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: Oct 5, 2025

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

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

Published on: October 27, 2023

Multimodal deep learning for biomedical data fusion: a review.

Sören Richard Stahlschmidt¹, Benjamin Ulfenborg¹, Jane Synnergren¹

¹Systems Biology Research Center, University of Skövde, Sweden.

Briefings in Bioinformatics

|January 28, 2022

Summary

Deep learning (DL) methods effectively fuse multimodal biomedical data, outperforming traditional approaches. Joint representation learning is key for modeling complex biological interactions, with gradual fusion and transfer learning showing future promise.

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

Targeted Delivery of mRNA to the Heart via Extracellular Vesicles or Lipid Nanoparticles.

Journal of extracellular vesicles·2026

Same author

AI-driven quality control of cell-based ATMPs: automated, accurate, and affordable.

Cytotherapy·2026

Same author

Mast cells in failing human hearts demonstrate transcriptomic activation of pathways involved in cardiac remodeling.

American journal of physiology. Heart and circulatory physiology·2026

Same author

Detecting TP53 mutations in paired liquid and tissue biopsies from patients with high-grade serous ovarian carcinoma.

International journal of cancer·2025

Same author

Novel glycovariant biomarkers of CA125 and CA15-3 and their diagnostic performance across histotypes of ovarian cancer: A multi-cohort study in Sweden and Finland.

European journal of obstetrics, gynecology, and reproductive biology·2025

Same author

Maternal Adiponectin Decreases Placenta Nutrient Transport in Mice.

FASEB journal : official publication of the Federation of American Societies for Experimental Biology·2025

Area of Science:

Biomedical data science
Computational biology
Artificial intelligence in medicine

Background:

Biomedical data is increasingly multimodal, capturing complex biological relationships.
Deep learning (DL) is a popular method for modeling nonlinear relationships in such data.

Purpose of the Study:

To review the state-of-the-art DL-based data fusion strategies for multimodal biomedical data.
To propose a taxonomy to guide choices in fusion strategies and foster research into novel methods.

Main Methods:

Comprehensive review of current deep learning data fusion techniques.
Development of a taxonomy categorizing fusion strategies.
Analysis of the performance of different fusion approaches.

Keywords:

data integration deep neural networks fusion strategies multi-omics multimodal machine learning representation learning

More Related Videos

Author Spotlight: Advancing Alzheimer's Research – Exploring Early Detection and Multi-Omics Approaches

Author Spotlight: Advancing Alzheimer's Research – Exploring Early Detection and Multi-Omics Approaches

Published on: December 15, 2023

Related Experiment Videos

Last Updated: Oct 5, 2025

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

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

Published on: October 27, 2023

Author Spotlight: Advancing Alzheimer's Research – Exploring Early Detection and Multi-Omics Approaches

Author Spotlight: Advancing Alzheimer's Research – Exploring Early Detection and Multi-Omics Approaches

Published on: December 15, 2023

Main Results:

Deep fusion strategies generally outperform unimodal and shallow methods.
Specific subcategories of fusion strategies exhibit distinct advantages and disadvantages.
Joint representation learning is a preferred intermediate fusion strategy for modeling biological interactions.

Conclusions:

Deep learning-based multimodal data fusion offers a powerful approach for understanding health and disease dynamics.
Gradual fusion and transfer learning represent promising avenues for future research.
Holistic models trained with multimodal DL can capture complex regulatory dynamics.