Histopathological Cancer Detection Using CNN

Histopathology involves the examination and analysis of diseases associated with human or animal tissues. It mainly comprises scrutinizing a batch of tissues or cells under a microscope, and it is one of the most frequently employed tools in the detection and diagnosis of cancer. The detection of metastatic cancer in pathological scans is an onerous and wearisome task. The possibility of melding artificial intelligence (AI) with this field holds immense potential. In this chapter, we attempt to build a system using a diverse ensemble of convolutional neural network (CNN) techniques, namely Inception Net V3 (or Inception V3), Xception Net, ResNet, and DenseNet, that can accurately analyze digital scans of tissues and identify traces of metastatic cancer in them. Further, based on the detailed dissection of the results, we aim to isolate the best technique from the ensemble. With this research, we aim to find a method that can surpass state-of-the-art methods tackling the same issue.