Purpose/Significance To explore the research topics and development trends of medical artificial intelligence (AI) in the past decade (2014—2023), and to provide references for related research decisions. Method/Process 152 398 entries of medical AI-related literatures from 2014 to 2023 are retrieved from PubMed and Web of Science databases. The BERTopic model is used to mine the research topics of the literatures. Based on the logical growth pattern of literature information, the development stages of medical AI are divided, and the characteristics of research topic evolution are analyzed. Result/Conclusion In terms of research topics, medical AI encompasses 22 research directions, including bioinformatics analysis, surgical robots, medical image analysis, and so on. Regarding the evolutionary characteristics, breakthroughs in deep learning technology, growth in medical demand, and sudden public health events have driven the continuous expansion and deepening of medical AI research. In recent years, the emergence of generative AI has injected new vitality into medical AI research.