Cancer prognosis prediction models based on deep learning

Abstract

Cancer is a hot issue that more and more people are concerned about at present. In recent years, the incidence of cancer in the world has been rising, but the treatment methods for cancer have been lacking in specificity. Some researchers have proposed the use of deep learning algorithms to assist cancer treatment, which can be analyzed according to the characteristics of patients, and has made progress in the prediction and treatment of a diverse of different cancers. However, there are few articles summarizing the application of deep learning in cancer prognosis. Therefore, this paper summarizes some advances in deep learning in the field of cancer prognosis in recent years, and discusses the limitations of deep learning and possible future development directions. The research believes that the current limitations of deep learning mainly come from the characteristics of data sets and algorithms themselves, and proposes their own solutions to the problems.