Breast Cancer Detection using Histopathology Images

One of the leading causes of death among women in the world is breast cancer and to make sure that the treatment process is effective and the survival rates could rise, the state of the disease should be diagnosed in time and rightly. The manual analysis of histopathology images under the scope of viewing them through a manual is excruciatingly time-consuming and prone to human error. In the recent years, the use of AI and DL methods can be used to automatize the process of diagnostics and ensure positive outcomes. The study will be focused on developing a deep learning model that will recognize and classify breast cancer histopathology images. The algorithm of the spatial and textual features extraction of the images is performed by the means of convolutional neural networks (CNNs), thus, it does not involve any hand-crafted feature engineering. The model is trained and tested on publicly available datasets, besides being able to make recommendations with considerable confidence about the diagnosis of benign and malignant specimens. The model provides diagnostic support to the pathologists that is quick and repeatable as well as consistent. As an example of potential of DL, the current paper discusses analysis of images as a supplement to computer-aided diagnosis of breast cancer and the automation of pathological processes.