Both contrasts are right. You might want to change the value from 0.5 to 1/6, but that won't affect the actual p-values (only the gamma values). The 2nd contrast is correct, but it maybe difficult to interpret if the first contrast has an effect. Technically, you should verify that there is no interaction between group and continuous covariates. If there is no effect, then re-run with DOSS (the non-interaction model).


On 10/12/16 1:01 PM, charles laidi wrote:
Dear Douglas and Trisanna,

Thank you very much for your explanations.

I just would like to be sure of what I am doing since there is no example on the wiki as complex as mine.

If I have the following classes :

Diagnosis : Patient or Control
Gender : Male or Female
Site : Site1 or Site2 or Site3

NregressorsDODS = Nclasses*(Nvariables+1) = 6*(0+1) = 2*2*3*(1+1) = 24
Regressor1: ones for PatientMaleSite1 subjects, 0 otherwise. Codes intercept for Group 1 Regressor2: ones for PatientMaleSite2 subjects, 0 otherwise. Codes intercept for Group 2 Regressor2: ones for PatientMaleSite3 subjects, 0 otherwise. Codes intercept for Group 3
....
Regressor12: ones for ControlFemaleSite3 subjects, 0 otherwise. Codes intercept for Group 12 Regressor13: age for PatientMaleSite1 subjects, 0 otherwise. Codes age slope for Group 1 Regressor14: age for PatientMaleSite2 subjects, 0 otherwise. Codes age slope for Group 2
...
Regressor24: age for ControlFemaleSite3 subjects, 0 otherwise. Codes age slope for Group 12

*I assume my FSGDF file should look like : *

GroupDescriptorFile 1
Title analysis
Class PatientMaleSite1
Class PatientMaleSite2
Class PatientMaleSite3
Class PatientFemaleSite1
Class PatientFemaleSite2
Class PatientFemaleSite3
Class ControlMaleSite1
Class ControlMaleSite2
Class ControlMaleSite3
Class ControlFemaleSite1
Class ControlFemaleSite2
Class ControlFemaleSite3
Variables Age
Input subject1 PatientMaleSite1 30
Input subject2 PatientFemaleSite2 45
Input subject3 PatientFemaleSite3 85
Input subject4 PatientFemaleSite4 75
...

As for the contrast analysis :

If my question is : *is there a difference between Patient and Controls age slope regressing out the effects of gender and site? *

_Contrast1 patient-control.slope.mtx_

0 0 0 0 0 0 0 0 0 0 0 0 0.5 0.5 0.5 0.5 0.5 0.5 -0.5 -0.5 -0.5 -0.5 -0.5 -0.5

with all regressor1 to regressor12 equal to *0*
with regressor13 to 18 equal to *0.5 *(regressor with Patient in it)
with regressor19 to 24 equal to *- 0.5* (regressor with Controls in it)

If my question is :*is there a difference between cortical thickness in patients and controls, regressing out the effects of age, gender and site ? *

_Contrast2 patient-controls.mtx_

0.5 0.5 0.5 0.5 0.5 0.5 -0.5 -0.5 -0.5 -0.5 -0.5 -0.5 0 0 0 0 0 0 0 0 0 0 0 0

with all regressor1 to regressor6 equal to *0.5
*with all regressor7 to regressor12 equal to -*0.5*
with all regressor13 to regressor24 equal to *0*

Thank you very much in advance for your help and sorry for taking your time.

Best,

Charles










2016-10-05 17:44 GMT+02:00 Douglas N Greve <gr...@nmr.mgh.harvard.edu <mailto:gr...@nmr.mgh.harvard.edu>>:

    This is a straightforward extension to the FSGD examples. You have 3
    discrete factors (2 diagnosis, 2 gender, 6 centers), this yields
    2*2*6=24 classes. With one covariate, you would have 24 covariate
    regressors (one for each class) for a total of 48. You would then need
    to create a contrast matrix that tests for an interaction between
    diagnosis and age which is also a straight-forward extension to the
    examples.

    Having said that, I think that 48 regressors is a lot unless you have
    about 500 subjects. It is also possible to have a less agressive model
    and just have two regressors, one for each diagnosis. But you'd
    have to
    create the design matrix yourself as this is outside of the FSGD
    specification.


    On 10/05/2016 09:22 AM, charles laidi wrote:
    > Dear FreeSurfers,
    >
    > I would like to study the interaction between age and cortical
    > thickness in patients and controls.
    > My hypothesis is that there is an interaction and that cortical
    > thickness is decreasing faster with age in patients than in
    controls.
    > I have both Male and Female included in 6 different centers.
    > I would like to consider also Gender and Site (6 centers) as
    confounds.
    >
    > My understanding is that I should perform a surface based group
    > analysis with freesurfer.
    >
    > I am not able to find an example for my problem in the documentation
    > https://surfer.nmr.mgh.harvard.edu/fswiki/Fsgdf2G0V
    <https://surfer.nmr.mgh.harvard.edu/fswiki/Fsgdf2G0V>
    >
    > Would anyone had some tips to build the Fsgd file and the
    contrast file ?
    >
    > Thank you very much in advance.
    >
    > Best,
    >
    > Charles
    >
    >
    >
    > _______________________________________________
    > Freesurfer mailing list
    > Freesurfer@nmr.mgh.harvard.edu
    <mailto:Freesurfer@nmr.mgh.harvard.edu>
    > https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
    <https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer>

    --
    Douglas N. Greve, Ph.D.
    MGH-NMR Center
    gr...@nmr.mgh.harvard.edu <mailto:gr...@nmr.mgh.harvard.edu>
    Phone Number: 617-724-2358 <tel:617-724-2358>
    Fax: 617-726-7422 <tel:617-726-7422>

    Bugs: surfer.nmr.mgh.harvard.edu/fswiki/BugReporting
    <http://surfer.nmr.mgh.harvard.edu/fswiki/BugReporting>
    FileDrop: https://gate.nmr.mgh.harvard.edu/filedrop2
    <https://gate.nmr.mgh.harvard.edu/filedrop2>
    www.nmr.mgh.harvard.edu/facility/filedrop/index.html
    <http://www.nmr.mgh.harvard.edu/facility/filedrop/index.html>
    Outgoing:
    ftp://surfer.nmr.mgh.harvard.edu/transfer/outgoing/flat/greve/
    <ftp://surfer.nmr.mgh.harvard.edu/transfer/outgoing/flat/greve/>

    _______________________________________________
    Freesurfer mailing list
    Freesurfer@nmr.mgh.harvard.edu <mailto:Freesurfer@nmr.mgh.harvard.edu>
    https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
    <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
    <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

_______________________________________________
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.

Reply via email to