Prediction of Cancer Treatment Using Advancements in Machine Learning

Abstract

Abstract: Many cancer patients die due to their treatment failing because of their disease's resistance to chemotherapy and other forms of radiation therapy. Resistance may develop at any stage of therapy, even at the beginning. Several factors influence current therapy, including the type of cancer and the existence of genetic abnormalities. The response to treatment is not always predicted by the existence of a genetic mutation and might vary for various cancer subtypes. It is clear that cancer patients must be assigned a particular treatment or combination of drugs based on prediction models. Preliminary studies utilizing artificial intelligence-based prediction models have shown promising results. Building therapeutically useful models is still difficult despite enormous increases in computer capacity due to the lack of adequate clinically important pharmacogenomics data. Machine learning is the most widely used branch of artificial intelligence. Here, we review the current state in the area of using machine learning to predict treatment response. In addition, examples of machine learning algorithms being employed in clinical practice are offered.