Improving the Breast Cancer Image Classification using Autoencoders and CNN

Breast cancer is the most common disease among female rather than male, affecting 2.1 million women every year. Globally more than 70,000 women die from breast cancer every year. Deep learning architectures such as Convolution neural networks are mostly used for image classification. They fit well in classifying breast cancer images also. Several feature extraction mechanism have been already available. The CNN is also used for the feature extraction and for the image classification. For improving the classification accuracy, the Convolution auto encoders are used to extract the features and the output of the auto encoders are fed into the Convolution neural network. The objective of the work is to improve the classification accuracy by combining the feature extraction mechanism such as auto encoders along with the Convolution neural network for various types of breast cancer images like Histology images, Mammogram images and Sonogram images.