Classification of microcalcifications on digital mammograms based of FPGA

Abstract

Breast cancer is a disease that is the death cause of many women around the world [2] and the best way to treat this disease is by detecting it in its early stages and starting the treatment right away [6]. As computers can be used to manipulate images, one way to aid radiologists in breast cancer early diagnosis is by the detection and classification of microcalcifications on digital mammograms, small calcium accumulations [3] that can be the first sign of a tumor that cannot yet be detected by palpable exams. Michele Le Gal developed a classification scheme to aid radiologists to determine whether a breast tumor is malignant or not without the need for surgeries [5], reducing the need for therapies and recurrent treatment [6]. This classification scheme is based on the shape of microcalcifications, and Figure 1 shows samples from each type of microcalcification, while Table 1 shows the percentage of malignant cancer associated with each type of microcalcification.