Automatic scoring of CDMAM using a model of the recognition threshold of the human visual system: R*

Author

Prieto, Gabriel ; Chevalier, Margarita ; Guibelalde, Eduardo

Author_Institution

Dept. Radiologia, Univ. Complutense de Madrid, Madrid, Spain

fYear

2009

fDate

7-10 Nov. 2009

Firstpage

2489

Lastpage

2492

Abstract

A software tool is presented for the automatic evaluation of the CDMAM phantom images that are currently used for the quality assessment of the image quality in mammography. This software tool is based on the use of the cross-correlation component of the index MS-SSIM*, R*, oriented to the recognition threshold of different image representations. Given an image sequence, whose images begin as unrecognizable and are gradually refined to include more information, the recognition threshold corresponds to first the image in the sequence in which an observer identifies the content. We have validated our software tool by comparing our readouts with those obtained by a total of 4 expert observers in the evaluation of 8 CDMAM images. The correlations obtained between both readout sets are better than 0.99 and the range of useful correlation comprises diameters from 0.16 mm to 2.0 mm of the gold disks inside the CDMAM.

Keywords

image recognition; image representation; image sequences; CDMAM phantom images; automatic scoring; cross-correlation component; human visual system; image representations; image sequence; recognition threshold; software tool; Humans; Image quality; Image recognition; Image representation; Image sequences; Imaging phantoms; Mammography; Quality assessment; Software tools; Visual system; CDMAM phantom; Human visual system modeling; MS-SSUN* index; digital mammography; quality control;

fLanguage

English

Publisher

ieee

Conference_Titel

Image Processing (ICIP), 2009 16th IEEE International Conference on

Conference_Location

Cairo

ISSN

1522-4880

Print_ISBN

978-1-4244-5653-6

Electronic_ISBN

1522-4880

Type

conf

DOI

10.1109/ICIP.2009.5413924

Filename

5413924