Latest research indicates that GPT-4V performs exceptionally well in medical licensing exams, offering new possibilities for AI applications in future medical practice. Despite significant progress in medical imaging diagnostics, there remains a trust issue among doctors regarding AI diagnostic results. The study suggests evaluating AI's professional knowledge and skills through medical licensing exams, emphasizing the importance of human-machine collaboration in future medical diagnostics. However, GPT-4V still needs to improve the quality and reliability of its explanations to adapt to real clinical scenarios. This research provides a direction for future development of more sophisticated AI collaboration systems.