/MLM

Mulivariate analysis for Neuroimaging data.

Primary LanguageHTML

MLM

Mulivariate analysis for Neuroimaging data. Multivariate Methods. General multivariate tool. Contains different procedures to explore MRI or PET DATA The standard way of using the MM package is to first perform a standard spm analysis, that will provide a first apriori model and the corresponding estimated parameters and residual sum of square images. As the temporal filter or the normalisation chosen is of importance, please keep in mind the parameters used for a meaningful interpretation of your MM results.

MM embeds the concept of several spatial space such that if the regression model is performed separately on several subjects (1..N) or regions of interest have the same temporal space, MM allows you to consider your data as a matrix with dimension common-time-dim X (subject1-space-dim + subject2-space-dim + ... + subjectN-space-dim) In such a case, the result of the analysis for the first component is one time dimension eigen vector and one space dimension vector with size : (subject1-space-dim + subject2-space-dim + ... + subjectN-space-dim), which can also be considered as one eigenimage per subject (or region of interest).

More often, the MM is performed on a matrix with dimension (subject1-time-dim + subject2-time-dim + ... + subjectN-time-dim) X common-space-dim

  1. SVD analysis: Given image files and a contrast of a general linear model, this procedure perform PCA analysis on the projected data in the sub-space defined by the contrast. orthogonal projection allow to study the residual part of a model.

  2. MLM analysis: Given Beta images and a contrast of a general linear model this procedure allow to study the relatiom between the data and the model. MLM is adapted from Worsley et al (1997).

----- References ---------- MLM : Worsley KJ, Poline JB, Friston KJ, Evans AC. "Characterizing the response of PET and fMRI data using multivariate linear models." Neuroimage 1997 Nov;6(4):305-19
SVD : K.J. Friston, J.-B. Poline, S. Strother, A.P. Holmes, C.D. Frith, et R.S.J. Frackowiak, "A multivariate analysis of PET activation studies" Human brain mapping. 4:140-151, 1996.

================================================================================ CREDITS

This package was developped by Ferath Kherif primarily

A number of functions used in the toolbox belong to the SPM core package from the Wellcome Department of Cognitive Neurology, (also distributed under GNU General Public License). See www.fil.ion.ucl.ac.uk

COPYING / DISTRIBUTING

You can redistribute it and/or modify it under the terms of the GNU General Public License version 2 as published by the Free Software Foundation, which is displayed in the accompanying COPYING file. See the GNU General Public License for more details.

Please redirect requests for the toolbox to us. For bugs, remarks, additions, etc, please contact ferath.kherif@chuv.ch

If you are using this material for publication, please see with us how you can acknowledge our work.

================================================================================ WARNING : THIS SOFTWARE IS DISTRIBUTED FREE WITHOUT ANY GUARANTY !

================================================================================

  • Copyright (C) LREN

================================================================================ Uptaed version 2019