2006
DOI: 10.1117/12.650976
|View full text |Cite
|
Sign up to set email alerts
|

Mammographic texture synthesis using genetic programming and clustered lumpy background

Abstract: In this work we investigated the digital synthesis of images which mimic real textures observed in mammograms. Such images could be produced in an unlimited number with tunable statistical properties in order to study human performance and model observer performance in perception experiments. We used the previously developed clustered lumpy background (CLB) technique and optimized its parameters with a genetic algorithm (GA). In order to maximize the realism of the textures, we combined the GA objective approa… Show more

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
4
1

Citation Types

0
9
0

Year Published

2007
2007
2008
2008

Publication Types

Select...
1
1

Relationship

1
1

Authors

Journals

citations
Cited by 2 publications
(9 citation statements)
references
References 17 publications
0
9
0
Order By: Relevance
“…Lumpy backgrounds [22] and clustered lumpy backgrounds [23] (CLBs) were developed in order to better reproduce anatomical variations. A new version of CLBs [24] was specifically developed for mammographic textures. Yet aside from the visual similarity and commonality of global image statistics between the CLBs and the real mammographic backgrounds, it is still unknown whether human observer strategies are similar across both types of backgrounds.…”
Section: Introductionmentioning
confidence: 99%
“…Lumpy backgrounds [22] and clustered lumpy backgrounds [23] (CLBs) were developed in order to better reproduce anatomical variations. A new version of CLBs [24] was specifically developed for mammographic textures. Yet aside from the visual similarity and commonality of global image statistics between the CLBs and the real mammographic backgrounds, it is still unknown whether human observer strategies are similar across both types of backgrounds.…”
Section: Introductionmentioning
confidence: 99%
“…2,800 256 by 256 pixels square ROIs were manually selected, and resampled to 154 by 154 pixels, in order to emulate a magnification factor of 1.5 on the display screen, reproducing typical clinical settings [9]. The CLB had been designed in order to mimic digital mammograms ROIs, and their statistical and visual properties assessed by radiologists in a previous study [8].…”
Section: Backgroundsmentioning
confidence: 99%
“…When studying mass detection in a typical radiological task, use of real backgrounds and masses is an option for getting closer to reality, but the collection of hundreds or thousands of similar images can be difficult and time-consuming. Therefore, mammographic non-stationary backgrounds are most of the time replaced by white noise [1,2], power-law filtered white noise [3][4][5], or lumpy backgrounds [6][7][8][9]. Similarly, masses are generally approximated by disks [2,3,10], phantoms elements [9], Gaussian or Gabor functions [11][12][13].…”
Section: Introductionmentioning
confidence: 99%
See 2 more Smart Citations