price Non-Member Price
PDF $17.00 $21.00

Paper Abstract

Contrast-enhanced (CE) digital breast tomosynthesis (DBT) provides a technique to increase the contrast of radiographic imaging agents by suppressing soft-tissue signal variation. By reducing the effect of the soft-tissue anatomical noise, it is then possible to quantify the signal from an iodinated contrast agent. The combination of dual-energy and tomographic acquisitions allows for both the accurate quantification and localization of an iodinated lesion. Here, we present our findings demonstrating the relationship that exists between the signal difference to noise ratio (SDNR) and reader detectability of iodinated lesions in a physical anthropomorphic phantom. The observer study was conducted using the ViewDEX software platform with a total of nine readers. The readers were asked to score each of the iodinated lesions on a scale from 1 (entire boundary and area are visible) to 5 (not visible). Both SDNR and lesion detectability were found to improve as the concentration of the iodine increases, and the thickness of the phantom decreases. Lesion detectability was better in the tomographic slice that best matches the focal plane of the imaged object. However, SDNR does not significantly change with focal plane. Our results demonstrated that observer lesion detectability correlated well with SDNR. Lesions whose SDNR fell below 1 were difficult to distinguish from the background and were in general not visible. Lesions that were rated entirely visible corresponded to those with SDNR values above 3. Lesions with intermediate SDNR values were visualized but not confidently from the surrounding background. These threshold SDNR values can be used to optimize the imaging parameters in CE-DBT.

Paper Details

Date Published: 19 March 2013
PDF: 9 pages
Proc. SPIE 8668, Medical Imaging 2013: Physics of Medical Imaging, 866863 (19 March 2013); doi: 10.1117/12.2008114
Show Author Affiliations
Roshan Karunamuni, Univ. of Pennsylvania (United States)
Swathiu Kanamaluru, Drexel Univ. (United States)
Kristen Lau, Univ. of Pennsylvania (United States)
Sara Gavenonis, Univ. of Pennsylvania (United States)
Predrag R. Bakic, Univ. of Pennsylvania (United States)
Andrew D. A. Maidment, Univ. of Pennsylvania (United States)

Published in SPIE Proceedings Vol. 8668:
Medical Imaging 2013: Physics of Medical Imaging
Robert M. Nishikawa; Bruce R. Whiting; Christoph Hoeschen, Editor(s)

© SPIE. Terms of Use
Back to Top
Sign in to read the full article
Create a free SPIE account to get access to
premium articles and original research
Forgot your username?