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

Challenges in AI Based Tumor Board Case Summarization and Recommendations.

Research square·2026
Same author

Automated Identification of Surgical Site Infections From Electronic Medical Records: Retrospective Observational Predictive Modeling Study.

JMIR perioperative medicine·2026
Same author

Automated identification of incidentalomas requiring follow-up: A multi-anatomy evaluation of LLM-based and supervised approaches.

Journal of biomedical informatics·2026
Same author

Using large language models to identify prediagnostic clinical features of ovarian cancer from healthcare records: a population-based case-control study.

The British journal of general practice : the journal of the Royal College of General Practitioners·2026
Same author

Predicting Early-Onset Colorectal Cancer with Large Language Models.

AMIA ... Annual Symposium proceedings. AMIA Symposium·2026
Same author

WoundcareVQA: A multilingual visual question answering benchmark dataset for wound care.

Journal of biomedical informatics·2025

Related Experiment Video

Updated: Sep 29, 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

703

Automatic Assignment of Radiology Examination Protocols Using Pre-trained Language Models with Knowledge

Wilson Lau1, Laura Aaltonen2, Martin Gunn2

  • 1Department of Biomedical and Health Informatics.

AMIA ... Annual Symposium Proceedings. AMIA Symposium
|March 21, 2022
PubMed
Summary

This study introduces BERTrad, a deep learning model for automatically assigning computed tomography (CT) examination protocols. Knowledge distillation enhanced performance on rare protocols, improving overall accuracy in radiology workflow.

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

7.0K
Author Spotlight: An Efficient and Robust Software for Automated Fusion of Multiple Preclinical Imaging Modalities
07:13

Author Spotlight: An Efficient and Robust Software for Automated Fusion of Multiple Preclinical Imaging Modalities

Published on: October 27, 2023

1.4K

Related Experiment Videos

Last Updated: Sep 29, 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

703
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

7.0K
Author Spotlight: An Efficient and Robust Software for Automated Fusion of Multiple Preclinical Imaging Modalities
07:13

Author Spotlight: An Efficient and Robust Software for Automated Fusion of Multiple Preclinical Imaging Modalities

Published on: October 27, 2023

1.4K

Area of Science:

  • Artificial Intelligence in Radiology
  • Machine Learning for Medical Imaging
  • Natural Language Processing in Healthcare

Background:

  • Manual selection of radiology examination protocols is a significant bottleneck in clinical workflows.
  • Computed tomography (CT) examinations require specific protocols for optimal image acquisition and diagnostic accuracy.
  • Existing methods for protocol assignment are often time-consuming and prone to errors.

Purpose of the Study:

  • To develop and evaluate a deep learning approach for automated protocol assignment in CT examinations.
  • To address the challenge of data imbalance across different examination protocols.
  • To compare the performance of the proposed model against traditional machine learning and baseline deep learning models.

Main Methods:

  • Pre-training of a domain-specific BERT model (BERTrad) on radiology examination data.
  • Implementation of a knowledge distillation technique combined with data augmentation to handle class imbalance.
  • Comparison of BERTrad with n-gram models (SVM, GBM, RF) and the BERTbase model using macro-averaged F1 scores.

Main Results:

  • BERTrad achieved a higher classification performance (F1 score of 0.63) compared to BERTbase (0.61) and traditional models (max F1 score of 0.60).
  • Knowledge distillation significantly improved performance on minority classes, resulting in an overall F1 score of 0.66.
  • The proposed approach demonstrates superior ability in accurately assigning diverse radiology examination protocols.

Conclusions:

  • Automated protocol assignment using the BERTrad model offers a promising solution to streamline radiology workflows.
  • The integration of knowledge distillation effectively mitigates challenges posed by data imbalance in medical imaging datasets.
  • This deep learning approach has the potential to reduce manual effort and improve efficiency in CT protocol selection.