Improving the use of LLMs in radiology through prompt engineering: from precision prompts to zero-shot learning
RöFo - Fortschritte auf dem Gebiet der Röntgenstrahlen und der bildgebenden Verfahren · 2024-11 · 196 (11) · 1166-1170
DOI: 10.1055/a-2264-5631 ↗Kurz gefasst
Fachartikel in der Zeitschrift RöFo - Fortschritte auf dem Gebiet der Röntgenstrahlen und der bildgebenden Verfahren, online erschienen am 26.02.2024. Verfasst von Maximilian Frederik Russe und 3 weiteren Autor:innen. Beteiligte Einrichtung: Department of Radiology, University Hospital Freiburg, Freiburg, Germany. Hauptthemen laut MEDLINE: Radiology. Die Arbeit wurde bisher 52-mal zitiert (Crossref) und nennt 18 Literaturangaben.
Abstract (Originalsprache)
Purpose
Large language models (LLMs) such as ChatGPT have shown significant potential in radiology. Their effectiveness often depends on prompt engineering, which optimizes the interaction with the chatbot for accurate results. Here, we highlight the critical role of prompt engineering in tailoring the LLMs’ responses to specific medical tasks. Materials and Methods Using a clinical case, we elucidate different prompting strategies to adapt the LLM ChatGPT using GPT4 to new tasks without additional training of the base model. These approaches range from precision prompts to advanced in-context methods such as few-shot and zero-shot learning. Additionally, the significance of embeddings, which serve as a data representation technique, is discussed.
Results
Prompt engineering substantially improved and focused the chatbot’s output. Moreover, embedding of specialized knowledge allows for more transparent insight into the model’s decision-making and thus enhances trust.
Conclusion
Despite certain challenges, prompt engineering plays a pivotal role in harnessing the potential of LLMs for specialized tasks in the medical domain, particularly radiology. As LLMs continue to evolve, techniques like few-shot learning, zero-shot learning, and embedding-based retrieval mechanisms will become indispensable in delivering tailored outputs. Key Points Citation Format
Schlagwörter
MeSH-Schlagwörter (MEDLINE)
Autor:innen und Institutionen
- Department of Radiology, University Hospital Freiburg, Freiburg, Germany.
ORCID-Profile: Alexander Rau
Zitieren
APA
Russe, M. F., Reisert, M., Bamberg, F., & Rau, A. (2024). Improving the use of LLMs in radiology through prompt engineering: from precision prompts to zero-shot learning. RöFo - Fortschritte auf dem Gebiet der Röntgenstrahlen und der bildgebenden Verfahren, 196(11), 1166-1170. https://doi.org/10.1055/a-2264-5631
Vancouver
Russe MF, Reisert M, Bamberg F, Rau A. Improving the use of LLMs in radiology through prompt engineering: from precision prompts to zero-shot learning. RöFo - Fortschritte auf dem Gebiet der Röntgenstrahlen und der bildgebenden Verfahren. 2024;196(11):1166-1170. doi:10.1055/a-2264-5631
BibTeX
@article{Russe2024,
author = {Russe, Maximilian Frederik and Reisert, Marco and Bamberg, Fabian and Rau, Alexander},
title = {Improving the use of LLMs in radiology through prompt engineering: from precision prompts to zero-shot learning},
journal = {RöFo - Fortschritte auf dem Gebiet der Röntgenstrahlen und der bildgebenden Verfahren},
year = {2024},
volume = {196},
number = {11},
pages = {1166--1170},
doi = {10.1055/a-2264-5631}
}
EUVIMED