Repository logo
 
Publication

Explainable prototype-based image classification using adaptive feature extractors in medical images

2025-12Journal article Open access
Simple item page
datacite.subject.fosCiências Médicas::Ciências da Saúde
datacite.subject.fosEngenharia e Tecnologia
datacite.subject.sdg03:Saúde de Qualidade
datacite.subject.sdg09:Indústria, Inovação e Infraestruturas
datacite.subject.sdg10:Reduzir as Desigualdades
dc.contributor.authorVasconcellos, Nicolas
dc.contributor.authorTavora, Luis M. N.
dc.contributor.authorMiragaia, Rolando
dc.contributor.authorGrilo, Carlos
dc.contributor.authorThomaz, Lucas
dc.date.accessioned2026-01-06T16:30:14Z
dc.date.available2026-01-06T16:30:14Z
dc.date.issued2025-12
dc.descriptionArticle number - 111322.
dc.description.abstractPrototype-based classifiers are a category of Explainable Artificial Intelligence methods that use representative samples from the data, called prototypes, to classify new inputs based on a similarity criterion. However, these methods often rely on pre-trained Convolutional Neural Networks as feature extractors, which may not be adapted for the specific type of data being used, thus not suited for identifying the most representative prototypes. In this paper, we propose a method named Explainable Prototype-based Image Classification, a cluster-oriented training strategy that enhances the performance and explainability of prototype-based classifiers. Our method uses a novel loss function, called Cluster Density Error, to fine-tune the feature extractor and preserve the most representative feature vectors in the latent space. We also use Principal Component Analysis-based approach to reduce the dimensionality and complexity of the feature vectors. We conduct experiments on four medical image datasets and compare the results with those from different prototype-based classifiers and state-of-the-art non-explainable learning methods. The proposed method demonstrated superior explainable capabilities and comparable classification performance to the compared methods. Specifically, the proposed method achieved up to 95.01% accuracy and 0.992 AUC using only 43 prototypes. This translated to an improvement in accuracy and AUC score of 21.54% and 9.06%, respectively, and a substantial reduction in the number of prototypes by 98,38%eng
dc.description.sponsorshipThis work is funded by the Polytechnic of Leiria under the project ‘‘1◦ Prémio + Ciência 2024’’, national funds through FCT – Fundação para a Ciência e a Tecnologia, I.P. under projects CoMBINNe 2022.09914.PTDC (DOI: 10.54499/2022.09914.PTDC), 2023.07886. CEECIND (DOI: 10.54499/2023.07886.CEECIND/CP2862/CT0003), and UID/4524/2025 (DOI: 10.54499/UID/04524/2025), and, when eligible, co-funded by EU funds under project/support UID/50008/2025 – Instituto de Telecomunicações, with DOI identifier 10.54499/UID/5008/2025 and LA/P/0109/2020 (DOI: 10.54499/LA/P/0109/2020).
dc.identifier.citationNicolas Vasconcellos, Luis M.N. Tavora, Rolando Miragaia, Carlos Grilo, Lucas A. Thomaz, Explainable prototype-based image classification using adaptive feature extractors in medical images, Computers in Biology and Medicine, Volume 199, 2025, 111322, ISSN 0010-4825, https://doi.org/10.1016/j.compbiomed.2025.111322
dc.identifier.doi10.1016/j.compbiomed.2025.111322
dc.identifier.issn0010-4825
dc.identifier.urihttp://hdl.handle.net/10400.8/15231
dc.language.isoeng
dc.peerreviewedyes
dc.publisherElsevier
dc.relationCompression of Multimodal Biomedical Images using Neural Networks
dc.relationInstitute of Telecommunications
dc.relation.hasversionhttps://www.sciencedirect.com/science/article/pii/S0010482525016762?via%3Dihub
dc.relation.ispartofComputers in Biology and Medicine
dc.rights.urihttp://creativecommons.org/licenses/by-nc-nd/4.0/
dc.subjectExplainable artificial intelligence
dc.subjectExplainable classification
dc.subjectPrototype-based classifiers
dc.subjectClustering
dc.subjectAdaptive feature extractors
dc.subjectMedical imaging
dc.subjectMedical imaging classification
dc.titleExplainable prototype-based image classification using adaptive feature extractors in medical imageseng
dc.typejournal article
dspace.entity.typePublication
oaire.awardTitleCompression of Multimodal Biomedical Images using Neural Networks
oaire.awardTitleInstitute of Telecommunications
oaire.awardURIinfo:eu-repo/grantAgreement/FCT/3599-PPCDT/2022.09914.PTDC/PT
oaire.awardURIinfo:eu-repo/grantAgreement/FCT/6817 - DCRRNI ID/LA%2FP%2F0109%2F2020/PT
oaire.citation.titleComputers in Biology and Medicine
oaire.citation.volume199
oaire.fundingStream3599-PPCDT
oaire.fundingStream6817 - DCRRNI ID
oaire.versionhttp://purl.org/coar/version/c_970fb48d4fbd8a85
person.affiliation.nameCIIC / ESTG
person.familyNameFreire Vasconcellos
person.familyNamede Oliveira Pegado de Noronha E Távora
person.familyNameMiragaia
person.familyNameGrilo
person.familyNameThomaz
person.givenNameNicolas David
person.givenNameLuís Miguel
person.givenNameRolando
person.givenNameCarlos
person.givenNameLucas
person.identifier.ciencia-id8410-201C-07BD
person.identifier.ciencia-id121C-FADA-D750
person.identifier.ciencia-idC712-E02E-0ED2
person.identifier.ciencia-id081D-025A-33BC
person.identifier.orcid0009-0009-9208-0020
person.identifier.orcid0000-0002-8580-1979
person.identifier.orcid0000-0003-4213-9302
person.identifier.orcid0000-0001-9727-905X
person.identifier.orcid0000-0002-1004-7772
person.identifier.ridGLS-3615-2022
person.identifier.ridM-9551-2013
person.identifier.scopus-author-id26422369700
person.identifier.scopus-author-id23466972200
project.funder.identifierhttp://doi.org/10.13039/501100001871
project.funder.identifierhttp://doi.org/10.13039/501100001871
project.funder.nameFundação para a Ciência e a Tecnologia
project.funder.nameFundação para a Ciência e a Tecnologia
relation.isAuthorOfPublicationfeba3b54-be4c-49e4-84f4-e4e05bb13e67
relation.isAuthorOfPublication71940f24-f333-4ab6-abf6-00c7119a07c2
relation.isAuthorOfPublicationc3934650-8cbe-40cd-bb29-31c57baa49e2
relation.isAuthorOfPublicationf2075503-40b3-49da-81c3-e0c50f97c9ab
relation.isAuthorOfPublication5aaecc29-3e2f-49a6-9fd9-7000d9f6085f
relation.isAuthorOfPublication.latestForDiscovery71940f24-f333-4ab6-abf6-00c7119a07c2
relation.isProjectOfPublication5fe22f63-2c55-4683-bae7-89d5dba06107
relation.isProjectOfPublication43215f6c-bfb6-4829-8e75-6096384ce6db
relation.isProjectOfPublication.latestForDiscovery5fe22f63-2c55-4683-bae7-89d5dba06107

Files

Original bundle
Now showing 1 - 1 of 1
Thumbnail Image
Name:
1-s2.0-S0010482525016762-main.pdf
Size:
4.74 MB
Format:
Adobe Portable Document Format
Download
License bundle
Now showing 1 - 1 of 1
No Thumbnail Available
Name:
license.txt
Size:
1.32 KB
Format:
Item-specific license agreed upon to submission
Description:
Download

Collections

CIIC - Artigos em Revistas com Peer Review
ESTG - Artigos em revistas internacionais