Research in Computing Science, Vol. 131, pp. 131-141, 2017.
Abstract: Primary prevention in early stages of breast cancer becomes crucial for diagnosis and, at the same time, complicated to achieve since the causes remain practically unknown. X-ray mammography is the most widely used screening technique due to cost effectiveness and its capability to provide valuable structural information thus allowing detection of characteristic cancer signatures such as masses and microcalcifications. Recently, thermography has gained a lot of interest since it has been demonstrated to be a non-invasive technique capable of revealing the health condition of the breasts in terms of physiological changes due to cancer formation. In this paper an approach based on intensity-based segmentation by means of morphological operators is proposed in order to detect regions potentially containing cancer in digital mammograms and thermograms. The algorithm is tested over several images taken from the Digital Database for Screening Mammography and the American College of Clinical Thermography, and the results suggest that the proposed algorithm is a suitable tool to successfully identify and extract regions of interest in a variety of environments and conditions, among which are several types of cancer and different image angles.
PDF: Digital Mammography and Thermography Processing for Breast Cancer Detection
PDF: Digital Mammography and Thermography Processing for Breast Cancer Detection