Advances in Artificial Intelligence for Precision Preoperative Assessment in Liver Surgery:A Comprehensive Review

Shi-yi Qian^1,2*Jiang Haitao^1,2*Ya-Hui Liu^2,3

• ¹Department of General Surgery, Ningbo No 2 Hospital, Ningbo, China
• ²School of Clinical Medicine, Hangzhou Medical College, Hangzhou, China
• ³Department of Scientific Research, Ningbo No 2 Hospital, Ningbo, China

The success rate of liver surgery is closely related to patients' individualized characteristics such as lesion benignity and malignancy, vascular variability, and liver function reserve, and the traditional assessment relies on surgeons' experience, which has the limitations of high subjectivity and low reproducibility. With the rapid development of artificial intelligence (AI) technology, especially the application of multimodal data fusion and deep learning, the goal of accurate decision-making and personalized medicine in liver surgery has been greatly promoted, facilitating the shift from empirical decision-making to accurate prediction of data. However, most of the current studies are small-sample, single-center data and there is no standardized liver imaging database, which is associated with data bias and overfitting of AI algorithms. In addition, the complexity of the liver structure and the dynamics of vascular variation greatly increase the difficulty of intraoperative monitoring. This paper systematically describes the key technologies and assessment areas of AI in preoperative assessment of liver surgery, focusing on the limitations of the current study and the future development direction, with the aim of providing a reference for the precise preoperative assessment of clinical AI, with a view to achieving better surgical results, improving the quality of postoperative recovery and the overall therapeutic effect of the patients.