Show simple item record

dc.contributor.authorRakowski, Alexander G.
dc.contributor.authorVeličković, Petar
dc.contributor.authorDall’Ara, Enrico
dc.contributor.authorLiò, Pietro
dc.date.accessioned2020-02-22T04:17:06Z
dc.date.available2020-02-22T04:17:06Z
dc.date.issued2020-02-21
dc.date.submitted2019-06-14
dc.identifier.otherpone-d-19-16620
dc.identifier.urihttps://www.repository.cam.ac.uk/handle/1810/302596
dc.description.abstractChronoMID—neural networks for temporally-varying, hence Chrono, Medical Imaging Data—makes the novel application of cross-modal convolutional neural networks (X-CNNs) to the medical domain. In this paper, we present multiple approaches for incorporating temporal information into X-CNNs and compare their performance in a case study on the classification of abnormal bone remodelling in mice. Previous work developing medical models has predominantly focused on either spatial or temporal aspects, but rarely both. Our models seek to unify these complementary sources of information and derive insights in a bottom-up, data-driven approach. As with many medical datasets, the case study herein exhibits deep rather than wide data; we apply various techniques, including extensive regularisation, to account for this. After training on a balanced set of approximately 70000 images, two of the models—those using difference maps from known reference points—outperformed a state-of-the-art convolutional neural network baseline by over 30pp (> 99% vs. 68.26%) on an unseen, balanced validation set comprising around 20000 images. These models are expected to perform well with sparse data sets based on both previous findings with X-CNNs and the representations of time used, which permit arbitrarily large and irregular gaps between data points. Our results highlight the importance of identifying a suitable description of time for a problem domain, as unsuitable descriptors may not only fail to improve a model, they may in fact confound it.
dc.languageen
dc.publisherPublic Library of Science
dc.rightsAttribution 4.0 International (CC BY 4.0)en
dc.rights.urihttps://creativecommons.org/licenses/by/4.0/en
dc.subjectResearch Article
dc.subjectComputer and information sciences
dc.subjectBiology and life sciences
dc.subjectResearch and analysis methods
dc.subjectMedicine and health sciences
dc.subjectEngineering and technology
dc.titleChronoMID—Cross-modal neural networks for 3-D temporal medical imaging data
dc.typeArticle
dc.date.updated2020-02-22T04:17:06Z
prism.issueIdentifier2
prism.publicationNamePLOS ONE
prism.volume15
dc.identifier.doi10.17863/CAM.49664
dcterms.dateAccepted2020-01-27
rioxxterms.versionofrecord10.1371/journal.pone.0228962
rioxxterms.versionVoR
rioxxterms.licenseref.urihttp://creativecommons.org/licenses/by/4.0/
datacite.contributor.supervisoreditor: Pławiak, Paweł
dc.contributor.orcidRakowski, Alexander G. [0000-0002-7512-4169]
dc.identifier.eissn1932-6203
pubs.funder-project-idEngineering and Physical Sciences Research Council (EP/K03877X/1)
pubs.funder-project-idNational Centre for the Replacement, Refinement and Reduction of Animals in Research (NC/R001073/1)


Files in this item

Thumbnail
Thumbnail
Thumbnail
Thumbnail

This item appears in the following Collection(s)

Show simple item record

Attribution 4.0 International (CC BY 4.0)
Except where otherwise noted, this item's licence is described as Attribution 4.0 International (CC BY 4.0)