Dies ist eine Übersichtsseite mit Metadaten zu dieser wissenschaftlichen Arbeit. Der vollständige Artikel ist beim Verlag verfügbar.
Is Chat-GPT4 a qualified surgical oncologist?
4
Zitationen
5
Autoren
2023
Jahr
Abstract
Introduced in November 2022, Chat Generative Pre-trained Transformer (ChatGPT) is a large-scale language model designed to answer questions and provide information on a wide range of topics [1] .In the field of medicine, Chat-GPT has demonstrated potential applications: Gilson et al. [2] found that Chat-GPT surpassed a 60% threshold in Step 1 of the National Board of Medical Examiners, equivalent to the passing score of third-year medical students.In another study, Sarraju et al. [3] assessed the appropriateness of Chat-GPT's responses to fundamental cardiovascular disease prevention questions, identifying 21 out of 25 responses as appropriate.In this study, we aimed to explore the potential application of Chat-GPT4 in surgical oncology by taking the three major urological cancers (prostate cancer, bladder cancer, and renal cell carcinoma) as examples, focusing on determining surgical indications, selecting appropriate surgical techniques, and developing perioperative management plans.We posed 40 questions related to prostate cancer, bladder cancer, and renal cell carcinoma, to which Chat-GPT4 (23 March 2023 Version) provided responses (Supplementary Material, Supplemental Digital Content 1, http://links.lww.com/JS9/ A546).Each response was independently evaluated by three urology experts based on guidelines from the European Association of Urology (EAU), American Urological Association (AUA), and National Comprehensive Cancer Network (NCCN), as well as their own clinical practice experience.The responses were scored as 0, 1, 2, or 3, representing complete noncompliance with clinical guidelines and practices or irrelevance; less than 50% compliance with clinical guidelines and practices, with limited guidance value; more than 50% compliance with clinical guidelines and practices, offering some guidance value; and nearly complete compliance with clinical guidelines and practices,
Ähnliche Arbeiten
Explainable Artificial Intelligence (XAI): Concepts, taxonomies, opportunities and challenges toward responsible AI
2019 · 8.418 Zit.
Stop explaining black box machine learning models for high stakes decisions and use interpretable models instead
2019 · 8.288 Zit.
High-performance medicine: the convergence of human and artificial intelligence
2018 · 7.726 Zit.
Proceedings of the 19th International Joint Conference on Artificial Intelligence
2005 · 5.781 Zit.
Peeking Inside the Black-Box: A Survey on Explainable Artificial Intelligence (XAI)
2018 · 5.516 Zit.