Classification of histopathological breast cancer images by deep learning methods

Processing of visual data has been used frequently in recent years, especially in the medical field. By using deep learning methods, a sub-branch of artificial intelligence, together with image processing methods; It has been successfully applied in many areas, especially tissue analysis and automatic classification of diseases. Deep learning experienced a stagnation due to the inadequacy of the existing hardware for a while, but it started to become a popular field again with the 2000s. With the developments in GPU, there has been a transition from shallow artificial neural networks to deep artificial neural networks. Thanks to back propagation algorithms between layers, hidden and distinctive features in large data sets are discovered. This reveals the effectiveness of artificial neural networks and deep learning methods. In this study, histopathological breast cancer images in the BreakHis dataset were classified using image processing and deep learning methods. At the same time, studies in the literature were also examined. The aim of the thesis is to present an automatic method for the early and accurate diagnosis of breast cancer, which is very common in women in recent years. As a result of the studies carried out in this way, a study has been put forward that will support medical professionals in their decisions. Keywords: Image Processing, Artificial Intelligence, Deep Learning, Artificial Neural Networks, Histopathological Images, Breast Cancer.