Zero-shot Thoracic Oncologic History Generation for Radiologists Using Retrieval-augmented Large Language Model Pipeline
Radiology · 01.06.2026 · 319 (3)
DOI: 10.1148/radiol.251581 ↗Kurz gefasst
Fachartikel in der Zeitschrift Radiology, erschienen am 01.06.2026. Verfasst von Karan Jani und 6 weiteren Autor:innen. Beteiligte Einrichtung: Mallinckrodt Institute of Radiology, Washington University School of Medicine, 510 S Kingshighway Blvd, St Louis, MO 63110. Hauptthemen laut MEDLINE: Large Language Models, Thoracic Neoplasms, Electronic Health Records, Information Storage and Retrieval. Die Arbeit nennt 25 Literaturangaben.
Abstract (Originalsprache)
A large language model pipeline automatically summarized oncologic histories with high efficiency, accuracy, and completeness; speed and performance varied depending on the model used (GPT-4o mini, o3-mini, GPT-5-mini [OpenAI]).
Schlagwörter
MeSH-Schlagwörter (MEDLINE)
Autor:innen und Institutionen
- Mallinckrodt Institute of Radiology, Washington University School of Medicine, 510 S Kingshighway Blvd, St Louis, MO 63110.
ORCID-Profile: Karan Jani, Govind Mattay, Vamsi Narra, Anup Shetty, Glenwood Stancil, Andrew Bierhals, Yasasvi Tadavarthi
Zitieren
APA
Jani, K., Mattay, G., Narra, V., Shetty, A., Stancil, G., Bierhals, A., & Tadavarthi, Y. (2026). Zero-shot Thoracic Oncologic History Generation for Radiologists Using Retrieval-augmented Large Language Model Pipeline. Radiology, 319(3). https://doi.org/10.1148/radiol.251581
Vancouver
Jani K, Mattay G, Narra V, Shetty A, Stancil G, Bierhals A, et al. Zero-shot Thoracic Oncologic History Generation for Radiologists Using Retrieval-augmented Large Language Model Pipeline. Radiology. 2026;319(3). doi:10.1148/radiol.251581
BibTeX
@article{Jani2026,
author = {Jani, Karan and Mattay, Govind and Narra, Vamsi and Shetty, Anup and Stancil, Glenwood and Bierhals, Andrew and Tadavarthi, Yasasvi},
title = {Zero-shot Thoracic Oncologic History Generation for Radiologists Using Retrieval-augmented Large Language Model Pipeline},
journal = {Radiology},
year = {2026},
volume = {319},
number = {3},
doi = {10.1148/radiol.251581}
}
EUVIMED