Automatic evaluation of the image quality of a mammographic phantom

Abstract

[EN] In this work a method has been developed to analyse the digital image quality of a mammographic phantom by means of automatic process techniques. The techniques used for the digital image treatment are standard techniques as the image thresholding to detect objects, the regional growing for pixels pooling and the morphological operator application to determine the objects shape and size, etc. This study allows the obtention of information about the phantom characteristics, that due to its small size and lowly contrast can be obtained very difficultly by direct observation. The final aim of this work is to obtain one or more parameters to characterize the reference phantom quality image in an objective way. These parameters will serve to compare images obtained at different mammographic centers and also, to study the temporal evolution of the image quality produced by determined mammographic equipment.