Högskolan i Skövde

his.sePublications
Change search
CiteExportLink to record
Permanent link

Direct link
Cite
Citation style
  • apa
  • apa-cv
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • Other style
More styles
Language
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Other locale
More languages
Output format
  • html
  • text
  • asciidoc
  • rtf
Multimodal deep learning for biomedical data fusion: a review
University of Skövde, School of Bioscience. University of Skövde, Systems Biology Research Environment. (Translational bioinformatics)ORCID iD: 0000-0003-4191-8435
University of Skövde, Systems Biology Research Environment. University of Skövde, School of Bioscience. (Translational bioinformatics)ORCID iD: 0000-0001-9242-4852
University of Skövde, Systems Biology Research Environment. University of Skövde, School of Bioscience. (Translational bioinformatics)ORCID iD: 0000-0003-4697-0590
2022 (English)In: Briefings in Bioinformatics, ISSN 1467-5463, E-ISSN 1477-4054, Vol. 23, no 2, article id bbab569Article, review/survey (Refereed) Published
Abstract [en]

Biomedical data are becoming increasingly multimodal and thereby capture the underlying complex relationships among biological processes. Deep learning (DL)-based data fusion strategies are a popular approach for modeling these nonlinear relationships. Therefore, we review the current state-of-the-art of such methods and propose a detailed taxonomy that facilitates more informed choices of fusion strategies for biomedical applications, as well as research on novel methods. By doing so, we find that deep fusion strategies often outperform unimodal and shallow approaches. Additionally, the proposed subcategories of fusion strategies show different advantages and drawbacks. The review of current methods has shown that, especially for intermediate fusion strategies, joint representation learning is the preferred approach as it effectively models the complex interactions of different levels of biological organization. Finally, we note that gradual fusion, based on prior biological knowledge or on search strategies, is a promising future research path. Similarly, utilizing transfer learning might overcome sample size limitations of multimodal data sets. As these data sets become increasingly available, multimodal DL approaches present the opportunity to train holistic models that can learn the complex regulatory dynamics behind health and disease.

Place, publisher, year, edition, pages
Oxford University Press, 2022. Vol. 23, no 2, article id bbab569
Keywords [en]
data integration, deep neural networks, fusion strategies, multi-omics, multimodal machine learning, representation learning
National Category
Bioinformatics (Computational Biology)
Research subject
Bioinformatics
Identifiers
URN: urn:nbn:se:his:diva-20873DOI: 10.1093/bib/bbab569ISI: 000804196500091PubMedID: 35089332Scopus ID: 2-s2.0-85127534700OAI: oai:DiVA.org:his-20873DiVA, id: diva2:1633465
Funder
Knowledge Foundation, 20170302Knowledge Foundation, 20200014
Note

CC BY-NC 4.0

Corresponding author: Sören Richard Stahlschmidt. Systems Biology Research Center, University of Skövde, Skövde, Sweden. E-mail: soren.richard.stahlschmidt@his.se

Published: 28 January 2022

This work was supported by the University of Skövde, Sweden under grants from the Knowledge Foundation (20170302, 20200014).

Available from: 2022-01-31 Created: 2022-01-31 Last updated: 2022-06-23Bibliographically approved

Open Access in DiVA

fulltext(1198 kB)376 downloads
File information
File name FULLTEXT02.pdfFile size 1198 kBChecksum SHA-512
485999c6c50f2482bbcd2de3efd72f8dacb892601ba75b7ff9313d324bd209e34abc6dd7faf519edfe390cf33c942b12ef84355af791934e6b4f1037fb4929e6
Type fulltextMimetype application/pdf

Other links

Publisher's full textPubMedScopus

Authority records

Stahlschmidt, Sören RichardUlfenborg, BenjaminSynnergren, Jane

Search in DiVA

By author/editor
Stahlschmidt, Sören RichardUlfenborg, BenjaminSynnergren, Jane
By organisation
School of BioscienceSystems Biology Research Environment
In the same journal
Briefings in Bioinformatics
Bioinformatics (Computational Biology)

Search outside of DiVA

GoogleGoogle Scholar
Total: 465 downloads
The number of downloads is the sum of all downloads of full texts. It may include eg previous versions that are now no longer available

doi
pubmed
urn-nbn

Altmetric score

doi
pubmed
urn-nbn
Total: 986 hits
CiteExportLink to record
Permanent link

Direct link
Cite
Citation style
  • apa
  • apa-cv
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • Other style
More styles
Language
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Other locale
More languages
Output format
  • html
  • text
  • asciidoc
  • rtf