A novel framework for edge detection of microcalcifications using a non-linear enhancement operator and morphological filter

Abstract

Minute calcium deposits that appear as bright spots on mammograms are known as microcalcifications. Calcifications in ducts and lobules are common markers of malignancy; therefore its detection plays a vital role in early detection of breast cancer. This paper proposes a novel framework for edge detection of microcalcifications using a combination of non-linear enhancement operator and morphological filter. The contrast improvement of the region of interest for better visualization of mammographic abnormalities is obtained using the proposed non-linear operator. This enhancement operator is versatile in approach, producing promising results for dense as well as non dense breast tissues. The enhanced ROI is then convolved with the proposed mask of two stage iterative discrete morphological gradient operator with variable sized structuring elements leading to improved edge detection of microcalcifications.