Share Email Print
cover

Proceedings Paper

Optimized AEC for scanning digital mammography based on local variation of scan velocity
Author(s): Magnus Aslund; Bjorn Cederstrom; Mats Lundqvist; Mats Danielsson
Format Member Price Non-Member Price
PDF $14.40 $18.00

Paper Abstract

In mammography, there is an optimal photon energy and current time product that produce the required image quality at the minimal dose. The task of an automatic exposure control (AEC), in full field digital mammography (FFDM) is to minimize the dose by using optimized exposure settings. Each point in a mammogram has different radiological thickness. A conventional AEC samples the thickness in some regions to set the current time product and possibly also the beam quality. We define an ideal AEC as one that optimizes the beam quality and exposure in each point to produce a constant contrast-to-noise ratio (CNR) of structures of interest throughout the image. This paper presents the results from a theoretical evaluation of an AEC proposed for a scanning photon-counting FFDM system. The geometry enables the AEC to use information from the leading detector line to adjust the scan velocity during the scan. Thus, the irradiation can be better optimized in the scanning-direction as compared to a conventional AEC. The scan time is further reduced by increased velocity over sections that contain no tissue. The results are quantified in terms of reduction of entrance dose and scan time. The presented AEC is compared to an ideal AEC, a conventional AEC and is also benchmarked against an ideal regulator. The effect of the detector width is evaluated. Compared to a conventional AEC, both evaluated on a set of 266 mammograms, the ideal AEC would reduce the entrance dose by 39% on average while the proposed AEC for scanning systems reduces the entrance dose by 10-20% and scan-time by 25-32% on average, depending on detector width.

Paper Details

Date Published: 20 April 2005
PDF: 10 pages
Proc. SPIE 5745, Medical Imaging 2005: Physics of Medical Imaging, (20 April 2005); doi: 10.1117/12.596667
Show Author Affiliations
Magnus Aslund, Royal Institute of Technology (Sweden)
Bjorn Cederstrom, Royal Institute of Technology (Sweden)
Mats Lundqvist, Sectra Mamea AB (Sweden)
Mats Danielsson, Sectra Mamea AB (Sweden)


Published in SPIE Proceedings Vol. 5745:
Medical Imaging 2005: Physics of Medical Imaging
Michael J. Flynn, Editor(s)

© SPIE. Terms of Use
Back to Top