Dies ist eine Übersichtsseite mit Metadaten zu dieser wissenschaftlichen Arbeit. Der vollständige Artikel ist beim Verlag verfügbar.
The Use of Chatbots in Head and Neck Mucosal Malignancy Treatment Recommendations
9
Zitationen
7
Autoren
2024
Jahr
Abstract
OBJECTIVE: As cancer patients increasingly use chatbots, it is crucial to recognize ChatGPT's potential in enhancing health literacy while ensuring validation to prevent misinformation. This study aims to assess ChatGPT-3.5's capability to provide appropriate staging and treatment recommendations for head and neck mucosal malignancies for vulnerable populations. STUDY DESIGN AND SETTING: Forty distinct clinical vignettes were introduced into ChatGPT to inquire about staging and treatment recommendations for head and neck mucosal malignancies. METHODS: Prompts were created based on head and neck cancer (HNC) disease descriptions (cancer location, tumor size, lymph node involvement, and symptoms). Staging and treatment recommendations according to the 2021 National Comprehensive Cancer Network (NCCN) guidelines were scored by three fellowship-trained HNC surgeons from two separate tertiary care institutions. HNC surgeons assessed the accuracy of staging and treatment recommendations, such as the completeness of surgery and the appropriateness of treatment modality. RESULTS: Whereas ChatGPT's responses were 95% accurate at recommending the correct first-line treatment based on the 2021 NCCN guidelines, 55% of the responses contained inaccurate staging. Neck dissection was incorrectly omitted from treatment recommendations in 50% of the cases. Moreover, 40% of ChatGPT's treatment recommendations were deemed unnecessary. CONCLUSION: This study emphasizes ChatGPT's potential in HNC patient education, aligning with NCCN guidelines for mucosal malignancies, but highlights the importance of ongoing refinement and scrutiny due to observed inaccuracies in tumor, nodal, metastasis staging, incomplete surgery options, and inappropriate treatment recommendations. Otolaryngologists can use this information to caution patients, families, and trainees regarding the use of ChatGPT for HNC education without expert guidance.
Ähnliche Arbeiten
Explainable Artificial Intelligence (XAI): Concepts, taxonomies, opportunities and challenges toward responsible AI
2019 · 8.693 Zit.
Stop explaining black box machine learning models for high stakes decisions and use interpretable models instead
2019 · 8.598 Zit.
High-performance medicine: the convergence of human and artificial intelligence
2018 · 8.124 Zit.
BioBERT: a pre-trained biomedical language representation model for biomedical text mining
2019 · 6.871 Zit.
Proceedings of the 19th International Joint Conference on Artificial Intelligence
2005 · 5.781 Zit.