Two decades of artificial intelligence in colon cancer diagnosis and treatment: a bibliometric analysis of research trends (2003-2023).

This study aims to analyse research trends, identify key areas of focus, and explore developmental patterns in AI applications for colon cancer diagnosis and treatment from 2003 to 2023. A systematic search of the Web of Science Core Collection database was conducted to identify relevant studies. Bibliometric analysis was performed using VOSviewer to visualize collaborations among countries (regions), institutions, and authors, as well as author cocitations, keyword co-occurrences, and overlay analyses. CiteSpace enabled institutional betweenness centrality analysis, journal dual-map overlay analysis, cluster analysis of cocited literature from the past five years, timeline visualization of cocited literature clusters, and burst detection analysis of references. Microsoft Excel was used to create bar charts of publication volumes and descriptive analysis tables for countries (regions), institutions, journals, authors, cocited authors, cited references, and keywords. The analysis included 1456 publications, revealing a consistent upwards trend in annual publication volume from 2003 to 2023, with a sharp increase from 2020 and a peak in 2023. China was the most productive country (region), the Chinese Academy of Sciences the leading institution, and Mori Yuichi et al. the leading author. Jemal A, Siegel RL, and Kather JN were identified as the most influential researchers based on cocitation analysis. Cancers published the most articles, while Gastroenterology received the highest number of cocitations. Citing journals focused predominantly on the “Molecular, Biology, Immunology” and “Medicine, Medical, Clinical” domains, while cited journals focused on the “Molecular, Biology, Genetics” and “Health, Nursing, Medicine” fields. The most frequently cocited reference was Global Cancer Statistics 2020: GLOBOCAN Estimates of Incidence and Mortality Worldwide for 36 Cancers in 185 Countries. Keyword clustering revealed four main research areas: AI-assisted treatment and prognosis prediction, endoscopic diagnosis, pathological diagnosis, and biological research in colon cancer. Current research hotspots include deep learning, convolutional neural networks, radiomics, gastrointestinal endoscopy, pathology, and immunotherapy. This bibliometric analysis highlights the expanding role of AI in colon cancer research, with growing interest from the scientific community. AI applications span various aspects of colon cancer management, including biology, diagnosis, staging, efficacy assessment, and prognosis prediction. These findings provide valuable insights for researchers and clinicians working at the intersection of AI and colon cancer.