Detection of cancerous tissue in histopathological images using Dual-Channel Residual Convolutional Neural Networks (DCRCNN)

Computer-aided detection techniques to improve precision diagnostic capability and efficiency in the diagnosis process has been regarded as one of the most important topics in the field of computer vision. The medical imaging data with respect to a patient is primarily considered as one of the most important sources to derive the information regarding the biomarkers of a particular disease. But the successful detection of biomarkers requires the radiologist and the pathologist to have long term experience in this field. Therefore, the development of computer-aided detection is one of the primary concerns that need to be discussed. Moreover with the advent of Deep Learning and Artificial Intelligence, now the detection of anomalies and aneurysms in the medical imagery can become much more precise and efficient. Therefore this particular paper presents a dual-channel residual convolution neural network (CNN) model for the automated classification and detection of cancerous tissues in histopathological images. The proposed CNN model has been trained with 220,025 histopathological images and has achieved an overall accuracy of 96.475%, average recall of 95.72% and an average precision of 95.92% respectively.