A Systematic Review of Machine Learning Prediction Models for Colorectal Cancer Patient Survival Using Clinical Data and Gene Expression Profiles.

Colorectal cancer persistently ranks among the top causes of cancer-related mortality globally. The development of superior predictive methodologies is imperative for augmenting survival outcomes. This systematic review, conducted in accordance with PRISMA-P guidelines, scrutinizes studies carried out between 2013 and 2023 that apply machine learning models to prognosticate survival in colorectal cancer patients, particularly those models incorporating clinical data and gene expression profiles. Criteria for inclusion comprised studies employing machine learning techniques, with specific emphasis on those integrating clinical data and gene expression profiles for predictive purposes. Studies devoid of explicit methodological delineation or not written in English were excluded. Decision trees, neural networks, and support vector machines emerged as the most frequently scrutinized models in the review. While some models manifested high accuracy, others underscored areas requiring refinement. Predominant data sources included patient clinical records, gene expression datasets, and molecular profiling. The results underscore the potential of machine learning in bolstering predictive precision, thereby implicating a trajectory for future research targeting the optimization of patient prognosis and treatment outcomes in colorectal cancer. [ABSTRACT FROM AUTHOR]