Artificial Intelligence and Radiomics in Primary Liver Cancer Imaging: A Bibliometric and Visualized Analysis

Background: Combining artificial intelligence (AI) with radiomics for primary liver cancer (PLC) enhances diagnostic precision, sharpens risk stratification, and facilitates personalized treatment. The study aims to conduct a bibliometric analysis of this field, explore its research status and emerging hotspots, and provide data support and academic insights for subsequent research. Methods: A bibliometric analysis of 2890 publications on PLC, AI, and radiomics from 2008 to 2025 was performed, using data retrieved from the Web of Science Core Collection (WoSCC) and Scopus, followed by manual screening and deduplication. The finalized dataset was analyzed and visualized using tools such as VOSviewer, CiteSpace, and R to examine trends in annual publication counts, the geographic distribution of research, the institutions involved, journals, authors, references, and keywords. Results: were identified as the most productive and influential journals, respectively. The most frequently occurring keywords were "hepatocellular carcinoma", "deep learning", and "magnetic resonance imaging", while "image reconstruction", "liver cancer classification", and "deep supervision" have emerged as prominent recent research hotspots. Conclusion: Applications of AI and radiomics in imaging for PLC are gaining increasing attention. Future trends are expected to focus on enhancing algorithmic accuracy and advancing clinical prediction of microvascular invasion, postoperative outcomes after hepatectomy, and the effectiveness of transarterial chemoembolization.