Microcalcification detection using wavelet transform

Abstract

This paper presents a new method for detection of microcalcification using wavelet transform based on statistical methods. Digitized mammograms are decomposed using the wavelet transform without down sampling process at several levels in the transform space. In order to improve the contrast enhancement of images, the multiscale adaptive gain as an enhancement method was applied. Skewness, kurtosis and boxplot outlier were applied as detection method of the previous modification image with a specific size of region of interest. We have simulated this algorithm by using 30 variations of images as part of 18 digitized mammograms. Preliminary results show visually that applied detecting method has 96% in an effectiveness level.