Abstract: Primary liver cancer, cholangiocarcinoma, and gallbladder cancer are highly prevalent and fatal forms of malignant tumors of the digestive system worldwide. Radical surgical resection is the core treatment approach for these patients to achieve long-term survival. Hepatobiliary surgical procedures are characterised by intricate anatomical structures and a high prevalence of anatomical variations. Not only is the risk control during the operation extremely difficult, but the diagnosis and treatment decisions also highly rely on the clinical experience of the surgeon. This has thereby become the key bottleneck restricting the development of hepatobiliary surgery towards more precise and minimally invasive techniques. In recent years, artificial intelligence （AI） technology, centered on deep learning, computer vision, and machine learning, has witnessed rapid development and has deeply permeated the entire diagnosis and treatment process of hepatobiliary tumor surgery, covering preoperative assessment and clinical decision-making, intraoperative intelligent assistance and safety guarantee, postoperative outcome prediction, and perioperative management, as well as the skills training and standardized training of surgeons. This review discussed the application status and research progress of AI technology in the field of hepatobiliary tumor surgery, analyzed the core challenges faced by current technology in clinical implementation, and put forward to future development directions, aiming to provide reliable theoretical references and practical ideas for the intelligent development of hepatobiliary tumor surgery.

Key words: Artificial intelligence, Hepatobiliary tumor, Surgery, Deep learning, Computer vision