Parameters from site classification to harmonize MRI clinical studies: application to a multi-site Parkinson's disease dataset

Vista sencilla de ítem
dc.contributor.authorMonté Rubio, Gemma Cristina
dc.contributor.authorSegura, Barbara
dc.contributor.authorStrafella, Antonio P.
dc.contributor.authorEimeren, Thilo van
dc.contributor.authorIbarretxe Bilbao, Naroa
dc.contributor.authorDíez Cirarda, María
dc.contributor.authorEggers, Carsten
dc.contributor.authorLucas Jiménez, Olaia
dc.contributor.authorOjeda del Pozo, Natalia
dc.contributor.authorPeña Lasa, Javier
dc.contributor.authorRuppert, Marina C.
dc.contributor.authorSala Llonch, Roser
dc.contributor.authorTheis, Hendrik
dc.contributor.authorUribe, Carme
dc.contributor.authorJunqué i Plaja, Carme
dc.date.accessioned2025-05-19T09:43:21Z
dc.date.available2025-05-19T09:43:21Z
dc.date.issued2022-03-19
dc.date.updated2025-05-19T09:43:21Z
dc.description.abstractMulti-site MRI datasets are crucial for big data research. However, neuroimaging studies must face the batch effect. Here, we propose an approach that uses the predictive probabilities provided by Gaussian processes (GPs) to harmonize clinical-based studies. A multi-site dataset of 216 Parkinson's disease (PD) patients and 87 healthy subjects (HS) was used. We performed a site GP classification using MRI data. The outcomes estimated from this classification, redefined like Weighted HARMonization PArameters (WHARMPA), were used as regressors in two different clinical studies: A PD versus HS machine learning classification using GP, and a VBM comparison (FWE-p <.05, k = 100). Same studies were also conducted using conventional Boolean site covariates, and without information about site belonging. The results from site GP classification provided high scores, balanced accuracy (BAC) was 98.39% for grey matter images. PD versus HS classification performed better when the WHARMPA were used to harmonize (BAC = 78.60%; AUC = 0.90) than when using the Boolean site information (BAC = 56.31%; AUC = 0.71) and without it (BAC = 57.22%; AUC = 0.73). The VBM analysis harmonized using WHARMPA provided larger and more statistically robust clusters in regions previously reported in PD than when the Boolean site covariates or no corrections were added to the model. In conclusion, WHARMPA might encode global site-effects quantitatively and allow the harmonization of data. This method is user-friendly and provides a powerful solution, without complex implementations, to clean the analyses by removing variability associated with the differences between sites.en
dc.description.sponsorshipRegarding the University of Barcelona, this study was sponsored by the Spanish Ministry of Economy and Competitiveness (PSI2010-161, PSI2013-41393, PSI2017-86930-P) cofinanced by Agencia Estatal de Investigación (AEI) and the European Regional Development Fund (ERDF), and PID2020-114640GB-I00/AEI/ 10.13039/501100011033, by Generalitat de Catalunya (2017SGR748), Fundació La Marató de TV3 20142310, and supported by María de Maeztu Unit of Excellence (Institute of Neurosciences, University of Barcelona) MDM-2017-0729, Ministry of Science, Innovation and Universities. Regarding the University of Cologne, the study was partially sponsored by the German Research Foundation (project numbers 101434521 and 431549029). Regarding the University of Deusto, this study was sponsored by the Department of Health of the Basque Government (2011111117), the Spanish Ministry of Economy and Competitiveness (PSI2012-32441), and by the A-grade Research Team IT 946-16 from the Basque Government. AS is a consultant for Hoffman La Roche; received honoraria from GE Health Care Canada LTD, Hoffman La Roche. TvE received honoraria by Shire, Lilly, Lundbeck, and Orion Pharma and reports no conflict of interest regarding this study. CU was supported by the European Union's Horizon 2020 research and innovation programme under the Marie Sklodowska-Curie fellowship [grant agreement 888692]en
dc.identifier.citationC. Monte-Rubio, G., Segura, B., P. Strafella, A., van Eimeren, T., Ibarretxe-Bilbao, N., Diez-Cirarda, M., Eggers, C., Lucas-Jiménez, O., Ojeda, N., Peña, J., Ruppert, M. C., Sala-Llonch, R., Theis, H., Uribe, C., & Junque, C. (2022). Parameters from site classification to harmonize MRI clinical studies: application to a multi-site Parkinson's disease dataset. Human Brain Mapping, 43(10), 3130-3142. https://doi.org/10.1002/HBM.25838
dc.identifier.doi10.1002/HBM.25838
dc.identifier.eissn1097-0193
dc.identifier.issn1065-9471
dc.identifier.urihttp://hdl.handle.net/20.500.14454/2765
dc.language.isoeng
dc.publisherJohn Wiley & Sons, Inc
dc.rights© 2022 The Authors
dc.subject.otherGaussian process
dc.subject.otherMachine learning
dc.subject.otherMRI harmonization
dc.subject.otherParkinson's disease
dc.subject.otherPredictive probabilities
dc.subject.otherSite-effects
dc.subject.otherVBM
dc.titleParameters from site classification to harmonize MRI clinical studies: application to a multi-site Parkinson's disease dataseten
dc.typejournal article
dcterms.accessRightsopen access
oaire.citation.endPage3142
oaire.citation.issue10
oaire.citation.startPage3130
oaire.citation.titleHuman Brain Mapping
oaire.citation.volume43
oaire.licenseConditionhttps://creativecommons.org/licenses/by-nc-nd/4.0/
oaire.versionVoR
Archivos
Bloque original
Mostrando 1 - 1 de 1
Miniatura
Nombre:
monte_parameters_2022.pdf
Tamaño:
5.88 MB
Formato:
Adobe Portable Document Format
Descargar
Colecciones
Artículos