liu.seSearch for publications in DiVA
Change search
ReferencesLink to record
Permanent link

Direct link
Multivariate fMRI Analysis using Canonical Correlation Analysis instead of Classifiers, Comment on Todd et al
Virginia Tech, Carilion Research Institute, USA. (School of Medicine)
Linköping University, Department of Biomedical Engineering, Medical Informatics. Linköping University, The Institute of Technology.ORCID iD: 0000-0002-9091-4724
2013 (English)In: figshare.comArticle in journal, Editorial material (Other academic) Published
Abstract [en]

Multivariate pattern analysis (MVPA) is a popular method for making inference about functional magnetic resonance imaging (fMRI) data. One approach is to train a classifier with voxels within a certain radius from the center voxel, to classify between different brain states. This approach is commonly known as the searchlight algorithm. As recently pointed out by Todd and colleagues, inference at the group level can however be confounded by the fact that the direction of the effect is lost if the per subject classification performance is used to generate group results. Here we show that canonical correlation analysis (CCA) can in some aspects be a better approach to multivariate fMRI analysis, than classification based analysis (CBA).

Place, publisher, year, edition, pages
National Category
Signal Processing Medical Image Processing
URN: urn:nbn:se:liu:diva-95668DOI: 10.6084/m9.figshare.787696OAI: diva2:637194
Linnaeus research environment CADICS
Available from: 2013-07-16 Created: 2013-07-16 Last updated: 2015-06-10Bibliographically approved

Open Access in DiVA

No full text

Other links

Publisher's full text

Search in DiVA

By author/editor
Knutsson, Hans
By organisation
Medical InformaticsThe Institute of Technology
Signal ProcessingMedical Image Processing

Search outside of DiVA

GoogleGoogle Scholar
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

Altmetric score

Total: 606 hits
ReferencesLink to record
Permanent link

Direct link