Dies ist eine Übersichtsseite mit Metadaten zu dieser wissenschaftlichen Arbeit. Der vollständige Artikel ist beim Verlag verfügbar.
Gastroenterology-Specific AI Model GastroGPT Outperforms Attending Physicians’ and ChatGPT in Analyses of Clinical Notes of Real-World Endoscopy Cases
0
Zitationen
7
Autoren
2024
Jahr
Abstract
Aims Endoscopic procedures involve non-endoscopist collaboration in healthcare presentations. Despite AI potential, like GastroGPT, its use is limited, emphasizing the need for greater clinical integration. Our team developed GastroGPT, a Gastroenterology specific LLM for clinical based text generation. Previous studies showed this approach outperforms general LLMs and attending physicians in general gastroenterology tasks and simulated cases. However, its success in an endoscopy field and comparison to physician note texts is not known. As such, the study aims to assess GastroGPT's abilities in comparison with attending physicians and ChatGPT-4 serving as a reference model, across seven different components of patient care sequence using authentic patient data from real-life endoscopy cases.
Ähnliche Arbeiten
Explainable Artificial Intelligence (XAI): Concepts, taxonomies, opportunities and challenges toward responsible AI
2019 · 8.239 Zit.
Stop explaining black box machine learning models for high stakes decisions and use interpretable models instead
2019 · 8.095 Zit.
High-performance medicine: the convergence of human and artificial intelligence
2018 · 7.463 Zit.
Proceedings of the 19th International Joint Conference on Artificial Intelligence
2005 · 5.776 Zit.
Peeking Inside the Black-Box: A Survey on Explainable Artificial Intelligence (XAI)
2018 · 5.428 Zit.