International Journal For Technological Research In Engineering Volume 3, Issue 10, June-2016 ISSN (Online): 2347 - 4718 www.ijtre.com Copyright 2016.All rights reserved. 2528 A REVIEW ON PREPROCESSING AND SEGMENTATION TECHNIQUES FOR THE DETECTION OF MICRO- CALCIFICATIONS IN MAMMOGRAMS Ramandeep Kaur 1 , Rashwinder Singh 2 , Navneet Kaur Mavi 3 1 Department of Computer Engineering, Punjabi University Patiala 2 Assistant Professor, Department of Electronics and communication Engg CGC-CEO Landran 3 Assistant Professor, Department of Computer Engineering, Punjabi University Patiala Abstract: Mammography is a popular method for the detection of breast cancer and removal of primary tumor. To monitor and control breast cancer breast cancer segmentation is required. Many researchers worked in this area and it is still a challenging problem. In this paper we present different techniques used for preprocessing of a mammogram and various segmentation algorithms to detect the occurrence of micro-calcifications in digital mammograms from mini-MIAS database. The images of mammograms are noisy and of low contrast and is divided into various regions. We introduce a simple approach from enhancement of mammogram to preprocessing and then followed by segmentation of cancerous region. I. INTRODUCTION Breast cancer is a leading cause of death among all cancer diseases for middle-aged and older women. In U.S. Breast cancer death rates are higher than that of any other cancers for women. Mammography is a well-known method for detection of breast tumors. Early detection and removal of the primary tumor is an essential and effective method to enhance survival rate and reduce mortality for the radiologists. Early and efficient detection is most effective way to reduce morality, and currently a screening program based on mammography is considered one the best and popular approach for the detection of breast cancer. Mammography is a low-dose X-ray produced that visualize the internal structure of breast on a mammogram. On average, mammography will detect about 80-90 percent of breast cancers in women without symptoms. The first digital mammography system received U.S. Food and Drug Administration (FDA) approval in year 2000. It has been proposed that a computer-aided diagnostic (CAD) system be used as a second reader to assist the radiologist, leaving the final decision to the human. CAD systems are very helpful for radiologists to study mammograms. CAD can increase the diagnostic accuracy and efficiency with high reproducibility. The abnormalities in mammograms can be divided into two types: (1) micro-calcifications or (2) masses. Calcifications are tiny mineral deposits or calcium within the breast tissue. They look like small white spots on a mammogram. The calcifications may be of different types and may differ in distribution. A mass is usually something a little more substantial and clearer than a lesion. Specifically, a mass has volume and occupies space. On a mammogram, it tends to be denser in the middle than towards the edges. Masses may also have different shapes and margins, may differ in size, location, and orientation, and may have different backgrounds. If the mass appears more like a lobule than a purely round or oval shape, then it is somewhat more suspicious for breast cancer. Masses with irregular shapes are highly suspicious for breast cancer. There is another type of breast cancer known as speculated lesions. Spiculated lesions have a central tumor mass that is surrounded by a radiating pattern of linear spicules. Most spiculated lesions are malignant. The first step in mammography is the selecting the image from the database. The two databases can be used: mini-MIAS database and DDSM database. The next step is preprocessing of mammographic image which is used for noise removal and enhancement of image. After that the region of interest is detected by using segmentation and clustering of various parts of the mammogram. Fig.1 Block diagram of the proposed mammographic micro- calcifications detection scheme II. LITERATURE REVIEW In mammograms there are two main views: The first is MLO view that is Medio-lateral Oblique view and other is CC view that is Cranio-caudal view. The MLO view is used in preprocessing step because it presents the whole view of mammogram whereas some information may be lost in CC view [8]. It is important to enhance the mammographic image to remove noise and other artifacts in preprocessing step. Mario and Mislav [1] presented an approach for the