Dear Freesurfers,
Is there a way to include a per voxel regressor in an LME model within the matlab LME toolbox? Just like you can do with mri_glmfit?
Cheers, Marnie
Hey Marnie, I don't think so. If Jorge is still following the FS list, he can answer more definitely. Jorge, Marnie is asking whether a different design matrix can be used at each voxel based upon a map passed to it. This is used for multimodal integration in a lot of applications (eg, does thickness at a vertex predict changes in fMRI at that vertex). doug
On 04/30/2014 06:37 PM, Marnie Shaw wrote:
Dear Freesurfers,
Is there a way to include a per voxel regressor in an LME model within the matlab LME toolbox? Just like you can do with mri_glmfit?
Cheers,
Marnie
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
The lme toolbox allows you to have different regression models across locations. If there is a missed data value at some location (vertex/voxel) for a given time-point for a given subject then you can exclude the corresponding row of the full design matrix at that location using the parameter Xrows of :
[stats,st] = lme_mass_fit_vw(X,Zcols,Y,ni,maskvtx,fname,prs,e,Xrows)
This allows you to use a different number of rows from your design matrix across locations. Xrows is just a matrix of zeros and ones. Look at the header of lme_mass_fit_vw for more info.
If you also want to have a different number of columns from your "maximal model" design matrix (this the design mtrix for the model that contains the maximum number of regressors you are going to use at any location) then you need to work directly with
[stats,st] = lme_mass_fit(X,Xcols,Xrows,Zcols,Y,ni) (this script is call by lme_mass_fit_vw)
Look at the header of this script for more info. The parameter Xcols is a matrix of zeros and ones that allows you to choose a different subset of columns from your “maximal design matrix” at each location.
Hope this makes sense.
Best -Jorge El Miércoles 30 de abril de 2014 18:47, Marnie Shaw marnie.shaw@anu.edu.au escribió:
Dear Freesurfers, Is there a way to include a per voxel regressor in an LME model within the matlab LME toolbox? Just like you can do with mri_glmfit? Cheers, Marnie _______________________________________________ Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
The information in this e-mail is intended only for the person to whom it is addressed. If you believe this e-mail was sent to you in error and the e-mail contains patient information, please contact the Partners Compliance HelpLine at http://www.partners.org/complianceline . If the e-mail was sent to you in error but does not contain patient information, please contact the sender and properly dispose of the e-mail.
Hi Jorge,
Thanks for your reply. Actually I don’t want to have different numbers of columns in the design matrix for different vertices. Instead I want to add an extra column to the design matrix, the value of which varies across vertices.
It looks like I can just add an extra column in the design matrix for each vertex within lme_mass_fit i.e. within the loop over vertices (~ line 174)
X=[X,pvr(:,i)];
Where pvr is my per-voxel-regressor and i is the vertex no.
Do you see any problem with this approach?
Thanks again,
Marnie
From: jorge luis [mailto:jbernal0019@yahoo.es] Sent: Friday, 2 May 2014 1:51 AM To: Freesurfer support list Cc: Marnie Shaw; Douglas N Greve Subject: Re: [Freesurfer] per voxel regressor in LME toolbox?
The lme toolbox allows you to have different regression models across locations. If there is a missed data value at some location (vertex/voxel) for a given time-point for a given subject then you can exclude the corresponding row of the full design matrix at that location using the parameter Xrows of :
[stats,st] = lme_mass_fit_vw(X,Zcols,Y,ni,maskvtx,fname,prs,e,Xrows)
This allows you to use a different number of rows from your design matrix across locations. Xrows is just a matrix of zeros and ones. Look at the header of lme_mass_fit_vw for more info.
If you also want to have a different number of columns from your "maximal model" design matrix (this the design mtrix for the model that contains the maximum number of regressors you are going to use at any location) then you need to work directly with
[stats,st] = lme_mass_fit(X,Xcols,Xrows,Zcols,Y,ni) (this script is call by lme_mass_fit_vw)
Look at the header of this script for more info. The parameter Xcols is a matrix of zeros and ones that allows you to choose a different subset of columns from your “maximal design matrix” at each location.
Hope this makes sense.
Best -Jorge El Miércoles 30 de abril de 2014 18:47, Marnie Shaw <marnie.shaw@anu.edu.aumailto:marnie.shaw@anu.edu.au> escribió: Dear Freesurfers,
Is there a way to include a per voxel regressor in an LME model within the matlab LME toolbox? Just like you can do with mri_glmfit?
Cheers, Marnie
_______________________________________________ Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edumailto:Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
The information in this e-mail is intended only for the person to whom it is addressed. If you believe this e-mail was sent to you in error and the e-mail contains patient information, please contact the Partners Compliance HelpLine at http://www.partners.org/complianceline . If the e-mail was sent to you in error but does not contain patient information, please contact the sender and properly dispose of the e-mail.
Yes you can do something like this:
XM = [X, pvr(:,i)];
Best -Jorge
El Jueves 1 de Mayo de 2014 23:39, Marnie Shaw marnie.shaw@anu.edu.au escribió:
Hi Jorge,
Thanks for your reply. Actually I don’t want to have different numbers of columns in the design matrix for different vertices. Instead I want to add an extra column to the design matrix, the value of which varies across vertices. It looks like I can just add an extra column in the design matrix for each vertex within lme_mass_fit i.e. within the loop over vertices (~ line 174) X=[X,pvr(:,i)]; Where pvr is my per-voxel-regressor and i is the vertex no. Do you see any problem with this approach? Thanks again, Marnie From:jorge luis [mailto:jbernal0019@yahoo.es] Sent: Friday, 2 May 2014 1:51 AM To: Freesurfer support list Cc: Marnie Shaw; Douglas N Greve Subject: Re: [Freesurfer] per voxel regressor in LME toolbox? The lme toolbox allows you to have different regression models across locations. If there is a missed data value at some location (vertex/voxel) for a given time-point for a given subject then you can exclude the corresponding row of the full design matrix at that location using the parameter Xrows of : [stats,st] = lme_mass_fit_vw(X,Zcols,Y,ni,maskvtx,fname,prs,e,Xrows) This allows you to use a different number of rows from your design matrix across locations. Xrows is just a matrix of zeros and ones. Look at the header of lme_mass_fit_vw for more info. If you also want to have a different number of columns from your "maximal model" design matrix (this the design mtrix for the model that contains the maximum number of regressors you are going to use at any location) then you need to work directly with [stats,st] = lme_mass_fit(X,Xcols,Xrows,Zcols,Y,ni) (this script is call by lme_mass_fit_vw) Look at the header of this script for more info. The parameter Xcols is a matrix of zeros and ones that allows you to choose a different subset of columns from your “maximal design matrix” at each location. Hope this makes sense. Best -Jorge El Miércoles 30 de abril de 2014 18:47, Marnie Shaw marnie.shaw@anu.edu.au escribió: Dear Freesurfers,
Is there a way to include a per voxel regressor in an LME model within the matlab LME toolbox? Just like you can do with mri_glmfit? Cheers, Marnie
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
The information in this e-mail is intended only for the person to whom it is addressed. If you believe this e-mail was sent to you in error and the e-mail contains patient information, please contact the Partners Compliance HelpLine at http://www.partners.org/complianceline . If the e-mail was sent to you in error but does not contain patient information, please contact the sender and properly dispose of the e-mail.
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
The information in this e-mail is intended only for the person to whom it is addressed. If you believe this e-mail was sent to you in error and the e-mail contains patient information, please contact the Partners Compliance HelpLine at http://www.partners.org/complianceline . If the e-mail was sent to you in error but does not contain patient information, please contact the sender and properly dispose of the e-mail.
freesurfer@nmr.mgh.harvard.edu
-
Douglas N Greve -
jorge luis -
Marnie Shaw