Share Email Print
cover

Proceedings Paper

Dose sensitivity of three methods of image quality assessment in digital mammography
Author(s): Johann Hummel; Marcus Kaar; Rainer Hoffmann; Heinrich Kaldarar; Friedrich Semturs; Peter Homolka; Michael Figl
Format Member Price Non-Member Price
PDF $14.40 $18.00

Paper Abstract

Image quality assurance is one of the key issues in breast screening protocols. Although image quality can always be improved by increasing dose this mechanism is restricted by limiting values given by the standards. Therefore, it is crucial for system adjustment to describe the dependency of the image quality parameters on small changes in dose. This dose sensitivity was tested for three image quality evaluation methods. The European protocol requires the use of the CDMAM phantom which is a conventional contrast-detail phantom, while in North America the American College of Radiology (ACR) accreditation phantom is proposed. In contrast to these visual test methods the German PAS 1054 phantom uses digital image processing to derive image quality parameters like the noise-equivalent number of quanta (NEQ). We varied the dose within the range of clinical use. For the ACR phantom the examined parameter was the number of detected objects. With the CDMAM phantom we chose the diameters 0,10, 0.13, 0.20, 0.31 and 0.5 mm and recorded the threshold thicknesses. With respect to the PAS 1054 measurements we evaluated the NEQ at typical spatial frequencies to calculate the relative changes. NEQ versus dose increment shows a linear relationship and can be described by a linear function (R = .998). Every current-time product increment can be detected. With the ACR phantom the number of detected objects increases only in the lower dose range and reaches saturation at about 100mAs. The CDMAM can detect a 50% increase in dose confidently although the parameter increase is not monotonous. We conclude that an NEQ based method can be used as a simple and highly sensitive procedure for weekly quality assurance.

Paper Details

Date Published: 3 March 2012
PDF: 8 pages
Proc. SPIE 8313, Medical Imaging 2012: Physics of Medical Imaging, 83134K (3 March 2012); doi: 10.1117/12.910917
Show Author Affiliations
Johann Hummel, Medical Univ. of Vienna (Austria)
Wilhelminenspital Vienna (Austria)
Marcus Kaar, Medical Univ. of Vienna (Austria)
Rainer Hoffmann, Medical Univ. of Vienna (Austria)
Heinrich Kaldarar, Wilhelminenspital Vienna (Austria)
Friedrich Semturs, Medical Univ. of Vienna (Austria)
Peter Homolka, Medical Univ. of Vienna (Austria)
Michael Figl, Medical Univ. of Vienna (Austria)


Published in SPIE Proceedings Vol. 8313:
Medical Imaging 2012: Physics of Medical Imaging
Norbert J. Pelc; Robert M. Nishikawa; Bruce R. Whiting, Editor(s)

© SPIE. Terms of Use
Back to Top