Paper Title
Convolutional Neural Networks for Automated Diagnosis of Breast Cancer

Abstract
Now-a-days Breast Cancer is the fast growing and commonly occurring disease in women and it is the main reason for rise in death rate globally. Among these deaths, many are due to the lack of proper diagnosis and early detection. This death rate can be decreased by accurate prediction of the stage of cancer. To identify the stage of cancer, histological samples are examined under the microscope which takes more time and thus automated systems are used for fast diagnosis. Automated systems are playing a crucial role in the field of human health care. The main aim of these techniques is to reduce human error in judgment for diagnosis and thus to decrease the human death rate. For the past few years, deep learning has been used in health care for disease diagnosis. In this paper, Convolutional Neural Network (CNN) is used to classify the type of the cancer. A dataset of 7909 breast cancer histopathology images obtained on 82 patients is taken from BreakHis Database. The dataset consist images of two different classes benign and malignant each with different magnifications namely 40X, 100X, 200X and 400X. We have proposed two different architectures to train and extract the patches of the image. An image is given as input to classify the type of cancer. The 40X magnification level has acquired highest accuracy of 74.68% compared with other architectures and other magnification levels. Keywords - Convolutional Neural Networks, Breast Cancer, Deep learning, Image Classification