Search research articles

ABOUT JoVE

Overview Leadership Blog JoVE Help Center

AUTHORS

Publishing Process Editorial Board Scope & Policies Peer Review FAQ Submit

LIBRARIANS

Testimonials Subscriptions Access Resources Library Advisory Board FAQ

RESEARCH

JoVE Journal Methods Collections JoVE Encyclopedia of Experiments Archive

EDUCATION

JoVE Core JoVE Business JoVE Science Education JoVE Lab Manual Faculty Resource Center Faculty Site

Terms & Conditions of Use

Search research articles

Related Experiment Video

Updated: May 7, 2026

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

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

Published on: August 16, 2020

Leveraging Fine-Tuned Large Language Models for Interpretable Pancreatic Cystic Lesion Feature Extraction and Risk

Ebrahim Rasromani¹, Stella K Kang², Yanqi Xu¹

¹Center for Data Science, New York University, 60 5th Ave, New York, NY 10011. United States.

AJR. American Journal of Roentgenology

|May 6, 2026

Summary

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

Neurogenic diarrhoea after superior mesenteric artery divestment or resection for pancreatic adenocarcinoma following preoperative chemotherapy: international multicentre retrospective study.

BJS open·2026

Same author

Impact of preoperative radiotherapy on the risk of postpancreatectomy haemorrhage and major morbidity after resection of pancreatic adenocarcinoma including arterial divestment or arterial resection.

BJS open·2026

Same author

ASO Author Reflections: Mortality During the Real-World Adoption of Robotic Pancreaticoduodenectomy in the USA.

Annals of surgical oncology·2026

Same author

ASO Visual Abstract: Increased Mortality with Surgeon Adoption of Robotic Pancreaticoduodenectomy-A National EHR Study of Outcomes.

Annals of surgical oncology·2026

Same author

Postpancreatectomy liver injury: A relevant entity in the modern era of pancreatic cancer surgery with hepatic vessel resection. A monocentric retrospective cohort study.

Surgery·2026

Same author

Increased Mortality with Surgeon Adoption of Robotic Pancreaticoduodenectomy: A National EHR Study of Outcomes.

Annals of surgical oncology·2026

Same journal

The Banality of Cancer: Entropy As a Third Pillar of Lung Nodule Risk Assessment.

AJR. American journal of roentgenology·2026

Same journal

A Narrow Window for Artificial Intelligence-Generated Synthetic Temporal Bone CT From MRI.

AJR. American journal of roentgenology·2026

Same journal

From Uncertainty to Actionable Management: The Isolated Abnormal Axillary Lymph Node.

AJR. American journal of roentgenology·2026

Same journal

Beyond Detection: Translating Artificial Intelligence-Driven Opportunistic Screening Into Clinical Action.

AJR. American journal of roentgenology·2026

Same journal

Navigating PSMA PET Radiopharmaceuticals: Clinical and Operational Factors.

AJR. American journal of roentgenology·2026

Same journal

From Mesenteric Ischemia to Intestinal Stroke.

AJR. American journal of roentgenology·2026

See all related articles

This summary is machine-generated.

Large language models (LLMs) can automatically extract pancreatic cystic lesion (PCL) features from radiology reports, comparable to human experts. Open-source LLMs fine-tuned with chain-of-thought (CoT) reasoning show high accuracy and potential for large-scale PCL research.

Area of Science:

Artificial Intelligence in Radiology
Medical Informatics
Natural Language Processing

Background:

Manual extraction of pancreatic cystic lesion (PCL) features from radiology reports is time-consuming, hindering large-scale research.
Automating this process is crucial for advancing PCL studies.

Purpose of the Study:

To evaluate the performance of GPT-4o, Llama, and DeepSeek large language models (LLMs) for extracting PCL features.
To assess the impact of chain-of-thought (CoT) reasoning on LLM performance for PCL feature extraction.

Main Methods:

A dataset of 6469 abdominal MRI/CT reports was curated.
Open-source LLMs (Llama, DeepSeek) were fine-tuned using CoT labels generated by GPT-4o.
Performance was evaluated using exact match accuracy, risk categorization F1 scores, and radiologist agreement (Fleiss' kappa).

Related Experiment Videos

Last Updated: May 7, 2026

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

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

Published on: August 16, 2020

Main Results:

Fine-tuned LLMs achieved high feature extraction accuracy (97-98%) and risk categorization F1 scores (0.93-0.97).
Open-source LLMs performed comparably to GPT-4o, with no significant difference in radiologist agreement.
Object identification and clinical reasoning were identified as primary error sources.

Conclusions:

LLMs demonstrate feasibility for automated PCL feature extraction from radiology reports.
Fine-tuned open-source LLMs offer a viable alternative to closed-source models.
CoT reasoning enhances accuracy and facilitates interpretable error analysis, supporting population-level PCL research.