Histopathological analyses of breast cancer using deep learning

Deep Learning hosts a plethora of variants and models in Convolution Neural Networks (CNN), where the prudence of these methods is algorithmically proven when implemented with sturdy datasets. Much number of haphazard structures and textures are found in the histopathological images of breast cancer, where dealing with such multicolor and multi-structure components in the images is a challenging task. Working with such data in wet labs proves clinically consistent results, but added with the computational models will improvise them empirically. In this paper, we proposed a model to diagnose breast cancer using raw images of breast cancer with different resolutions, irrespective of the structures and textures. The floating image is mapped with the healthy reference image and examined using different statistics such as cross correlations and phase correlations. Experiments are carried out with the aim of establishing the optimal performance on histopathological images. The model attained satisfactory results and are proved good for decision making in cancer diagnosis.