ChatGPT在皮肤病诊断中的作用。

The Role of ChatGPT in Dermatology Diagnostics.

作者信息

Khamaysi Ziad, Awwad Mahdi, Jiryis Badea, Bathish Naji, Shapiro Jonathan

机构信息

Department of Dermatology, Rambam Health Care Campus, Haifa 3109601, Israel.

Bruce Rappaport Faculty of Medicine, Technion Institute of Technology, Haifa 3525433, Israel.

出版信息

Diagnostics (Basel). 2025 Jun 16;15(12):1529. doi: 10.3390/diagnostics15121529.

DOI:10.3390/diagnostics15121529

PMID:40564849

原文链接:https://pmc.ncbi.nlm.nih.gov/articles/PMC12191462/

Abstract

Artificial intelligence (AI), especially large language models (LLMs) like ChatGPT, has disrupted different medical disciplines, including dermatology. This review explores the application of ChatGPT in dermatological diagnosis, emphasizing its role in natural language processing (NLP) for clinical data interpretation, differential diagnosis assistance, and patient communication enhancement. ChatGPT can enhance a diagnostic workflow when paired with image analysis tools, such as convolutional neural networks (CNNs), by merging text and image data. While it boasts great capabilities, it still faces some issues, such as its inability to perform any direct image analyses and the risk of inaccurate suggestions. Ethical considerations, including patient data privacy and the responsibilities of the clinician, are discussed. Future perspectives include an integrated multimodal model and AI-assisted framework for diagnosis, which shall improve dermatology practice.

摘要

人工智能（AI），尤其是像ChatGPT这样的大型语言模型（LLM），已经对包括皮肤科在内的不同医学学科产生了冲击。本综述探讨了ChatGPT在皮肤病诊断中的应用，强调了其在自然语言处理（NLP）中对临床数据解读、鉴别诊断辅助和改善医患沟通方面的作用。ChatGPT与卷积神经网络（CNN）等图像分析工具相结合时，通过融合文本和图像数据，可以优化诊断流程。尽管它具有强大的功能，但仍然面临一些问题，例如无法进行任何直接的图像分析以及存在给出不准确建议的风险。文中还讨论了包括患者数据隐私和临床医生责任在内的伦理考量。未来展望包括用于诊断的集成多模态模型和人工智能辅助框架，这将改善皮肤科的诊疗实践。