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

Combination Therapies and Personalized Medicine02:50

Combination Therapies and Personalized Medicine

4.9K
Combining two or more treatment methods increases the life span of cancer patients while reducing damage to vital organs or tissue from the overuse of a single treatment. Combination therapy also targets different cancer-inducing pathways, thus reducing the chances of developing resistance to treatment.
The combination of the drug acetazolamide and sulforaphane is a good example of combination therapy to treat cancer. The cells in the interior of a large tumor often die due to the hypoxic and...
4.9K
Ultrasonography01:17

Ultrasonography

4.5K
Ultrasonography is an imaging technique that uses high-frequency sound waves to visualize the body's internal structures. It is a non-invasive and safe procedure that does not involve the use of ionizing radiation, making it widely used in various medical fields. Ultrasonography is used to study heart function, blood flow in the neck or extremities, certain conditions such as gallbladder disease, and fetal growth and development.
During an ultrasonography procedure, a handheld device called...
4.5K

You might also read

Related Articles

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

Sort by
Same author

Clinically grounded retinal representation learning from minimal supervision.

Scientific reports·2026
Same author

Early fibrotic niches establish tumour-permissive microenvironments.

Nature·2026
Same author

scMILD: Single-cell multiple instance learning for sample classification and associated subpopulation discovery.

iScience·2026
Same author

Data augmentation method for computer-aided diagnosis using specular reflection.

Biomedical engineering letters·2026
Same author

scTELL: a single-cell ATAC-seq tool for locus-specific transposable element identification in chromatin accessibility.

Mobile DNA·2026
Same author

Artificial intelligence based prediction of first recurrence in neovascular age related macular degeneration with validation by 19 experts.

Scientific reports·2026

Related Experiment Video

Updated: Jul 17, 2025

Evidence-based Knowledge Synthesis and Hypothesis Validation: Navigating Biomedical Knowledge Bases via Explainable AI and Agentic Systems
05:47

Evidence-based Knowledge Synthesis and Hypothesis Validation: Navigating Biomedical Knowledge Bases via Explainable AI and Agentic Systems

Published on: June 13, 2025

276

Selective UMLS knowledge infusion for biomedical question answering.

Hyeryun Park1,2, Jiye Son1,2, Jeongwon Min1,2

  • 1Interdisciplinary Program for Bioengineering, Seoul National University Graduate School, Seoul, Republic of Korea.

Scientific Reports
|August 30, 2023
PubMed
Summary

This study introduces an efficient method for integrating biomedical knowledge into AI language models using adapters. Pretraining adapters on specific knowledge graph partitions improves biomedical question-answering performance, reducing computational costs.

More Related Videos

A Knowledge Graph Approach to Elucidate the Role of Organellar Pathways in Disease via Biomedical Reports
07:35

A Knowledge Graph Approach to Elucidate the Role of Organellar Pathways in Disease via Biomedical Reports

Published on: October 13, 2023

1.7K
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

618

Related Experiment Videos

Last Updated: Jul 17, 2025

Evidence-based Knowledge Synthesis and Hypothesis Validation: Navigating Biomedical Knowledge Bases via Explainable AI and Agentic Systems
05:47

Evidence-based Knowledge Synthesis and Hypothesis Validation: Navigating Biomedical Knowledge Bases via Explainable AI and Agentic Systems

Published on: June 13, 2025

276
A Knowledge Graph Approach to Elucidate the Role of Organellar Pathways in Disease via Biomedical Reports
07:35

A Knowledge Graph Approach to Elucidate the Role of Organellar Pathways in Disease via Biomedical Reports

Published on: October 13, 2023

1.7K
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

618

Area of Science:

  • Biomedical Informatics
  • Artificial Intelligence
  • Natural Language Processing

Background:

  • Knowledge-intensive question-answering is a key AI application in biomedicine.
  • Domain expertise is critical, necessitating efficient knowledge infusion into language models.
  • Current methods of transferring large knowledge graphs are computationally expensive.

Purpose of the Study:

  • To propose an efficient method for infusing biomedical knowledge into pretrained language models for question-answering.
  • To investigate the necessity of using all knowledge graph semantics.
  • To explore strategies for partitioning knowledge graphs for efficient pretraining.

Main Methods:

  • Leveraging adapters to inject Unified Medical Language System (UMLS) knowledge into pretrained language models.
  • Investigating strategies for partitioning knowledge graphs, including discarding or merging semantic groups.
  • Evaluating performance on three biomedical question-answering finetuning datasets.

Main Results:

  • Adapters pretrained on semantically partitioned knowledge graph groups demonstrated improved efficiency in evaluation metrics, parameter count, and time.
  • Discarding knowledge groups with fewer concepts is more effective for smaller datasets.
  • Merging these groups is more beneficial for larger datasets.
  • Adapter methodology showed insensitivity to specific group formulations, with slight metric improvements.

Conclusions:

  • Adapter-based knowledge infusion offers an efficient approach to enhance biomedical question-answering models.
  • Strategic partitioning and selection of knowledge graph components optimize pretraining efficiency.
  • The method is robust across different dataset sizes and knowledge graph structures.