Mammography visual enhancement in CAD-based breast cancer diagnosis.

This paper presents a novel approach to detect and discriminate abnormal and cueing signatures in mammography through enhancing the imaging contrast. Partial gland and adipose tissues are removed, and thus, the visual effect of mammography will be enhanced. Inspired by single image haze removal, we remove the majority of background tissues by introducing the idea of image matting. Experimental results show the feasibility and performance on distinguishing focuses from healthy tissues in the enhanced mammography. The method has potential applications on breast cancer diagnosis in computer-aided detection.