Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by
Same author

Investigating Age-Related Differences in Survival Outcomes and Mutation Status of Patients with Intrahepatic Cholangiocarcinoma.

Annals of surgical oncology·2026
Same author

Pancreatic Cancer: Translating Tumor Biology into Actionability.

Cancer discovery·2026
Same author

Long-term Outcomes After Resection of Solitary Colorectal Liver Metastases.

Annals of surgical oncology·2026
Same author

ASO Visual Abstract: Long-Term Outcomes After Resection of Solitary Colorectal Liver Metastases.

Annals of surgical oncology·2026
Same author

Pancreas Cancer Risk During Intraductal Papillary Mucinous Neoplasm Surveillance - A SEER-Based Comparison.

Annals of surgery·2026
Same author

Social and Surgical Drivers of Survival in Colorectal Liver Metastases: An NCBD Analysis.

Journal of surgical oncology·2026

Related Experiment Video

Updated: Sep 16, 2025

Augmenting Large Language Models via Vector Embeddings to Improve Domain-Specific Responsiveness
03:14

Augmenting Large Language Models via Vector Embeddings to Improve Domain-Specific Responsiveness

Published on: December 6, 2024

692

Data Extraction and Curation from Radiology Reports for Pancreatic Cyst Surveillance Using Large Language Models.

Ankur P Choubey1, Emanuel Eguia1, Alexander Hollingsworth1,2

  • 1From the Department of Surgery, Memorial Sloan Kettering Cancer Center, New York, NY (Choubey, Eguia, Hollingsworth, D'Angelica, Jarnagin, Wei, Soares).

Journal of the American College of Surgeons
|July 10, 2025
PubMed
Summary
This summary is machine-generated.

Large language models (LLMs) can accurately extract radiographic features from radiology reports for pancreatic cyst surveillance. This technology enables efficient data curation for longitudinal studies and potential AI-driven surveillance models.

More Related Videos

Machine Learning Algorithms for Early Detection of Bone Metastases in an Experimental Rat Model
07:15

Machine Learning Algorithms for Early Detection of Bone Metastases in an Experimental Rat Model

Published on: August 16, 2020

6.9K
Author Spotlight: A 3D Digital Model for the Diagnosis and Treatment of Pulmonary Nodules
10:26

Author Spotlight: A 3D Digital Model for the Diagnosis and Treatment of Pulmonary Nodules

Published on: May 19, 2023

2.1K

Related Experiment Videos

Last Updated: Sep 16, 2025

Augmenting Large Language Models via Vector Embeddings to Improve Domain-Specific Responsiveness
03:14

Augmenting Large Language Models via Vector Embeddings to Improve Domain-Specific Responsiveness

Published on: December 6, 2024

692
Machine Learning Algorithms for Early Detection of Bone Metastases in an Experimental Rat Model
07:15

Machine Learning Algorithms for Early Detection of Bone Metastases in an Experimental Rat Model

Published on: August 16, 2020

6.9K
Author Spotlight: A 3D Digital Model for the Diagnosis and Treatment of Pulmonary Nodules
10:26

Author Spotlight: A 3D Digital Model for the Diagnosis and Treatment of Pulmonary Nodules

Published on: May 19, 2023

2.1K

Area of Science:

  • Radiology
  • Artificial Intelligence
  • Medical Informatics

Background:

  • Manual data extraction from pancreatic cyst registries is labor-intensive and hinders widespread implementation.
  • Automating the extraction of radiographic features is crucial for efficient longitudinal evaluation of pancreatic cysts.

Purpose of the Study:

  • To assess the feasibility and accuracy of using large language models (LLMs) for extracting clinical variables from radiology reports.
  • To evaluate LLMs' performance in curating data for pancreatic cyst surveillance.

Main Methods:

  • A retrospective study utilized LLMs (GPT-4) with a zero-shot learning approach to extract nine radiographic elements from 3198 scans of 991 patients.
  • Data extracted by LLMs were compared against a manually annotated institutional cyst database serving as ground truth.

Main Results:

  • LLMs demonstrated high accuracy in extracting categorical variables (97%-99%) and continuous variables (92%-97%) related to pancreatic cyst progression.
  • Accuracy for cyst size was 92% (Cohen's Kappa 0.92) and for main pancreatic duct size was 97% (Cohen's Kappa 0.82).
  • The lowest accuracy (81%) was observed for the multi-class variable concerning the number of cysts.

Conclusions:

  • LLMs can accurately and reliably extract data from radiology reports for pancreatic cyst surveillance.
  • This technology facilitates the assembly of longitudinal databases and may enable the development of AI-based surveillance models.