Zhang Jiaxian, Sun Ying, Rong Yihong, Li Haoyang, Jiang Botian, Zhao Chen, Liu Hongwei
Key Laboratory of Regenerative Medicine of Ministry of Education, Department of Plastic Surgery of the First Affiliated Hospital of Jinan University, Institute of New Technology of Plastic Surgery of Jinan University, Guangzhou, 510630, China.
Advanced Medical and Dental Institute, Universiti Sains Malaysia, 350180, Penang, Malaysia.
Aesthetic Plast Surg. 2025 Aug 8. doi: 10.1007/s00266-025-05103-4.
The use of artificial intelligence (AI) chatbots has demonstrated considerable promise in assisting medical consultations. However, their potential for application in online hair transplantation consultations remains largely unexplored.
This study aims to assess the effectiveness of AI chatbots in responding to patient inquiries during online hair transplantation consultations.
We evaluated responses to 10 common patient questions collected from online hair transplantation clinics, comparing answers generated by three AI chatbots-ChatGPT-4o mini, Claude 3.5 Sonnet, and Gemini Advanced-with those from senior surgeons. Each response was scored based on medical accuracy, empathy, understandability, actionability, and readability, with a focus on determining how well AI can match or exceed human expert performance.
All three AI chatbots matched or outperformed the response capabilities of senior surgeons in medical accuracy, empathy, understandability, actionability, and readability. Among them, Gemini Advanced showed the most comprehensive advantages, including significantly higher scores in medical accuracy (4.5 vs. 3.9, P < .001), empathy (4.9 vs. 2.5, P < .001), and understandability (82.7% vs. 63.9%, P < .001). Additionally, Gemini Advanced demonstrated a lower Flesch-Kincaid Grade Level (10.5 vs. 18.7, P < .001) and higher Flesch Reading Ease Score (40.3 vs. 16.2, P < .001), suggesting better readability.
AI chatbots show strong potential for use in online hair transplantation consultations, providing accurate, empathetic, and easily understandable responses. Nevertheless, challenges such as privacy concerns, ethical considerations, and potential biases need to be addressed before their adoption in clinical practice.
This journal requires that authors assign a level of evidence to each article. For a full description of these Evidence-Based Medicine ratings, please refer to the Table of Contents or the online Instructions to Authors www.springer.com/00266 .
人工智能（AI）聊天机器人在协助医疗咨询方面已展现出可观的前景。然而，其在在线植发咨询中的应用潜力在很大程度上仍未得到探索。
本研究旨在评估人工智能聊天机器人在在线植发咨询中回应患者询问的有效性。
我们评估了对从在线植发诊所收集的10个常见患者问题的回答，将三个人工智能聊天机器人（ChatGPT-4o mini、Claude 3.5 Sonnet和Gemini Advanced）生成的答案与资深外科医生的答案进行比较。每个回答根据医学准确性、同理心、易懂性、可操作性和可读性进行评分，重点是确定人工智能在多大程度上能够匹配或超越人类专家的表现。
在医学准确性、同理心、易懂性、可操作性和可读性方面，所有三个人工智能聊天机器人的回答能力均与资深外科医生相当或更胜一筹。其中，Gemini Advanced表现出最全面的优势，包括在医学准确性（4.5对3.9，P <.001）、同理心（4.9对2.5，P <.001）和易懂性（82.7%对63.9%，P <.001）方面得分显著更高。此外，Gemini Advanced的弗莱什-金凯德年级水平较低（10.5对18.7，P <.001），弗莱什阅读易读性得分较高（40.3对16.2，P <.001），表明可读性更好。
人工智能聊天机器人在在线植发咨询中显示出强大的应用潜力，能够提供准确、有同理心且易于理解的回答。然而，在临床实践中采用之前，隐私问题、伦理考量和潜在偏差等挑战需要得到解决。
证据级别IV：本期刊要求作者为每篇文章指定证据级别。有关这些循证医学评级的完整描述，请参阅目录或在线作者指南www.springer.com/00266 。
