Re: [Freesurfer] Design Matrix Creation

[Douglas N Greve]

I don't think you can do that. A subject must belong to both a gender 
and a dx class, so you have to do it as an interaction.


On 07/06/2016 03:38 PM, Timothy Hendrickson wrote:
> Hi Doug,
> I just wanted to make sure that we were on the same page. I may have been a 
> little unclear in

> what I would ideally like to do.
> For #2 what I would like to do is take into account age slopes and total 
> intracranial slopes for

>   gender, and age slopes and total intracranial slopes for diagnosis.
> I was not implying that I wanted diagnosis and gender to interact.
>
> Assuming that you understood this previously could you please explain why you 
> would model

> the matrix this way? I do not fully understand.
>
> Respectfully,
>
> -Tim
> >I did not get the same matrix. Here's how I would do it
> >Columns 1-16 model the intercepts for all your classes
> >Columns 17-20 model the ages for your gender-x-diagnosis subclasses
> >Columns 21-24 model the ICV for your gender-x-diagnosis subclasses
> On 07/06/2016 02:04 PM, Timothy Hendrickson wrote:
> > Freesurfer Support,
> >
> > Thank you for your support with my previous line of questions regarding
> > design matrix creation.
> > I have manually created another design matrix and want to ensure that I
> > designed it correctly.
> >
> > An example of my FSGD file is as follows below:
> > Class SITE 1-Male-Control
> > Class SITE 1-Male-PATIENT
> > Class SITE 1-Female-Control
> > Class SITE 1-Female-PATIENT
> > Class SITE 2-Male-Control
> > Class SITE 2-Male-PATIENT
> > Class SITE 2-Female-Control
> > Class SITE 2-Female-PATIENT
> > Class SITE 3-Male-Control
> > Class SITE 3-Male-PATIENT
> > Class SITE 3-Female-Control
> > Class SITE 3-Female-PATIENT
> > Class SITE 4-Male-Control
> > Class SITE 4-Male-PATIENT
> > Class SITE 4-Female-Control
> > Class SITE 4-Female-PATIENT
> > Variables age_at_scan total_intracranial_volume
> > study site levels = 1, 2, 3, 4
> > gender levels = Male, Female
> > diagnosis levels = PATIENT, Control
> > age_at_scan = covariate age
> > total_intracranial_volume = covariate total intracranial volume
> >
> > What I would ideally like to do is:
> >
> > 1) Take into account offset differences amongst diagnosis, gender, and study
> > site.
> >
> > 2) Allowing a difference in age slope, and total intracranial volume slope
> > amongst the diagnosis and gender levels.
> >
> > 3) Modeling the age slope and total intracranial volume slope as the same 
> > for
> > the study site levels.
> > Let's image that the first participant is from site 1, Male, and control, 
> > and
> > is 12 and has a TIV of 30,000.
> > The second participant is from site 2, Female, and PATIENT, and is 14, and
> > has a TIV of 25,000.
> > My understanding of the design matrix would be as follows:
> > 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 12 0 0 0 3 0 0 0
> > 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 14 0 0 0 25000
> > -Tim
> >
> > >Yes, create your matrix manually.
> > >
> > >Those matrix lines are not quite right. The ages are in the correct
> > >column, but you need a 1 somewhere in columns 1-8 to indicate the class
> > >(ie, site/dx) that the subject is in.
> > On 05/26/2016 12:57 PM, Timothy Hendrickson wrote:
> > > Hi Doug,
> > >
> > > Thank you for such a prompt response. Just to be clear you are
> > > recommending that I manually create the matrix file right?
> > >
> > > If so I want to ensure that I am understanding how to design the
> > > matrix file properly.
> > >
> > > Let's imagine that the first participant is a control and is 13 and
> > > the second is a patient and is 15. My understanding is that the matrix
> > > file would be as follows:
> > > 0 0 0 0 0 0 0 0 13 0
> > > 0 0 0 0 0 0 0 0 0 15.
> > >
> > > -Tim
> > > Previous correspondences are below:
> > >
> > > You'll need a regressor for each of the 8 classes you describe below.
> > > You can use mri_glmfit to generate this (Xg.dat file)
> > > You'll need two more regressors for age, one for each diagnosis. If a
> > > subject (ie, row) is a control then the two values will be AGE 0. If the
> > > subject of the row is a patient, then the two values will be 0 AGE. You
> > > can then set up a Controls-Patients age (ie, interaction between dx and
> > > age) contrast like
> > > [0 0 0 0 0 0 0 0 1 -1]
> > > On 05/24/2016 02:30 PM, Timothy Hendrickson wrote:
> > > >
> > > > Freesurfer Support,
> > > >
> > > > I'd like to create a design matrix for a group analysis outside of the
> > > > DODS and DOSS models. I understand that in order to do this the -X
> > > > flag must be used. However, I have been unable to find examples of how
> > > > to do this.
> > > >
> > > > I am hoping to reveal a difference in thickness or gyrification
> > > > amongst a clinical population. The data set contains two factors:
> > > > diagnosis, and study site and one covariate: age. Diagnosis has two
> > > > levels: controls, and patients. Study site has four levels, one level
> > > > for each location the data has been collected from.
> > > >
> 

[Freesurfer] Design Matrix Creation

[Timothy Hendrickson]

Hi Doug,

I just wanted to make sure that we were on the same page. I may have
been a little unclear in what I would ideally like to do.

For #2 what I would like to do is take into account age slopes and
total intracranial slopes for gender, and age slopes and total
intracranial slopes for diagnosis.
I was not implying that I wanted diagnosis and gender to interact.

Assuming that you understood this previously could you please explain
why you would model the matrix this way? I do not fully understand.

Respectfully,

-Tim


>I did not get the same matrix. Here's how I would do it
>Columns 1-16 model the intercepts for all your classes
>Columns 17-20 model the ages for your gender-x-diagnosis subclasses
>Columns 21-24 model the ICV for your gender-x-diagnosis subclasses

On 07/06/2016 02:04 PM, Timothy Hendrickson wrote:
> Freesurfer Support,
>
> Thank you for your support with my previous line of questions regarding
> design matrix creation.
> I have manually created another design matrix and want to ensure that I
> designed it correctly.
>
> An example of my FSGD file is as follows below:
> Class SITE 1-Male-Control
> Class SITE 1-Male-PATIENT
> Class SITE 1-Female-Control
> Class SITE 1-Female-PATIENT
> Class SITE 2-Male-Control
> Class SITE 2-Male-PATIENT
> Class SITE 2-Female-Control
> Class SITE 2-Female-PATIENT
> Class SITE 3-Male-Control
> Class SITE 3-Male-PATIENT
> Class SITE 3-Female-Control
> Class SITE 3-Female-PATIENT
> Class SITE 4-Male-Control
> Class SITE 4-Male-PATIENT
> Class SITE 4-Female-Control
> Class SITE 4-Female-PATIENT
> Variables age_at_scan total_intracranial_volume
> study site levels = 1, 2, 3, 4
> gender levels = Male, Female
> diagnosis levels = PATIENT, Control
> age_at_scan = covariate age
> total_intracranial_volume = covariate total intracranial volume
>
> What I would ideally like to do is:
>
> 1) Take into account offset differences amongst diagnosis, gender, and study
> site.
>
> 2) Allowing a difference in age slope, and total intracranial volume slope
> amongst the diagnosis and gender levels.
>
> 3) Modeling the age slope and total intracranial volume slope as the same for
> the study site levels.
> Let's image that the first participant is from site 1, Male, and control, and
> is 12 and has a TIV of 30,000.
> The second participant is from site 2, Female, and PATIENT, and is 14, and
> has a TIV of 25,000.
> My understanding of the design matrix would be as follows:
> 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 12 0 0 0 3 0 0 0
> 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 14 0 0 0 25000
> -Tim
>
> >Yes, create your matrix manually.
> >
> >Those matrix lines are not quite right. The ages are in the correct
> >column, but you need a 1 somewhere in columns 1-8 to indicate the class
> >(ie, site/dx) that the subject is in.
> On 05/26/2016 12:57 PM, Timothy Hendrickson wrote:
> > Hi Doug,
> >
> > Thank you for such a prompt response. Just to be clear you are
> > recommending that I manually create the matrix file right?
> >
> > If so I want to ensure that I am understanding how to design the
> > matrix file properly.
> >
> > Let's imagine that the first participant is a control and is 13 and
> > the second is a patient and is 15. My understanding is that the matrix
> > file would be as follows:
> > 0 0 0 0 0 0 0 0 13 0
> > 0 0 0 0 0 0 0 0 0 15.
> >
> > -Tim
> > Previous correspondences are below:
> >
> > You'll need a regressor for each of the 8 classes you describe below.
> > You can use mri_glmfit to generate this (Xg.dat file)
> > You'll need two more regressors for age, one for each diagnosis. If a
> > subject (ie, row) is a control then the two values will be AGE 0. If the
> > subject of the row is a patient, then the two values will be 0 AGE. You
> > can then set up a Controls-Patients age (ie, interaction between dx and
> > age) contrast like
> > [0 0 0 0 0 0 0 0 1 -1]
> > On 05/24/2016 02:30 PM, Timothy Hendrickson wrote:
> > >
> > > Freesurfer Support,
> > >
> > > I'd like to create a design matrix for a group analysis outside of the
> > > DODS and DOSS models. I understand that in order to do this the -X
> > > flag must be used. However, I have been unable to find examples of how
> > > to do this.
> > >
> > > I am hoping to reveal a difference in thickness or gyrification
> > > amongst a clinical population. The data set contains two factors:
> > > diagnosis, and study site and one covariate: age. Diagnosis has two
> > > levels: controls, and patients. Study site has four levels, one level
> > > for each location the data has been collected from.
> > >
> > > What I would ideally like to do is:
> > >
> > > 1) Take into account offset differences amongst diagnosis and study site.
> > >
> > > 2) Allowing a difference in age slope amongst the diagnosis levels
> > >
> > > 3) Modeling the age slope as the same for the study site levels
> > >
> > > My FSGD file is designed as follows
> > >
> > > Class SITE 1-Control
> > > Class SITE 1-PATIENT
> > > Class SITE 2-Control
> > > 

Re: [Freesurfer] Design Matrix Creation

[Douglas N Greve]

I did not get the same matrix. Here's how I would do it
Columns 1-16 model the intercepts for all your classes
Columns 17-20 model the ages for your gender-x-diagnosis subclasses
Columns 21-24 model the ICV for your gender-x-diagnosis subclasses


On 07/06/2016 02:04 PM, Timothy Hendrickson wrote:
> Freesurfer Support,
>
> Thank you for your support with my previous line of questions regarding 
> design matrix creation.
> I have manually created another design matrix and want to ensure that I 
> designed it correctly.
>
> An example of my FSGD file is as follows below:
> Class SITE 1-Male-Control
> Class SITE 1-Male-PATIENT
> Class SITE 1-Female-Control
> Class SITE 1-Female-PATIENT
> Class SITE 2-Male-Control
> Class SITE 2-Male-PATIENT
> Class SITE 2-Female-Control
> Class SITE 2-Female-PATIENT
> Class SITE 3-Male-Control
> Class SITE 3-Male-PATIENT
> Class SITE 3-Female-Control
> Class SITE 3-Female-PATIENT
> Class SITE 4-Male-Control
> Class SITE 4-Male-PATIENT
> Class SITE 4-Female-Control
> Class SITE 4-Female-PATIENT
> Variables age_at_scan total_intracranial_volume
> study site levels = 1, 2, 3, 4
> gender levels = Male, Female
> diagnosis levels = PATIENT, Control
> age_at_scan = covariate age
> total_intracranial_volume = covariate total intracranial volume
>
> What I would ideally like to do is:
>
> 1) Take into account offset differences amongst diagnosis, gender, and study 
> site.
>
> 2) Allowing a difference in age slope, and total intracranial volume slope 
> amongst the diagnosis and gender levels.
>
> 3) Modeling the age slope and total intracranial volume slope as the same for 
> the study site levels.
> Let's image that the first participant is from site 1, Male, and control, and 
> is 12 and has a TIV of 30,000.
> The second participant is from site 2, Female, and PATIENT, and is 14, and 
> has a TIV of 25,000.
> My understanding of the design matrix would be as follows:
> 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 12 0 0 0 3 0 0 0
> 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 14 0 0 0 25000
> -Tim
>
> >Yes, create your matrix manually.
> >
> >Those matrix lines are not quite right. The ages are in the correct
> >column, but you need a 1 somewhere in columns 1-8 to indicate the class
> >(ie, site/dx) that the subject is in.
> On 05/26/2016 12:57 PM, Timothy Hendrickson wrote:
> > Hi Doug,
> >
> > Thank you for such a prompt response. Just to be clear you are
> > recommending that I manually create the matrix file right?
> >
> > If so I want to ensure that I am understanding how to design the
> > matrix file properly.
> >
> > Let's imagine that the first participant is a control and is 13 and
> > the second is a patient and is 15. My understanding is that the matrix
> > file would be as follows:
> > 0 0 0 0 0 0 0 0 13 0
> > 0 0 0 0 0 0 0 0 0 15.
> >
> > -Tim
> > Previous correspondences are below:
> >
> > You'll need a regressor for each of the 8 classes you describe below.
> > You can use mri_glmfit to generate this (Xg.dat file)
> > You'll need two more regressors for age, one for each diagnosis. If a
> > subject (ie, row) is a control then the two values will be AGE 0. If the
> > subject of the row is a patient, then the two values will be 0 AGE. You
> > can then set up a Controls-Patients age (ie, interaction between dx and
> > age) contrast like
> > [0 0 0 0 0 0 0 0 1 -1]
> > On 05/24/2016 02:30 PM, Timothy Hendrickson wrote:
> > >
> > > Freesurfer Support,
> > >
> > > I'd like to create a design matrix for a group analysis outside of the
> > > DODS and DOSS models. I understand that in order to do this the -X
> > > flag must be used. However, I have been unable to find examples of how
> > > to do this.
> > >
> > > I am hoping to reveal a difference in thickness or gyrification
> > > amongst a clinical population. The data set contains two factors:
> > > diagnosis, and study site and one covariate: age. Diagnosis has two
> > > levels: controls, and patients. Study site has four levels, one level
> > > for each location the data has been collected from.
> > >
> > > What I would ideally like to do is:
> > >
> > > 1) Take into account offset differences amongst diagnosis and study site.
> > >
> > > 2) Allowing a difference in age slope amongst the diagnosis levels
> > >
> > > 3) Modeling the age slope as the same for the study site levels
> > >
> > > My FSGD file is designed as follows
> > >
> > > Class SITE 1-Control
> > > Class SITE 1-PATIENT
> > > Class SITE 2-Control
> > > Class SITE 2-PATIENT
> > > Class SITE 3-Control
> > > Class SITE 3-PATIENT
> > > Class SITE 4-Control
> > > Class SITE 4-PATIENT
> > > Variables age_at_scan
> > >
> > > study site levels = 1,2,3 and 4
> > > diagnosis levels = PATIENT and Control
> > > age_at_scan = covariate age
> > >
> > > Any advice would be greatly appreciated.
> > >
> > > Respectfully,
> > >
> > > Tim
> > >
> > > --
> > > Timothy Hendrickson
> > > Department of Psychiatry
> > > University of Minnesota
> > > Mobile: 507-259-3434  (texts okay)
>
>
> 

[Freesurfer] Design Matrix Creation

[Timothy Hendrickson]

Freesurfer Support,

Thank you for your support with my previous line of questions
regarding design matrix creation.
I have manually created another design matrix and want to ensure that
I designed it correctly.

An example of my FSGD file is as follows below:
Class SITE 1-Male-Control
Class SITE 1-Male-PATIENT
Class SITE 1-Female-Control
Class SITE 1-Female-PATIENT
Class SITE 2-Male-Control
Class SITE 2-Male-PATIENT
Class SITE 2-Female-Control
Class SITE 2-Female-PATIENT
Class SITE 3-Male-Control
Class SITE 3-Male-PATIENT
Class SITE 3-Female-Control
Class SITE 3-Female-PATIENT
Class SITE 4-Male-Control
Class SITE 4-Male-PATIENT
Class SITE 4-Female-Control
Class SITE 4-Female-PATIENT
Variables age_at_scan total_intracranial_volume

study site levels = 1, 2, 3, 4

gender levels = Male, Female

diagnosis levels = PATIENT, Control

age_at_scan = covariate age

total_intracranial_volume = covariate total intracranial volume

What I would ideally like to do is:
1) Take into account offset differences amongst diagnosis, gender, and
study site.
2) Allowing a difference in age slope, and total intracranial volume
slope amongst the diagnosis and gender levels.

3) Modeling the age slope and total intracranial volume slope as the
same for the study site levels.

Let's image that the first participant is from site 1, Male, and
control, and is 12 and has a TIV of 30,000.
The second participant is from site 2, Female, and PATIENT, and is 14,
and has a TIV of 25,000.

My understanding of the design matrix would be as follows:
1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 12 0 0 0 3 0 0 0
0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 14 0 0 0 25000

-Tim



>Yes, create your matrix manually.
>
>Those matrix lines are not quite right. The ages are in the correct
>column, but you need a 1 somewhere in columns 1-8 to indicate the class
>(ie, site/dx) that the subject is in.

On 05/26/2016 12:57 PM, Timothy Hendrickson wrote:
> Hi Doug,
>
> Thank you for such a prompt response. Just to be clear you are
> recommending that I manually create the matrix file right?
>
> If so I want to ensure that I am understanding how to design the
> matrix file properly.
>
> Let's imagine that the first participant is a control and is 13 and
> the second is a patient and is 15. My understanding is that the matrix
> file would be as follows:
> 0 0 0 0 0 0 0 0 13 0
> 0 0 0 0 0 0 0 0 0 15.
>
> -Tim
> Previous correspondences are below:
>
> You'll need a regressor for each of the 8 classes you describe below.
> You can use mri_glmfit to generate this (Xg.dat file)
> You'll need two more regressors for age, one for each diagnosis. If a
> subject (ie, row) is a control then the two values will be AGE 0. If the
> subject of the row is a patient, then the two values will be 0 AGE. You
> can then set up a Controls-Patients age (ie, interaction between dx and
> age) contrast like
> [0 0 0 0 0 0 0 0 1 -1]
> On 05/24/2016 02:30 PM, Timothy Hendrickson wrote:
> >
> > Freesurfer Support,
> >
> > I'd like to create a design matrix for a group analysis outside of the
> > DODS and DOSS models. I understand that in order to do this the -X
> > flag must be used. However, I have been unable to find examples of how
> > to do this.
> >
> > I am hoping to reveal a difference in thickness or gyrification
> > amongst a clinical population. The data set contains two factors:
> > diagnosis, and study site and one covariate: age. Diagnosis has two
> > levels: controls, and patients. Study site has four levels, one level
> > for each location the data has been collected from.
> >
> > What I would ideally like to do is:
> >
> > 1) Take into account offset differences amongst diagnosis and study site.
> >
> > 2) Allowing a difference in age slope amongst the diagnosis levels
> >
> > 3) Modeling the age slope as the same for the study site levels
> >
> > My FSGD file is designed as follows
> >
> > Class SITE 1-Control
> > Class SITE 1-PATIENT
> > Class SITE 2-Control
> > Class SITE 2-PATIENT
> > Class SITE 3-Control
> > Class SITE 3-PATIENT
> > Class SITE 4-Control
> > Class SITE 4-PATIENT
> > Variables age_at_scan
> >
> > study site levels = 1,2,3 and 4
> > diagnosis levels = PATIENT and Control
> > age_at_scan = covariate age
> >
> > Any advice would be greatly appreciated.
> >
> > Respectfully,
> >
> > Tim
> >
> > --
> > Timothy Hendrickson
> > Department of Psychiatry
> > University of Minnesota
> > Mobile: 507-259-3434  (texts okay)
___
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

Re: [Freesurfer] Design Matrix Creation

[Douglas N Greve]

Yes, create your matrix manually.

Those matrix lines are not quite right. The ages are in the correct 
column, but you need a 1 somewhere in columns 1-8 to indicate the class 
(ie, site/dx) that the subject is in.

On 05/26/2016 12:57 PM, Timothy Hendrickson wrote:
> Hi Doug,
>
> Thank you for such a prompt response. Just to be clear you are 
> recommending that I manually create the matrix file right?
>
> If so I want to ensure that I am understanding how to design the 
> matrix file properly.
>
> Let's imagine that the first participant is a control and is 13 and 
> the second is a patient and is 15. My understanding is that the matrix 
> file would be as follows:
> 0 0 0 0 0 0 0 0 13 0
> 0 0 0 0 0 0 0 0 0 15.
>
> -Tim
> Previous correspondences are below:
>
> You'll need a regressor for each of the 8 classes you describe below.
> You can use mri_glmfit to generate this (Xg.dat file)
> You'll need two more regressors for age, one for each diagnosis. If a
> subject (ie, row) is a control then the two values will be AGE 0. If the
> subject of the row is a patient, then the two values will be 0 AGE. You
> can then set up a Controls-Patients age (ie, interaction between dx and
> age) contrast like
> [0 0 0 0 0 0 0 0 1 -1]
> On 05/24/2016 02:30 PM, Timothy Hendrickson wrote:
> >
> > Freesurfer Support,
> >
> > I'd like to create a design matrix for a group analysis outside of the
> > DODS and DOSS models. I understand that in order to do this the -X
> > flag must be used. However, I have been unable to find examples of how
> > to do this.
> >
> > I am hoping to reveal a difference in thickness or gyrification
> > amongst a clinical population. The data set contains two factors:
> > diagnosis, and study site and one covariate: age. Diagnosis has two
> > levels: controls, and patients. Study site has four levels, one level
> > for each location the data has been collected from.
> >
> > What I would ideally like to do is:
> >
> > 1) Take into account offset differences amongst diagnosis and study site.
> >
> > 2) Allowing a difference in age slope amongst the diagnosis levels
> >
> > 3) Modeling the age slope as the same for the study site levels
> >
> > My FSGD file is designed as follows
> >
> > Class SITE 1-Control
> > Class SITE 1-PATIENT
> > Class SITE 2-Control
> > Class SITE 2-PATIENT
> > Class SITE 3-Control
> > Class SITE 3-PATIENT
> > Class SITE 4-Control
> > Class SITE 4-PATIENT
> > Variables age_at_scan
> >
> > study site levels = 1,2,3 and 4
> > diagnosis levels = PATIENT and Control
> > age_at_scan = covariate age
> >
> > Any advice would be greatly appreciated.
> >
> > Respectfully,
> >
> > Tim
> >
> > --
> > Timothy Hendrickson
> > Department of Psychiatry
> > University of Minnesota
> > Mobile: 507-259-3434  (texts okay)
> >
> >
> > ___
> > Freesurfer mailing list
> >Freesurfer@nmr.mgh.harvard.edu 
> >https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
>
> -- 
> Douglas N. Greve, Ph.D.
> MGH-NMR Center
> gr...@nmr.mgh.harvard.edu 
> Phone Number:617-724-2358 
> Fax:617-726-7422 
>
> Bugs:surfer.nmr.mgh.harvard.edu/fswiki/BugReporting 
> 
> FileDrop:https://gate.nmr.mgh.harvard.edu/filedrop2
> www.nmr.mgh.harvard.edu/facility/filedrop/index.html 
> 
> Outgoing:ftp://surfer.nmr.mgh.harvard.edu/transfer/outgoing/flat/greve/
>
> ___
> 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.
>
>
> On Tue, May 24, 2016 at 1:30 PM, Timothy Hendrickson  > wrote:
>
>
> Freesurfer Support,
>
> I'd like to create a design matrix for a group analysis outside of
> the DODS and DOSS models. I understand that in order to do this
> the -X flag must be used. However, I have been unable to find
> examples of how to do this.
>
> I am hoping to reveal a difference in thickness or gyrification
> amongst a clinical population. The data set contains two factors:
> diagnosis, and study site and one covariate: age. Diagnosis has
> two levels: controls, and patients. Study site has four levels,
> one level for each location the data has been collected from.
>
> What I would ideally like to do is:
>
> 1) Take 

Re: [Freesurfer] Design Matrix Creation

[Timothy Hendrickson]

Hi Doug,

Thank you for such a prompt response. Just to be clear you are recommending
that I manually create the matrix file right?

If so I want to ensure that I am understanding how to design the matrix
file properly.

Let's imagine that the first participant is a control and is 13 and the
second is a patient and is 15. My understanding is that the matrix file
would be as follows:
0 0 0 0 0 0 0 0 13 0
0 0 0 0 0 0 0 0 0 15.

-Tim

Previous correspondences are below:

You'll need a regressor for each of the 8 classes you describe below.
You can use mri_glmfit to generate this (Xg.dat file)
You'll need two more regressors for age, one for each diagnosis. If a
subject (ie, row) is a control then the two values will be AGE 0. If the
subject of the row is a patient, then the two values will be 0 AGE. You
can then set up a Controls-Patients age (ie, interaction between dx and
age) contrast like
[0 0 0 0 0 0 0 0 1 -1]

On 05/24/2016 02:30 PM, Timothy Hendrickson wrote:
>
> Freesurfer Support,
>
> I'd like to create a design matrix for a group analysis outside of the
> DODS and DOSS models. I understand that in order to do this the -X
> flag must be used. However, I have been unable to find examples of how
> to do this.
>
> I am hoping to reveal a difference in thickness or gyrification
> amongst a clinical population. The data set contains two factors:
> diagnosis, and study site and one covariate: age. Diagnosis has two
> levels: controls, and patients. Study site has four levels, one level
> for each location the data has been collected from.
>
> What I would ideally like to do is:
>
> 1) Take into account offset differences amongst diagnosis and study site.
>
> 2) Allowing a difference in age slope amongst the diagnosis levels
>
> 3) Modeling the age slope as the same for the study site levels
>
> My FSGD file is designed as follows
>
> Class SITE 1-Control
> Class SITE 1-PATIENT
> Class SITE 2-Control
> Class SITE 2-PATIENT
> Class SITE 3-Control
> Class SITE 3-PATIENT
> Class SITE 4-Control
> Class SITE 4-PATIENT
> Variables age_at_scan
>
> study site levels = 1,2,3 and 4
> diagnosis levels = PATIENT and Control
> age_at_scan = covariate age
>
> Any advice would be greatly appreciated.
>
> Respectfully,
>
> Tim
>
> --
> Timothy Hendrickson
> Department of Psychiatry
> University of Minnesota
> Mobile: 507-259-3434  (texts okay)
>
>
> ___
> Freesurfer mailing list
> Freesurfer@nmr.mgh.harvard.edu
> https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer

-- 
Douglas N. Greve, Ph.D.
MGH-NMR centergr...@nmr.mgh.harvard.edu
Phone Number: 617-724-2358
Fax: 617-726-7422

Bugs: surfer.nmr.mgh.harvard.edu/fswiki/BugReporting
FileDrop: 
https://gate.nmr.mgh.harvard.edu/filedrop2www.nmr.mgh.harvard.edu/facility/filedrop/index.html
Outgoing: ftp://surfer.nmr.mgh.harvard.edu/transfer/outgoing/flat/greve/

___
Freesurfer mailing
listfreesur...@nmr.mgh.harvard.eduhttps://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 athttp://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.



On Tue, May 24, 2016 at 1:30 PM, Timothy Hendrickson 
wrote:

>
> Freesurfer Support,
>
> I'd like to create a design matrix for a group analysis outside of the
> DODS and DOSS models. I understand that in order to do this the -X flag
> must be used. However, I have been unable to find examples of how to do
> this.
>
> I am hoping to reveal a difference in thickness or gyrification amongst a
> clinical population. The data set contains two factors: diagnosis, and
> study site and one covariate: age. Diagnosis has two levels: controls, and
> patients. Study site has four levels, one level for each location the data
> has been collected from.
>
> What I would ideally like to do is:
>
> 1) Take into account offset differences amongst diagnosis and study site.
>
> 2) Allowing a difference in age slope amongst the diagnosis levels
>
> 3) Modeling the age slope as the same for the study site levels
>
> My FSGD file is designed as follows
>
> Class SITE 1-Control
> Class SITE 1-PATIENT
> Class SITE 2-Control
> Class SITE 2-PATIENT
> Class SITE 3-Control
> Class SITE 3-PATIENT
> Class SITE 4-Control
> Class SITE 4-PATIENT
> Variables age_at_scan
>
> study site levels = 1,2,3 and 4
> diagnosis levels = PATIENT and Control
> age_at_scan = covariate age
>
> Any advice would be greatly appreciated.
>
> Respectfully,
>
> Tim
>
> --
> Timothy Hendrickson
> Department of Psychiatry
> University of Minnesota
> Mobile: 507-259-3434 (texts okay)
>



-- 
Timothy Hendrickson
Department of 

[Freesurfer] Design Matrix Creation

[Timothy Hendrickson]

Freesurfer Support,

I'd like to create a design matrix for a group analysis outside of the DODS
and DOSS models. I understand that in order to do this the -X flag must be
used. However, I have been unable to find examples of how to do this.

I am hoping to reveal a difference in thickness or gyrification amongst a
clinical population. The data set contains two factors: diagnosis, and
study site and one covariate: age. Diagnosis has two levels: controls, and
patients. Study site has four levels, one level for each location the data
has been collected from.

What I would ideally like to do is:

1) Take into account offset differences amongst diagnosis and study site.

2) Allowing a difference in age slope amongst the diagnosis levels

3) Modeling the age slope as the same for the study site levels

My FSGD file is designed as follows

Class SITE 1-Control
Class SITE 1-PATIENT
Class SITE 2-Control
Class SITE 2-PATIENT
Class SITE 3-Control
Class SITE 3-PATIENT
Class SITE 4-Control
Class SITE 4-PATIENT
Variables age_at_scan

study site levels = 1,2,3 and 4
diagnosis levels = PATIENT and Control
age_at_scan = covariate age

Any advice would be greatly appreciated.

Respectfully,

Tim

-- 
Timothy Hendrickson
Department of Psychiatry
University of Minnesota
Mobile: 507-259-3434 (texts okay)
___
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.