Wavelets in the computed evaluation of mammographic accreditation phantom images

Abstract

This paper presents a novel approach to the computed assessment of a mammographic phantom device. The approach shown here is fully automated and is based on the automatic selection of the region of interest, in the use of the discrete wavelet transform (DWT) and morphological operators to assess the quality of the American College of Radiology (ACR) mammographic phantom images. The algorithms developed here have succesfully scored 30 images obtained with different combinations of voltage applied to the tube and exposure and could notice the differences in the radiographs due to the different level of exposure to radiation. © 2013 Springer-Verlag.