Breast Cancer Detection : A Review On Mammograms Analysis Techniques Boulehmi Hela, Mahersia Hela, Hamrouni Kamel Université Tunis El Manar, Ecole Nationale d’Ingénieurs de Tunis LR SITI – Signal, Image et Technologies de l’information Tunis, Tunisia Boussetta Sana, Mnif Najla Université Tunis El Manar, Faculté de Médecine de Tunis UR Epidémiologie et Imagerie Médicale Tunis, Tunisia Abstract— Breast cancer is the most common cancer among women over 40 years. Studies have shown that early detection and appropriate treatment of breast cancer significantly increase the chances of survival. They have also shown that early detection of small lesions boosts prognosis and leads to a significant reduction in mortality. Mammography is in this case the best diagnostic technique for screening. However, the interpretation of mammograms is not easy because of small differences in densities of different tissues within the image. This is especially true for dense breasts. This paper is a survey of the automatic early detection of breast cancer by analyzing mammographic images. This analysis could provide radiologists a better understanding of stereotypes and provides, if it is detected at an early stage, a better prognosis inducing a significant decrease in mortality. Keywords- breast cancer; early detection; breast density; mammograms analysis I. INTRODUCTION Breast cancer is the most common cancer among women worldwide and has, in addition, a high mortality rate. However, no study has been able to identify with certainty why every year, one million breast cancer cases are discovered and 400 000 women die. Besides, according to the SFSPM (French Society of sinology and breast pathology), 7% of women with breast cancer are under 40 years in 2010, while the rate was 5.6% in 2002 [3, 4, 5, 6] which weakened its connection with the menopause. Detected early, breast cancer is easier to treat, with fewer risks and reduces mortality by 25%. This early detection can be achieved by subjecting women at risk (mainly postmenopausal women) to a mammography every two years, since it takes about five years for a breast tumor to reach 1 mm, two years longer to reach 5mm and one or two years to measure 2 cm, large enough to detect by palpation [3, 4, 5, 6, 78]. In this paper we remind the importance of mammography in early detection of breast cancer in the second section and the different forms of breast abnormalities in the third section. The fourth section explains how the breast density can be considered as a risk factor and the fifth section is a review on the different techniques of mammograms analysis. II. MAMMOGRAPHY AND BREAST CANCER DETECTION A mammogram is an X-ray, particularly suited to woman breast to detect nodules whose presence may indicate the existence of breast cancer [7]. It should be noted that mammography does not always provide a definitive diagnosis on the presence or absence of cancer: it allows doctors to see if there is an abnormality in the breast. The radiologist is consequently led to analyze the mammogram and perform a physical examination (appearance of the skin and nipple). Other tests are then necessary to establish the diagnosis (breast ultrasound, breast MRI and sampling). The computer-aided diagnosis (CAD) improves accuracy in the interpretation of considered mammograms, early detection of possible tumor and distinguishing between benign and malignant ones [26, 81, 83]. The presence, for example, of clusters of microcalcifications (3 or more per cm 2 ) on a mammogram shows an early sign of breast cancer. However, their detection is not trivial because of their small size and their similarity to the breast tissue. III. DIFFERENT FORMS OF BREAST ABNORMALITIES There are several forms of abnormality that may affect breast tissue. These abnormalities are often classified into three families: the opacities, microcalcifications and architectural distortions: A. Masses These are space occupying lesions, seen on two different impacts. They are characterized by their shape (round, oval, lobulated, irregular), their contour (circumscribed, micro- lobulated, obscured, indistinct, spiculated) and density (high, medium, low fat) (Fig. 3). Breast cancers are never made of fat (radio-transparent) though they may trap grease. Lesions containing fat are: oil cysts, lipomas, the galactocele and mixed lesions (hamartoma). Mass containing fat is always benign. Figure 3 - Examples of masses [1] SSD'13 1569677207 1 2013 10th International Multi-Conference on Systems, Signals & Devices (SSD) Hammamet, Tunisia, March 18-21, 2013 978-1-4673-6457-7/13/$31.00 ©2013 IEEE