Abstract

Ever since the world entered the age of information, scientists have looked into the developments and applications of the growing prospects of machine learning and neural networks. In particular, the ability for deep learning machines to determine the risk, survivability, and prognosis of tumors based on medical cancer databases has intrigued healthcare researchers seeking to improve these algorithms in recent years. There are distinct aspects of medical procedures where artificial intelligence (AI) training can be applied; for example, the calculation of risk scores for patients based on mammographic screening, analysis of the presence of biomarkers like spermine and other polyamines in fluids surrounding tumors, genomic and epigenetic assessments to map genes that influence cancer expression, as well as the utilization of metabolomic data from FTIR spectroscopy of a patient’s biofluids to help make a more reproducible and conclusive diagnosis. The goal of this review is to discuss the progress of AI and deep learning in clinical procedures and applications in recent years and evaluate the efficacy of certain AI methods for tumor diagnosis, prognosis, and prediction based on patient information from available medical databases.