Hello, Freesurfer experts:
I found to report subcortical volumetric measure, ie, the volume of amygdale, people tend to use icv as a control variable. Is there any corresponding measure to correct the cortical thickness/area/volume results from autorecon3. Because we have priori hypothesis of hemisphere dichotomy of cortical structures, we need some control variable for each hemisphere. We want to use an overall measure for each hemisphere as a control variable. For example, to measure the group differences on rh_paracentral_thickness, we will use the rh_whole_thickness as a contral variable (which is the sum of the thickness measures of all the ROIs in the right hemisphere produced by Freesurfer). Similarly, to measure the group differences on rh_paracentral_area and rh_paracentral_volume, we will use the rh_whole_area and rh_whole_area as a contral variable respectively. Is this a valid approach to report our results?
Thanks a lot!
Karl
Yes, I think that using average cortical thickness is a reasonable covariate to use in thickness analyses. And total cortical surface area likewise if you are analyzing regional surface areas, although in the case of surface area you need to decide whether to use areas based on the GM or WM surface, or some "mid"-surface in-between.
Use mris_anatomical_stats along with the {lh,rh}.cortex.label files to compute these statistics for the overall cortex. If the optional <surface name> at the end of the command is 'pial', then the area will be based on the pial surface. If nothing is specified, it'll be based on the white (default) surface. If you want the area based on the mid- surface, you either need to construct that surface first (if it isn't already computed by recon-all in the most recent FS versions) or run the command separately for the 'pial' and 'white' surfaces and then average the resulting areas].
cheers, -MH
On Fri, 2010-12-17 at 01:42 +0800, Liukarl wrote:
Hello, Freesurfer experts:
I found to report subcortical volumetric measure, ie, the volume of amygdale, people tend to use icv as a control variable. Is there any corresponding measure to correct the cortical thickness/area/volume results from autorecon3. Because we have priori hypothesis of hemisphere dichotomy of cortical structures, we need some control variable for each hemisphere. We want to use an overall measure for each hemisphere as a control variable. For example, to measure the group differences on rh_paracentral_thickness, we will use the rh_whole_thickness as a contral variable (which is the sum of the thickness measures of all the ROIs in the right hemisphere produced by Freesurfer). Similarly, to measure the group differences on rh_paracentral_area and rh_paracentral_volume, we will use the rh_whole_area and rh_whole_area as a contral variable respectively. Is this a valid approach to report our results?
Thanks a lot!
Karl
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.
BTW: You can't "sum" the thicknesses across the various regions in the aparc files, as that would be a meaningless measure. You either need to compute the average cortical thickness as I suggested, or compute a weighted mean from the data in the aparc files, in which you weight each region's thickness by its proportion of the total surface area.
cheers, -MH
On Thu, 2010-12-16 at 11:57 -0600, Michael Harms wrote:
Yes, I think that using average cortical thickness is a reasonable covariate to use in thickness analyses. And total cortical surface area likewise if you are analyzing regional surface areas, although in the case of surface area you need to decide whether to use areas based on the GM or WM surface, or some "mid"-surface in-between.
Use mris_anatomical_stats along with the {lh,rh}.cortex.label files to compute these statistics for the overall cortex. If the optional <surface name> at the end of the command is 'pial', then the area will be based on the pial surface. If nothing is specified, it'll be based on the white (default) surface. If you want the area based on the mid- surface, you either need to construct that surface first (if it isn't already computed by recon-all in the most recent FS versions) or run the command separately for the 'pial' and 'white' surfaces and then average the resulting areas].
cheers, -MH
On Fri, 2010-12-17 at 01:42 +0800, Liukarl wrote:
Hello, Freesurfer experts:
I found to report subcortical volumetric measure, ie, the volume of amygdale, people tend to use icv as a control variable. Is there any corresponding measure to correct the cortical thickness/area/volume results from autorecon3. Because we have priori hypothesis of hemisphere dichotomy of cortical structures, we need some control variable for each hemisphere. We want to use an overall measure for each hemisphere as a control variable. For example, to measure the group differences on rh_paracentral_thickness, we will use the rh_whole_thickness as a contral variable (which is the sum of the thickness measures of all the ROIs in the right hemisphere produced by Freesurfer). Similarly, to measure the group differences on rh_paracentral_area and rh_paracentral_volume, we will use the rh_whole_area and rh_whole_area as a contral variable respectively. Is this a valid approach to report our results?
Thanks a lot!
Karl
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
Hi, Michael
Thanks for the guidance. Still I am unclear of how to implement your ideas
1 For the mris_anatomical_stats, should I use -a subjid/label/lh.aparc.annot or use the �Ct option to check the thickness file (what��s the name of this file)? If I understand it correctly, I will get a measure of ��average cortical thickness�� of each hemisphere for every subject and further I can set this as a control variable for regional thickness comparison?
2 for each of the cortical regions, there is also a volumetric measure. What average or overall volume measure I can use as a control variable, the ICV or something new from autorecon3?
Thanks
Karl
From: mharms@conte.wustl.edu To: rememberer@hotmail.com Date: Thu, 16 Dec 2010 12:04:35 -0600 CC: freesurfer@nmr.mgh.harvard.edu Subject: Re: [Freesurfer] How to adjust cortical thickness/area/volume measures from autorecon3
BTW: You can't "sum" the thicknesses across the various regions in the aparc files, as that would be a meaningless measure. You either need to compute the average cortical thickness as I suggested, or compute a weighted mean from the data in the aparc files, in which you weight each region's thickness by its proportion of the total surface area.
cheers, -MH
On Thu, 2010-12-16 at 11:57 -0600, Michael Harms wrote:
Yes, I think that using average cortical thickness is a reasonable covariate to use in thickness analyses. And total cortical surface area likewise if you are analyzing regional surface areas, although in the case of surface area you need to decide whether to use areas based on the GM or WM surface, or some "mid"-surface in-between.
Use mris_anatomical_stats along with the {lh,rh}.cortex.label files to compute these statistics for the overall cortex. If the optional <surface name> at the end of the command is 'pial', then the area will be based on the pial surface. If nothing is specified, it'll be based on the white (default) surface. If you want the area based on the mid- surface, you either need to construct that surface first (if it isn't already computed by recon-all in the most recent FS versions) or run the command separately for the 'pial' and 'white' surfaces and then average the resulting areas].
cheers, -MH
On Fri, 2010-12-17 at 01:42 +0800, Liukarl wrote:
Hello, Freesurfer experts:
I found to report subcortical volumetric measure, ie, the volume of amygdale, people tend to use icv as a control variable. Is there any corresponding measure to correct the cortical thickness/area/volume results from autorecon3. Because we have priori hypothesis of hemisphere dichotomy of cortical structures, we need some control variable for each hemisphere. We want to use an overall measure for each hemisphere as a control variable. For example, to measure the group differences on rh_paracentral_thickness, we will use the rh_whole_thickness as a contral variable (which is the sum of the thickness measures of all the ROIs in the right hemisphere produced by Freesurfer). Similarly, to measure the group differences on rh_paracentral_area and rh_paracentral_volume, we will use the rh_whole_area and rh_whole_area as a contral variable respectively. Is this a valid approach to report our results?
Thanks a lot!
Karl
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 mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
You want to use the -l option. Something like: "mris_anatomical_stats -l lh.cortex.label -f cortexStatsFile.txt SUBJECTID lh <pial>" will I think to the trick.
(Use 'pial' at the end if you want the cortical surface area to be that of the pial surface instead of the white).
For volume, you have to chose what measure you want to use as a control variable -- eTIV (aka ICV), one of the "new" variables provided in the aparc.stats files, or one of your own construction. [In the past, I've used a "brain volume" covariate, derived as the volume enclosed within the pial surface minus the volume of the lateral and inferior ventricles (obtained from the aseg.stats)]. The choice depends on exactly what you want to control for (i.e., controlling for ICV and "brain volume" are not the same thing in the case of appreciable ventricular atrophy).
cheers, -MH
On Fri, 2010-12-17 at 05:29 +0800, Liukarl wrote:
Hi, Michael
Thanks for the guidance. Still I am unclear of how to implement your ideas
1 For the mris_anatomical_stats, should I use -a subjid/label/lh.aparc.annot or use the Ct option to check the thickness file (what’s the name of this file)? If I understand it correctly, I will get a measure of “average cortical thickness” of each hemisphere for every subject and further I can set this as a control variable for regional thickness comparison?
2 for each of the cortical regions, there is also a volumetric measure. What average or overall volume measure I can use as a control variable, the ICV or something new from autorecon3?
Thanks
Karl
From: mharms@conte.wustl.edu To: rememberer@hotmail.com Date: Thu, 16 Dec 2010 12:04:35 -0600 CC: freesurfer@nmr.mgh.harvard.edu Subject: Re: [Freesurfer] How to adjust cortical
thickness/area/volume measures from autorecon3
BTW: You can't "sum" the thicknesses across the various regions in
the
aparc files, as that would be a meaningless measure. You either need
to
compute the average cortical thickness as I suggested, or compute a weighted mean from the data in the aparc files, in which you weight
each
region's thickness by its proportion of the total surface area.
cheers, -MH
On Thu, 2010-12-16 at 11:57 -0600, Michael Harms wrote:
Yes, I think that using average cortical thickness is a reasonable covariate to use in thickness analyses. And total cortical surface
area
likewise if you are analyzing regional surface areas, although in
the
case of surface area you need to decide whether to use areas based
on
the GM or WM surface, or some "mid"-surface in-between.
Use mris_anatomical_stats along with the {lh,rh}.cortex.label
files to
compute these statistics for the overall cortex. If the optional <surface name> at the end of the command is 'pial', then the area
will
be based on the pial surface. If nothing is specified, it'll be
based
on the white (default) surface. If you want the area based on the
mid-
surface, you either need to construct that surface first (if it
isn't
already computed by recon-all in the most recent FS versions) or
run the
command separately for the 'pial' and 'white' surfaces and then
average
the resulting areas].
cheers, -MH
On Fri, 2010-12-17 at 01:42 +0800, Liukarl wrote:
&! gt; > Hello, Freesurfer experts:
I found to report subcortical volumetric measure, ie, the volume
of
amygdale, people tend to use icv as a control variable. Is there
any
corresponding measure to correct the cortical
thickness/area/volume
results from autorecon3. Because we have priori hypothesis of hemisphere dichotomy of cortical structures, we need some
control
variable for each hemisphere. We want to use an overall measure
for
each hemisphere as a control variable. For example, to measure
the
group differences on rh_paracentral_thickness, we will use the rh_whole_thickness as a contral variable (which is the sum of
the
thickness measures of all the ROIs in the right hemisphere
produced by
Freesurfer). Similarly, to measure the group differences on rh_paracentral_area and rh_paracentral_volume, we will use the rh_whole_area and rh_whole_area as a contral variable
respectively. Is
this a valid approach to report our results?
Thanks a lot!
Karl
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 &g! t; > > 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
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
-
Liukarl -
Michael Harms