• A new era in colorectal cancer: Artificial Intelligence at the forefront

    Aram Farhoudian, Arash Heidari, Reza Shahhosseini
    Comput Biol Med. 2025 Sep;196(Pt C):110926. doi: 10.1016/j.compbiomed.2025.110926. Epub 2025 Aug 15.

    Abstract

    Background: Colorectal Cancer (CRC) is the third most common malignancy and the second leading cause of cancer death globally. As radiology interlaces with various medical fields, the progress of Artificial Intelligence (AI) in radiology has impacted diverse areas of medicine, including CRC. Utilizing AI such as Machine Learning (ML), Deep Learning (DL), Explainable AI (XAI), and hybrid models in CRC screening leads to a notable decrease in the occurrence and mortality of CRC patients. Incorporation of bioinformatics tools within AI aids in screening and identifying a more significant number of CRC biomarkers. 

     Methods: Due to this demand, an enormous amount of research has been conducted. Therefore, in this study, we utilized a Systematic Literature Review (SLR) to encompass all facets of findings from relevant articles. Inspection of genetic factors inspection, CRC early detection, late-stage CRC rapid prediction, treatment plan selection, metastatic biomarkers discovery, CRC types classification, CRC risk prediction, and CRC survival rate prediction are the critical uses of applications employed in the CRC field. 

     Results: Random Forest (RF), Support Vector Machine (SVM), Conventional Neural Network (CNN), and other AI models are frequently used in such scenarios. A comprehensive assessment was conducted on the diverse issues and obstacles associated with implementing AI applications for this particular disease. According to the data, most papers are evaluated mainly on accuracy, delay time, data privacy, robustness, and dataset availability. ML models are employed in 50 % of the papers, while DL models account for 23.3 %. Furthermore, XAI is utilized in 10 % of cases, and hybrid models are implemented in 16.7 % of papers. 

     Conclusions: Inspection of genetic factors, early CRC detection, late-stage CRC rapid prediction, treatment plan selection, metastatic biomarkers discovery, CRC types classification, CRC risk prediction, and CRC survival rate prediction highlight the critical contributions of AI in the CRC field.