Statistical discrimination between mammary cancer and mastopathy / Diana Lucía Londoño Londoño.

Doctor en Estadística.

Breast cancer is currently considered a high frequency pathology in the world and one of the causes of higher mortality among women. According to recent statistical data published by the World Health Organisation (WHO), breast cancer accounts for 23% of all cancer related cases and 14% of all cancer related deaths among women worldwide (Zhou et al., 2017). Cancer is a group of diseases that cause cells in the body to change and grow out of control (Albertson and Pinkel, 2003). However, if breast anomalities are detected and diagnosis are made at early stages, studies show that the chances of survival can be greatly improved (Kestener et al., 2001; Martin et al., 1979). Currently the most reliable imaging technique for the detection of such anomalities is Mammography, or X-ray examination, that also plays an important role in control during and after the treatment. The early detection of breast cancer in asymptomatic women using breast screening mammography is currently the most effective imaging technique for the detection of such anomalities to reduce the morbidity and mortality associated with breast cancer (Lauby-Secretan et al., 2015). However, high imprecision and false / rate diagnostics makes problem hard. Mammographies are low dose X-ray projections of the breast, and it is the best method for detecting cancer at the early stage. One of the stages present in the analysis of these mammography views involves the identification and classification of breast lesions, such as breast masses and micro-calcifications. Microcalcification is an effective indicator of early breast cancer (Wang et al., 2016).This identification and classification is usually performed manually by a radiologist, who should give a more accurate diagnosis. Furthermore, efficacy is often highly correlated with radiologist expertise and workload. The interpretation of mammograms is however not an easy task. The mammographic appearance of normal tissue is highly variable and the radiological findings associated with breast cancer can be very complex. On the other hand, it has been estimated that 10−30% of cancers which could have been detected are missed (Martin et al., 1979) and a high percent of the patients that are called back at screening turn out not to have cancer (Kestener et al., 2001)...