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

Related Concept Videos

Positron Emission Tomography

Positron Emission Tomography

Positron emission tomography (PET) is a medical imaging technique involving radiopharmaceuticals — substances that emit short-lived radiation. Although the first PET scanner was introduced in 1961, it took 15 more years before radiopharmaceuticals were combined with the technique and revolutionized its potential.
One of the main requirements of a PET scan is a positron-emitting radioisotope, which is produced in a cyclotron and then attached to a substance used by the part of the body...

You might also read

Related Articles

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

Sort by

Same author

Pre-Imaging Clinical Factors Associated With Cardiac MR Image Quality Using Large Language Model-Enabled Data Extraction.

Journal of magnetic resonance imaging : JMRI·2026

Same author

Large Language Models in Radiologic Numerical Tasks: A Thorough Evaluation and Error Analysis.

Journal of imaging informatics in medicine·2026

Same author

Completing the Baby Album: AI Synthesizing Infant Brain MRI for Missing Time Points.

Radiology. Artificial intelligence·2025

Same author

Balancing Diagnostic Certainty and Locoregional Recurrence Risk in Stage I Non-Small Cell Lung Cancer.

Radiology·2025

Same author

Assessment of delays in diagnosis of lung cancer in interstitial lung disease.

European radiology·2025

Same author

Best Practices for the Safe Use of Large Language Models and Other Generative AI in Radiology.

Radiology·2025

Search research articles

Related Experiment Video

Updated: May 13, 2025

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

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

Published on: December 6, 2024

RadSearch, a Semantic Search Model for Accurate Radiology Report Retrieval with Large Language Model Integration.

Cody H Savage^1,2, Gunvant Chaudhari³, Andrew D Smith^2,4

¹Department of Diagnostic Radiology and Nuclear Medicine, University of Maryland School of Medicine, Baltimore, Md.

|April 15, 2025

Summary

A new semantic search model, RadSearch, significantly improves radiology report retrieval and diagnostic accuracy for large language models. This scalable method enhances clinical information access by outperforming existing models in finding relevant reports.

More Related Videos

A Metadata Extraction Approach for Clinical Case Reports to Enable Advanced Understanding of Biomedical Concepts

A Metadata Extraction Approach for Clinical Case Reports to Enable Advanced Understanding of Biomedical Concepts

Published on: September 20, 2018

Cloud-Based Phrase Mining and Analysis of User-Defined Phrase-Category Association in Biomedical Publications

Cloud-Based Phrase Mining and Analysis of User-Defined Phrase-Category Association in Biomedical Publications

Published on: February 23, 2019

Related Experiment Videos

Last Updated: May 13, 2025

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

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

Published on: December 6, 2024

A Metadata Extraction Approach for Clinical Case Reports to Enable Advanced Understanding of Biomedical Concepts

A Metadata Extraction Approach for Clinical Case Reports to Enable Advanced Understanding of Biomedical Concepts

Published on: September 20, 2018

Cloud-Based Phrase Mining and Analysis of User-Defined Phrase-Category Association in Biomedical Publications

Cloud-Based Phrase Mining and Analysis of User-Defined Phrase-Category Association in Biomedical Publications

Published on: February 23, 2019

Area of Science:

Medical Informatics
Artificial Intelligence in Radiology
Natural Language Processing

Background:

Current radiology report search relies on keywords, lacking semantic understanding and causing false positives.
Developing semantic search models requires scalable methods for generating radiology-specific training data.

Purpose of the Study:

To develop a scalable method for training semantic search models for radiology reports.
To evaluate the performance of a trained model, RadSearch, for clinical applications.

Main Methods:

A scalable method was used to generate training examples from CT and MRI reports (Dec 2021-Jan 2022).
RadSearch was trained and evaluated on internal and external test sets (Dec 2015-Jun 2023).
Performance was assessed for report retrieval, free-text query accuracy, and improving large language model (LLM) diagnostic capabilities.

Main Results:

RadSearch achieved 83.0% accuracy in retrieving reports with specified findings and 89.8% for matching location, significantly outperforming the GTE-large model (P < .001).
Integration with RadSearch improved LLM diagnostic accuracy from 30% to 61% (P < .001), surpassing GTE-large integration (47%, P = .03).
The training set comprised 16,690 reports, with test sets ranging from 6,178 to 13,958 reports.

Conclusions:

A scalable method successfully trained a state-of-the-art semantic search model for radiology reports.
RadSearch demonstrates superior performance in retrieving relevant clinical information and enhancing LLM diagnostic accuracy.