Classification of Nuclei in Colon Cancer Images using Ensemble of Deep Learned Features

Abstract

The classification of the cell nuclei in histopathological cancerous tissue images is a difficult problem due to the inhomogeneity of the cellular structures. In this study, ensemble learning and deep learning methods are used to reduce the difficulty of this problem. On the deep learning side, the features of histopathological images were extracted using convolution layers of convolutional neural networks. It has been suggested that these extracted features will be reduced with feature reduction methods and modeled with conventional machine learning algorithms. In order to get better results, the decisions of the learners were combined and a second proposition was made with ensemble learning. The proposed method in this study showed more robust results compared to the results of other studies in the literature. With this proposed approach, it is aimed to make more accurate analysis of cancer cell types and histopathological images.