Mass auto-detection in mammogram based on wavelet transform modulus maximum

Abstract

High accurate detection of mass in mammogram is critical for improving the performance and efficiency of computer-aided diagnosis (CAD) system. In this paper, we propose a novel approach to enhance the detection performance of mass in mammograms using Wavelet Transform Modulus Maximum (WTMM). First, hunt the region of interest (ROI) through the whole image and the ROI was approximately located by multi-threshold method. Then the contour of the ROI was extracted from the modulus image acquired by Wavelet Transform Modulus Maximum (WTMM) method. The region of interest was finally refined by the contour extracted. Experimental results indicate that the proposed method is able to detect not only isolate masses, but also the masses connected with the glandular tissues successfully. This technique could potentially improve the performance of CAD system and diagnosis accuracy in mammograms.