Department of Urology, University College of London Hospitals (UCLH), W1G8PH London, UK.
2nd Department of Urology, Sismanoglio Hospital, National and Kapodistrian University of Athens, 10431 Athens, Greece.
Arch Esp Urol. 2024 Aug;77(7):708-717. doi: 10.56434/j.arch.esp.urol.20247707.99.
This study aims to provide a comprehensive overview of the current literature on the utilisation of ChatGPT in the fields of clinical medicine, urology, and academic medicine, while also addressing the associated ethical challenges and potential risks.
This narrative review conducted an extensive search of the PubMed and MEDLINE databases, covering the period from January 2022 to January 2024. The search phrases employed were "urologic surgery" in conjunction with "artificial intelligence", "machine learning", "neural network", "ChatGPT", "urology", and "medicine". The initial studies were chosen from the screened research to examine the possible interaction between those entities. Research utilising animal models was excluded.
ChatGPT has demonstrated its usefulness in clinical settings by producing precise clinical correspondence, discharge summaries, and medical records, thereby assisting in these laborious tasks, especially with the latest iterations of ChatGPT. Furthermore, patients can access essential medical information by inquiring with ChatGPT. Nevertheless, there are multiple concerns regarding the correctness of the system, including allegations of falsified data and references. These issues emphasise the importance of having a doctor oversee the final result to guarantee patient safety. ChatGPT shows potential in academic medicine for generating drafts and organising datasets. However, the presence of guidelines and plagiarism-detection technologies is necessary to mitigate the risks of plagiarism and the use of faked data when using it for academic purposes.
ChatGPT should be utilised as a supplementary tool by urologists and academicians. However, it is now advisable to have human oversight to guarantee patient safety, uphold academic integrity, and maintain transparency.
本研究旨在全面概述 ChatGPT 在临床医学、泌尿科和学术医学领域的应用文献，同时探讨相关的伦理挑战和潜在风险。
本叙事性综述对 PubMed 和 MEDLINE 数据库进行了广泛搜索，涵盖了 2022 年 1 月至 2024 年 1 月期间的文献。使用的搜索词是“泌尿科手术”与“人工智能”、“机器学习”、“神经网络”、“ChatGPT”、“泌尿科”和“医学”相结合。从筛选出的研究中选择初始研究，以检查这些实体之间的可能相互作用。排除了使用动物模型的研究。
ChatGPT 通过生成精确的临床信函、出院小结和病历，在临床环境中展现了其有用性，从而辅助完成这些繁琐的任务，尤其是使用最新迭代的 ChatGPT 时。此外，患者可以通过向 ChatGPT 查询来获取重要的医疗信息。然而，系统的正确性存在多个问题，包括伪造数据和参考文献的指控。这些问题强调了在最终结果中由医生监督的重要性，以保证患者安全。ChatGPT 在学术医学领域具有生成草稿和组织数据集的潜力。然而，在学术用途中使用 ChatGPT 时，需要有指南和剽窃检测技术来降低剽窃和使用伪造数据的风险。
泌尿科医生和学者可以将 ChatGPT 用作辅助工具。然而，现在建议进行人工监督，以保证患者安全、维护学术诚信和保持透明度。
