Maryam Heidari Hratmh - Student, Faculty of Electrical Engineering, Islamic Azad University of opioid,
Seyed Alireza Mirzaee - Lecturer, Faculty of Electrical Engineering, Islamic Azad University of opioid,

Breast cancer is one of the most common causes of mortality among women considered. Early detection increases chances of survival of breast cancer, mammography is important to create a system to detect suspicious masses. The construction of systems to detect this type of cancer, such as in the fields of medical diagnosis systems is considered crucial. In this article and the implementation of two methods to automatically detect the type of mass is benign or malignant tumor in the breast that is presented. In the first method, the auto-thresholding segmentation to support regional growth by cellular neural networks (CNN) trained obtained. In the second proposal, part of cellular neural networks (CNN) parameters with genetic algorithm (GA) is determined.

Breast cancer cell growth, neural networks, genetic algorithms, neural networks