Dear freesurfer experts,I have a question regarding correction of cortical measures or global measures (ICV, total surface area). As I understand from literature and the freesurfer wiki, a model for thickness should not be corrected for ICV (since thickness does not scale with total brain volume), but surface area should be corrected since is does scale with total brain volume. Three questions:1. should I use ICV or should I use total surface area (TSA) as covariate in the model to correct surface area (by the way, average ICV and TSA do not differ between the groups to compare )?2. does lGI also need correction? does it scale with volume? If so, do I need to correct for ICV or for TSA? Usually I see no correction but some do use ICV...3. even if the average of certain measures (like ICV/TSA, or gender, or age) does not differ between groups, should they be added to the model? If I do not correct for gender, for example, my results change and most papers do use age and gender as covariates if they compare groups despite a lack of differences between groups... Thanks,Maaike
On 05/02/2016 07:39 AM, maaike rive wrote:
Dear freesurfer experts,
I have a question regarding correction of cortical measures or global measures (ICV, total surface area). As I understand from literature and the freesurfer wiki, a model for thickness should not be corrected for ICV (since thickness does not scale with total brain volume), but surface area should be corrected since is does scale with total brain volume. Three questions:
- should I use ICV or should I use total surface area (TSA) as
covariate in the model to correct surface area (by the way, average ICV and TSA do not differ between the groups to compare )?
Not sure. They will be different models. ICV should not change over the lifetime whereas total surface area will.
- does lGI also need correction? does it scale with volume? If so, do
I need to correct for ICV or for TSA? Usually I see no correction but some do use ICV...
I don't know, perhaps Marie will weigh in
- even if the average of certain measures (like ICV/TSA, or gender,
or age) does not differ between groups, should they be added to the model? If I do not correct for gender, for example, my results change and most papers do use age and gender as covariates if they compare groups despite a lack of differences between groups...
They can still reduce intersubject variance. doug
Thanks,
Maaike
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
Hi Doug, Thanks for the reply.I'm still struggling with this issue though. 1. For instance, if one corrects surface area for ICV in the GLM, shouldn't the fact that this relationship is non-linear be taken into account? Since volume increases with r^3 and surface with r^2... Same goes for lGI, how is the relationship between lGI en ICV? 2. If I choose to correct for TSA, it seems natural to correct thickness for mean (or total) thickness and lGI for some global measure (is there such thing as mean or total lGI?) as well, or am I mistaken? I don't understand why one measure should be corrected for a global measure with the same units, whereas another should not. 3. ICV is not influenced by the volumes of the different brain structures (it is not a summation of these volumes) so volume correction for ICV is unbiased. However TSA (or mean thickness, or "global lGI measure") is dependent of the area (or thickness or lGI) of each discrete brain structure, right? So for instance if one group has a smaller superior frontal area, total area will be smaller as well. Wouldn't correction remove part of the effect then? Maaike > To: freesurfer@nmr.mgh.harvard.edu; Marie.Schaer@unige.ch> From: greve@nmr.mgh.harvard.edu> Date: Thu, 5 May 2016 12:32:11 -0400> Subject: Re: [Freesurfer] does lGI needs correction for any global measure?> > > > On 05/02/2016 07:39 AM, maaike rive wrote:> > Dear freesurfer experts,> > > >> >> > > > I have a question regarding correction of cortical measures or global > > measures (ICV, total surface area). As I understand from literature > > and the freesurfer wiki, a model for thickness should not be corrected > > for ICV (since thickness does not scale with total brain volume), but > > surface area should be corrected since is does scale with total brain > > volume. Three questions:> >> > > >> >> > > > 1. should I use ICV or should I use total surface area (TSA) as > > covariate in the model to correct surface area (by the way, average > > ICV and TSA do not differ between the groups to compare )?> >> > > Not sure. They will be different models. ICV should not change over the > lifetime whereas total surface area will.> > > > 2. does lGI also need correction? does it scale with volume? If so, do > > I need to correct for ICV or for TSA? Usually I see no correction but > > some do use ICV...> >> > > I don't know, perhaps Marie will weigh in> > > > 3. even if the average of certain measures (like ICV/TSA, or gender, > > or age) does not differ between groups, should they be added to the > > model? If I do not correct for gender, for example, my results change > > and most papers do use age and gender as covariates if they compare > > groups despite a lack of differences between groups...> >> > > They can still reduce intersubject variance.> doug> > > >> >> > > > Thanks,> >> > > >> >> > > > Maaike> >> > > >> >> > _______________________________________________> > Freesurfer mailing list> > Freesurfer@nmr.mgh.harvard.edu> > https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer%3E > -- > Douglas N. Greve, Ph.D.> MGH-NMR Center> greve@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%3E 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%3E > > 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 05/06/2016 04:53 AM, maaike rive wrote:
Hi Doug,
Thanks for the reply.
I'm still struggling with this issue though.
- For instance, if one corrects surface area for ICV in the GLM,
shouldn't the fact that this relationship is non-linear be taken into account? Since volume increases with r^3 and surface with r^2... Same goes for lGI, how is the relationship between lGI en ICV?
Not necessarily. The surface of a sphere will increase as r^2, but we are not talking about a sphere but a *folded* surface. Consider a box of sheets of paper. If you double the volume of the box, you double the number of sheets of paper and double the surface area of those sheets, so surface area is linear with volume. The folded brain is not as efficient as a box of paper, but you'll still get something more than r^2. When I've looked at it before, it appears to be pretty linear.
- If I choose to correct for TSA, it seems natural to correct
thickness for mean (or total) thickness and lGI for some global measure (is there such thing as mean or total lGI?) as well, or am I mistaken? I don't understand why one measure should be corrected for a global measure with the same units, whereas another should not.
I think it just depends on what hypothesis you are trying to test. The normalization changes the hypothesis.
- ICV is not influenced by the volumes of the different brain
structures (it is not a summation of these volumes) so volume correction for ICV is unbiased. However TSA (or mean thickness, or "global lGI measure") is dependent of the area (or thickness or lGI) of each discrete brain structure, right? So for instance if one group has a smaller superior frontal area, total area will be smaller as well. Wouldn't correction remove part of the effect then?
Yes, see my answer to #2
Maaike
To: freesurfer@nmr.mgh.harvard.edu; Marie.Schaer@unige.ch From: greve@nmr.mgh.harvard.edu Date: Thu, 5 May 2016 12:32:11 -0400 Subject: Re: [Freesurfer] does lGI needs correction for any global
measure?
On 05/02/2016 07:39 AM, maaike rive wrote:
Dear freesurfer experts,
I have a question regarding correction of cortical measures or global measures (ICV, total surface area). As I understand from literature and the freesurfer wiki, a model for thickness should not be
corrected
for ICV (since thickness does not scale with total brain volume), but surface area should be corrected since is does scale with total brain volume. Three questions:
- should I use ICV or should I use total surface area (TSA) as
covariate in the model to correct surface area (by the way, average ICV and TSA do not differ between the groups to compare )?
Not sure. They will be different models. ICV should not change over the lifetime whereas total surface area will.
- does lGI also need correction? does it scale with volume? If
so, do
I need to correct for ICV or for TSA? Usually I see no correction but some do use ICV...
I don't know, perhaps Marie will weigh in
- even if the average of certain measures (like ICV/TSA, or gender,
or age) does not differ between groups, should they be added to the model? If I do not correct for gender, for example, my results change and most papers do use age and gender as covariates if they compare groups despite a lack of differences between groups...
They can still reduce intersubject variance. doug
Thanks,
Maaike
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 greve@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.
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
freesurfer@nmr.mgh.harvard.edu
-
Douglas N Greve -
Harms, Michael -
maaike rive