Hi there!
I was hoping you could help me with the following 'problem' I am experiencing.
Short background: I compared volumetric brain characteristics as output from FreeSurfer parcellations and segmentations between 3 groups, by extracting the volumes per anatomical label from FreeSurfer and analyzing these in SPSS. In SPSS I did a mixed-model analysis, and performed subsequent Benjamini-Hochberg FDR correction for multiple comparisons.
At the moment, I want to make brain maps (because this is a nice way to visualize the data and a reviewer suggested it). However... I believe I end up with only being able to use the uncorrected maps? When I perform an FDR-correction (or MonteCarlo, or clusterwise) in FreeSurfer all the group differences disappear. Unfortunately there is nobody that I can consult on this matter in my lab, thus I really hope someone here can help me? Personally, I think that the difference in the presence (SPSS) versus absence (FS) of groups effects is due to the correction method. I believe that in FreeSurfer I apply a whole brain multiple comparison correction, while in SPSS I merely apply a ROI-based FDR-correction I in SPSS (because I look at volume per label).
However, I am not sure if this is indeed (theoretically) sound? Could anyone advise me? Is there a way I could visualise my own (SPSS) results? If more info is needed please do say so, and I will provide it!
Thank you very much for your time and reply! Best, Siri
Dear Freesurfer experts,
I haven't had the pleasure of receiving a response (yet), and was wondering whether that is because more information is needed?
In addition to the general theoretical question, the second part of the question is as important to me: Is there is any way, known to you experts, for me to visualize my SPSS results in a statistical-map-way?
If anyone could spare some time to help me out I would greatly appreciate it!! Best, Siri
From: freesurfer-bounces@nmr.mgh.harvard.edu [mailto:freesurfer-bounces@nmr.mgh.harvard.edu] On Behalf Of Noordermeer, S.D.S. Sent: zaterdag 13 mei 2017 22:22 To: freesurfer@nmr.mgh.harvard.edu Subject: [Freesurfer] Question - multiple comparisons in FS versus SPSS
Hi there!
I was hoping you could help me with the following 'problem' I am experiencing.
Short background: I compared volumetric brain characteristics as output from FreeSurfer parcellations and segmentations between 3 groups, by extracting the volumes per anatomical label from FreeSurfer and analyzing these in SPSS. In SPSS I did a mixed-model analysis, and performed subsequent Benjamini-Hochberg FDR correction for multiple comparisons.
At the moment, I want to make brain maps (because this is a nice way to visualize the data and a reviewer suggested it). However... I believe I end up with only being able to use the uncorrected maps? When I perform an FDR-correction (or MonteCarlo, or clusterwise) in FreeSurfer all the group differences disappear. Unfortunately there is nobody that I can consult on this matter in my lab, thus I really hope someone here can help me? Personally, I think that the difference in the presence (SPSS) versus absence (FS) of groups effects is due to the correction method. I believe that in FreeSurfer I apply a whole brain multiple comparison correction, while in SPSS I merely apply a ROI-based FDR-correction I in SPSS (because I look at volume per label).
However, I am not sure if this is indeed (theoretically) sound? Could anyone advise me? Is there a way I could visualise my own (SPSS) results? If more info is needed please do say so, and I will provide it!
Thank you very much for your time and reply! Best, Siri
Hi Siri,
Are you trying to overlay the p value associated with each region on the fsaverage surface? This may help:
http://www.mail-archive.com/freesurfer@nmr.mgh.harvard.edu/msg50465.html
Best Wishes, Elijah
Dr. Elijah Mak, Research Associate Department of Psychiatry, Old Age Psychiatry Group University of Cambridge Trinity College, CB21TQ, UK http://www.neuroscience.cam.ac.uk/directory/profile.php?fkm24
On 23 May 2017 at 10:46:38, Noordermeer, S.D.S. (sds.noordermeer@vu.nl) wrote:
Dear Freesurfer experts,
I haven’t had the pleasure of receiving a response (yet), and was wondering whether that is because more information is needed?
In addition to the general theoretical question, the second part of the question is as important to me: Is there is any way, known to you experts, for me to visualize my SPSS results in a statistical-map-way?
If anyone could spare some time to help me out I would greatly appreciate it!!
Best,
Siri
From: freesurfer-bounces@nmr.mgh.harvard.edu [mailto:freesurfer-bounces@nmr.mgh.harvard.edu] On Behalf Of Noordermeer, S.D.S. Sent: zaterdag 13 mei 2017 22:22 To: freesurfer@nmr.mgh.harvard.edu Subject: [Freesurfer] Question - multiple comparisons in FS versus SPSS
Hi there!
I was hoping you could help me with the following 'problem' I am experiencing.
Short background:
I compared volumetric brain characteristics as output from FreeSurfer parcellations and segmentations between 3 groups, by extracting the volumes per anatomical label from FreeSurfer and analyzing these in SPSS. In SPSS I did a mixed-model analysis, and performed subsequent Benjamini-Hochberg FDR correction for multiple comparisons.
At the moment, I want to make brain maps (because this is a nice way to visualize the data and a reviewer suggested it). However... I believe I end up with only being able to use the uncorrected maps?
When I perform an FDR-correction (or MonteCarlo, or clusterwise) in FreeSurfer all the group differences disappear. Unfortunately there is nobody that I can consult on this matter in my lab, thus I really hope someone here can help me? Personally, I think that the difference in the presence (SPSS) versus absence (FS) of groups effects is due to the correction method. I believe that in FreeSurfer I apply a whole brain multiple comparison correction, while in SPSS I merely apply a ROI-based FDR-correction I in SPSS (because I look at volume per label).
However, I am not sure if this is indeed (theoretically) sound? Could anyone advise me? Is there a way I could visualise my own (SPSS) results?
If more info is needed please do say so, and I will provide it!
Thank you very much for your time and reply!
Best,
Siri
_______________________________________________ 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.
Dear Dr. Mak,
Thank you for your time and response! I looked into the link you suggested. Unfortunately, it’s not exactly what I need, although I realize my question may have been stated incorrectly.
Part of the rationale behind creating the map(s) is to be able to see if the findings we report on regional-level are highly focal within that region or rather more widely distributed / traversing regions. When overlaying the p-value over the entire region, that would indeed help me in terms of a visualization, but not in terms of useful information. Sorry for that misunderstanding.
Would you (or anybody else) know if that is possible? Or is the only solution what I’ve done: re-analyze in Freesurfer and study the uncorrected images? And maybe not even use those images in the manuscript, but merely state that the findings seem distributed? Since no findings remain after FDR correction in Freesurfer, which indicates that per voxel small differences are present which add up for the region to a larger (significant) difference?
Thank you again! Best, Siri
From: Elijah [mailto:elijahmak@gmail.com] On Behalf Of Elijah Mak Sent: dinsdag 23 mei 2017 11:52 To: Freesurfer support list; Noordermeer, S.D.S. Subject: Re: [Freesurfer] Question - multiple comparisons in FS versus SPSS
Hi Siri,
Are you trying to overlay the p value associated with each region on the fsaverage surface? This may help:
http://www.mail-archive.com/freesurfer@nmr.mgh.harvard.edu/msg50465.html
Best Wishes, Elijah
Dr. Elijah Mak, Research Associate Department of Psychiatry, Old Age Psychiatry Group University of Cambridge Trinity College, CB21TQ, UK http://www.neuroscience.cam.ac.uk/directory/profile.php?fkm24
On 23 May 2017 at 10:46:38, Noordermeer, S.D.S. (sds.noordermeer@vu.nlmailto:sds.noordermeer@vu.nl) wrote: Dear Freesurfer experts,
I haven’t had the pleasure of receiving a response (yet), and was wondering whether that is because more information is needed?
In addition to the general theoretical question, the second part of the question is as important to me: Is there is any way, known to you experts, for me to visualize my SPSS results in a statistical-map-way?
If anyone could spare some time to help me out I would greatly appreciate it!! Best, Siri
From: freesurfer-bounces@nmr.mgh.harvard.edumailto:freesurfer-bounces@nmr.mgh.harvard.edu [mailto:freesurfer-bounces@nmr.mgh.harvard.edu] On Behalf Of Noordermeer, S.D.S. Sent: zaterdag 13 mei 2017 22:22 To: freesurfer@nmr.mgh.harvard.edumailto:freesurfer@nmr.mgh.harvard.edu Subject: [Freesurfer] Question - multiple comparisons in FS versus SPSS
Hi there!
I was hoping you could help me with the following 'problem' I am experiencing.
Short background: I compared volumetric brain characteristics as output from FreeSurfer parcellations and segmentations between 3 groups, by extracting the volumes per anatomical label from FreeSurfer and analyzing these in SPSS. In SPSS I did a mixed-model analysis, and performed subsequent Benjamini-Hochberg FDR correction for multiple comparisons.
At the moment, I want to make brain maps (because this is a nice way to visualize the data and a reviewer suggested it). However... I believe I end up with only being able to use the uncorrected maps? When I perform an FDR-correction (or MonteCarlo, or clusterwise) in FreeSurfer all the group differences disappear. Unfortunately there is nobody that I can consult on this matter in my lab, thus I really hope someone here can help me? Personally, I think that the difference in the presence (SPSS) versus absence (FS) of groups effects is due to the correction method. I believe that in FreeSurfer I apply a whole brain multiple comparison correction, while in SPSS I merely apply a ROI-based FDR-correction I in SPSS (because I look at volume per label).
However, I am not sure if this is indeed (theoretically) sound? Could anyone advise me? Is there a way I could visualise my own (SPSS) results? If more info is needed please do say so, and I will provide it!
Thank you very much for your time and reply! Best, Siri
_______________________________________________ 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.
The ROI vs voxel-wise differences can come from a lot of places. When you do an ROI analysis, you get a lot of averaging so the noise can be expected to be much less than voxel-wise. If your effect is largely constrained to an ROI, then the ROI result will be more powerful. But they are really very different analyses with very different power.
On 05/23/2017 11:43 AM, Noordermeer, S.D.S. wrote:
Dear Dr. Mak,
Thank you for your time and response! I looked into the link you suggested. Unfortunately, it’s not exactly what I need, although I realize my question may have been stated incorrectly.
Part of the rationale behind creating the map(s) is to be able to see if the findings we report on regional-level are highly focal within that region or rather more widely distributed / traversing regions. When overlaying the p-value over the entire region, that would indeed help me in terms of a visualization, but not in terms of useful information. Sorry for that misunderstanding.
Would you (or anybody else) know if that is possible?
Or is the only solution what I’ve done: re-analyze in Freesurfer and study the uncorrected images?
And maybe not even use those images in the manuscript, but merely state that the findings seem distributed? Since no findings remain after FDR correction in Freesurfer, which indicates that per voxel small differences are present which add up for the region to a larger (significant) difference?
Thank you again!
Best,
Siri
*From:*Elijah [mailto:elijahmak@gmail.com] *On Behalf Of *Elijah Mak *Sent:* dinsdag 23 mei 2017 11:52 *To:* Freesurfer support list; Noordermeer, S.D.S. *Subject:* Re: [Freesurfer] Question - multiple comparisons in FS versus SPSS
Hi Siri,
Are you trying to overlay the p value associated with each region on the fsaverage surface? This may help:
http://www.mail-archive.com/freesurfer@nmr.mgh.harvard.edu/msg50465.html
Best Wishes,
Elijah
*Dr. Elijah Mak, Research Associate*
Department of Psychiatry, Old Age Psychiatry Group
University of Cambridge Trinity College, CB21TQ, UK
http://www.neuroscience.cam.ac.uk/directory/profile.php?fkm24
On 23 May 2017 at 10:46:38, Noordermeer, S.D.S. (sds.noordermeer@vu.nl mailto:sds.noordermeer@vu.nl) wrote:
Dear Freesurfer experts, I haven’t had the pleasure of receiving a response (yet), and was wondering whether that is because more information is needed? In addition to the general theoretical question, the second part of the question is as important to me: Is there is any way, known to you experts, for me to visualize my SPSS results in a statistical-map-way? If anyone could spare some time to help me out I would greatly appreciate it!! Best, Siri *From:*freesurfer-bounces@nmr.mgh.harvard.edu <mailto:freesurfer-bounces@nmr.mgh.harvard.edu> [mailto:freesurfer-bounces@nmr.mgh.harvard.edu] *On Behalf Of *Noordermeer, S.D.S. *Sent:* zaterdag 13 mei 2017 22:22 *To:* freesurfer@nmr.mgh.harvard.edu <mailto:freesurfer@nmr.mgh.harvard.edu> *Subject:* [Freesurfer] Question - multiple comparisons in FS versus SPSS Hi there! I was hoping you could help me with the following 'problem' I am experiencing. Short background: I compared volumetric brain characteristics as output from FreeSurfer parcellations and segmentations between 3 groups, by extracting the volumes per anatomical label from FreeSurfer and analyzing these in SPSS. In SPSS I did a mixed-model analysis, and performed subsequent Benjamini-Hochberg FDR correction for multiple comparisons. At the moment, I want to make brain maps (because this is a nice way to visualize the data and a reviewer suggested it). However... I believe I end up with only being able to use the uncorrected maps? When I perform an FDR-correction (or MonteCarlo, or clusterwise) in FreeSurfer all the group differences disappear. Unfortunately there is nobody that I can consult on this matter in my lab, thus I really hope someone here can help me? Personally, I think that the difference in the presence (SPSS) versus absence (FS) of groups effects is due to the correction method. I believe that in FreeSurfer I apply a whole brain multiple comparison correction, while in SPSS I merely apply a ROI-based FDR-correction I in SPSS (because I look at volume per label). However, I am not sure if this is indeed (theoretically) sound? Could anyone advise me? Is there a way I could visualise my own (SPSS) results? If more info is needed please do say so, and I will provide it! Thank you very much for your time and reply! Best, Siri _______________________________________________ Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu <mailto: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
Thank you for your answer! I already thought so, but it is good to hear it from the expert.
Would it be correct to state (as the reviewer persistently asks for it) that it may be somewhat informative to additionally do the voxel-wise analyses, but that the analyses are very different resulting in non-matching findings? And that the uncorrected findings are the best (or rather the least worse) comparison with the ROI (spss) analyses? So when the uncorrected voxel-wise maps from Freesurfer largely correspond (i.e. missing 1 out of 11 regions) with the ROI findings, this could (maybe) be seen as an indication for robust findings? And the absence of findings after correction may indicate distributed results, but that both suggestions are very tentative because of the different approaches?
Or would you rather say, 'Let's not do that, it's too different and therefore quite useless'?
With great gratefulness, Best, Siri
________________________________________ Van: freesurfer-bounces@nmr.mgh.harvard.edu [freesurfer-bounces@nmr.mgh.harvard.edu] namens Douglas N Greve [greve@nmr.mgh.harvard.edu] Verzonden: woensdag 24 mei 2017 0:05 Aan: freesurfer@nmr.mgh.harvard.edu Onderwerp: Re: [Freesurfer] Question - multiple comparisons in FS versus SPSS
The ROI vs voxel-wise differences can come from a lot of places. When you do an ROI analysis, you get a lot of averaging so the noise can be expected to be much less than voxel-wise. If your effect is largely constrained to an ROI, then the ROI result will be more powerful. But they are really very different analyses with very different power.
On 05/23/2017 11:43 AM, Noordermeer, S.D.S. wrote:
Dear Dr. Mak,
Thank you for your time and response! I looked into the link you suggested. Unfortunately, it’s not exactly what I need, although I realize my question may have been stated incorrectly.
Part of the rationale behind creating the map(s) is to be able to see if the findings we report on regional-level are highly focal within that region or rather more widely distributed / traversing regions. When overlaying the p-value over the entire region, that would indeed help me in terms of a visualization, but not in terms of useful information. Sorry for that misunderstanding.
Would you (or anybody else) know if that is possible?
Or is the only solution what I’ve done: re-analyze in Freesurfer and study the uncorrected images?
And maybe not even use those images in the manuscript, but merely state that the findings seem distributed? Since no findings remain after FDR correction in Freesurfer, which indicates that per voxel small differences are present which add up for the region to a larger (significant) difference?
Thank you again!
Best,
Siri
*From:*Elijah [mailto:elijahmak@gmail.com] *On Behalf Of *Elijah Mak *Sent:* dinsdag 23 mei 2017 11:52 *To:* Freesurfer support list; Noordermeer, S.D.S. *Subject:* Re: [Freesurfer] Question - multiple comparisons in FS versus SPSS
Hi Siri,
Are you trying to overlay the p value associated with each region on the fsaverage surface? This may help:
http://www.mail-archive.com/freesurfer@nmr.mgh.harvard.edu/msg50465.html
Best Wishes,
Elijah
*Dr. Elijah Mak, Research Associate*
Department of Psychiatry, Old Age Psychiatry Group
University of Cambridge Trinity College, CB21TQ, UK
http://www.neuroscience.cam.ac.uk/directory/profile.php?fkm24
On 23 May 2017 at 10:46:38, Noordermeer, S.D.S. (sds.noordermeer@vu.nl mailto:sds.noordermeer@vu.nl) wrote:
Dear Freesurfer experts, I haven’t had the pleasure of receiving a response (yet), and was wondering whether that is because more information is needed? In addition to the general theoretical question, the second part of the question is as important to me: Is there is any way, known to you experts, for me to visualize my SPSS results in a statistical-map-way? If anyone could spare some time to help me out I would greatly appreciate it!! Best, Siri *From:*freesurfer-bounces@nmr.mgh.harvard.edu <mailto:freesurfer-bounces@nmr.mgh.harvard.edu> [mailto:freesurfer-bounces@nmr.mgh.harvard.edu] *On Behalf Of *Noordermeer, S.D.S. *Sent:* zaterdag 13 mei 2017 22:22 *To:* freesurfer@nmr.mgh.harvard.edu <mailto:freesurfer@nmr.mgh.harvard.edu> *Subject:* [Freesurfer] Question - multiple comparisons in FS versus SPSS Hi there! I was hoping you could help me with the following 'problem' I am experiencing. Short background: I compared volumetric brain characteristics as output from FreeSurfer parcellations and segmentations between 3 groups, by extracting the volumes per anatomical label from FreeSurfer and analyzing these in SPSS. In SPSS I did a mixed-model analysis, and performed subsequent Benjamini-Hochberg FDR correction for multiple comparisons. At the moment, I want to make brain maps (because this is a nice way to visualize the data and a reviewer suggested it). However... I believe I end up with only being able to use the uncorrected maps? When I perform an FDR-correction (or MonteCarlo, or clusterwise) in FreeSurfer all the group differences disappear. Unfortunately there is nobody that I can consult on this matter in my lab, thus I really hope someone here can help me? Personally, I think that the difference in the presence (SPSS) versus absence (FS) of groups effects is due to the correction method. I believe that in FreeSurfer I apply a whole brain multiple comparison correction, while in SPSS I merely apply a ROI-based FDR-correction I in SPSS (because I look at volume per label). However, I am not sure if this is indeed (theoretically) sound? Could anyone advise me? Is there a way I could visualise my own (SPSS) results? If more info is needed please do say so, and I will provide it! Thank you very much for your time and reply! Best, Siri _______________________________________________ Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu <mailto: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
-- 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
I think you're first statement is accurate.
On 5/24/17 1:45 AM, Noordermeer, S.D.S. wrote:
Thank you for your answer! I already thought so, but it is good to hear it from the expert.
Would it be correct to state (as the reviewer persistently asks for it) that it may be somewhat informative to additionally do the voxel-wise analyses, but that the analyses are very different resulting in non-matching findings? And that the uncorrected findings are the best (or rather the least worse) comparison with the ROI (spss) analyses? So when the uncorrected voxel-wise maps from Freesurfer largely correspond (i.e. missing 1 out of 11 regions) with the ROI findings, this could (maybe) be seen as an indication for robust findings? And the absence of findings after correction may indicate distributed results, but that both suggestions are very tentative because of the different approaches?
Or would you rather say, 'Let's not do that, it's too different and therefore quite useless'?
With great gratefulness, Best, Siri
Van: freesurfer-bounces@nmr.mgh.harvard.edu [freesurfer-bounces@nmr.mgh.harvard.edu] namens Douglas N Greve [greve@nmr.mgh.harvard.edu] Verzonden: woensdag 24 mei 2017 0:05 Aan: freesurfer@nmr.mgh.harvard.edu Onderwerp: Re: [Freesurfer] Question - multiple comparisons in FS versus SPSS
The ROI vs voxel-wise differences can come from a lot of places. When you do an ROI analysis, you get a lot of averaging so the noise can be expected to be much less than voxel-wise. If your effect is largely constrained to an ROI, then the ROI result will be more powerful. But they are really very different analyses with very different power.
On 05/23/2017 11:43 AM, Noordermeer, S.D.S. wrote:
Dear Dr. Mak,
Thank you for your time and response! I looked into the link you suggested. Unfortunately, it’s not exactly what I need, although I realize my question may have been stated incorrectly.
Part of the rationale behind creating the map(s) is to be able to see if the findings we report on regional-level are highly focal within that region or rather more widely distributed / traversing regions. When overlaying the p-value over the entire region, that would indeed help me in terms of a visualization, but not in terms of useful information. Sorry for that misunderstanding.
Would you (or anybody else) know if that is possible?
Or is the only solution what I’ve done: re-analyze in Freesurfer and study the uncorrected images?
And maybe not even use those images in the manuscript, but merely state that the findings seem distributed? Since no findings remain after FDR correction in Freesurfer, which indicates that per voxel small differences are present which add up for the region to a larger (significant) difference?
Thank you again!
Best,
Siri
*From:*Elijah [mailto:elijahmak@gmail.com] *On Behalf Of *Elijah Mak *Sent:* dinsdag 23 mei 2017 11:52 *To:* Freesurfer support list; Noordermeer, S.D.S. *Subject:* Re: [Freesurfer] Question - multiple comparisons in FS versus SPSS
Hi Siri,
Are you trying to overlay the p value associated with each region on the fsaverage surface? This may help:
http://www.mail-archive.com/freesurfer@nmr.mgh.harvard.edu/msg50465.html
Best Wishes,
Elijah
*Dr. Elijah Mak, Research Associate*
Department of Psychiatry, Old Age Psychiatry Group
University of Cambridge Trinity College, CB21TQ, UK
http://www.neuroscience.cam.ac.uk/directory/profile.php?fkm24
On 23 May 2017 at 10:46:38, Noordermeer, S.D.S. (sds.noordermeer@vu.nl mailto:sds.noordermeer@vu.nl) wrote:
Dear Freesurfer experts, I haven’t had the pleasure of receiving a response (yet), and was wondering whether that is because more information is needed? In addition to the general theoretical question, the second part of the question is as important to me: Is there is any way, known to you experts, for me to visualize my SPSS results in a statistical-map-way? If anyone could spare some time to help me out I would greatly appreciate it!! Best, Siri *From:*freesurfer-bounces@nmr.mgh.harvard.edu <mailto:freesurfer-bounces@nmr.mgh.harvard.edu> [mailto:freesurfer-bounces@nmr.mgh.harvard.edu] *On Behalf Of *Noordermeer, S.D.S. *Sent:* zaterdag 13 mei 2017 22:22 *To:* freesurfer@nmr.mgh.harvard.edu <mailto:freesurfer@nmr.mgh.harvard.edu> *Subject:* [Freesurfer] Question - multiple comparisons in FS versus SPSS Hi there! I was hoping you could help me with the following 'problem' I am experiencing. Short background: I compared volumetric brain characteristics as output from FreeSurfer parcellations and segmentations between 3 groups, by extracting the volumes per anatomical label from FreeSurfer and analyzing these in SPSS. In SPSS I did a mixed-model analysis, and performed subsequent Benjamini-Hochberg FDR correction for multiple comparisons. At the moment, I want to make brain maps (because this is a nice way to visualize the data and a reviewer suggested it). However... I believe I end up with only being able to use the uncorrected maps? When I perform an FDR-correction (or MonteCarlo, or clusterwise) in FreeSurfer all the group differences disappear. Unfortunately there is nobody that I can consult on this matter in my lab, thus I really hope someone here can help me? Personally, I think that the difference in the presence (SPSS) versus absence (FS) of groups effects is due to the correction method. I believe that in FreeSurfer I apply a whole brain multiple comparison correction, while in SPSS I merely apply a ROI-based FDR-correction I in SPSS (because I look at volume per label). However, I am not sure if this is indeed (theoretically) sound? Could anyone advise me? Is there a way I could visualise my own (SPSS) results? If more info is needed please do say so, and I will provide it! Thank you very much for your time and reply! Best, Siri _______________________________________________ Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu <mailto: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
-- 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
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
Thanks again!! I will write it down like that and will use the uncorrected maps as possible supplementary material.
Best, Siri
-----Original Message----- From: freesurfer-bounces@nmr.mgh.harvard.edu [mailto:freesurfer-bounces@nmr.mgh.harvard.edu] On Behalf Of Douglas Greve Sent: woensdag 24 mei 2017 15:30 To: freesurfer@nmr.mgh.harvard.edu Subject: Re: [Freesurfer] Question - multiple comparisons in FS versus SPSS
I think you're first statement is accurate.
On 5/24/17 1:45 AM, Noordermeer, S.D.S. wrote:
Thank you for your answer! I already thought so, but it is good to hear it from the expert.
Would it be correct to state (as the reviewer persistently asks for it) that it may be somewhat informative to additionally do the voxel-wise analyses, but that the analyses are very different resulting in non-matching findings? And that the uncorrected findings are the best (or rather the least worse) comparison with the ROI (spss) analyses? So when the uncorrected voxel-wise maps from Freesurfer largely correspond (i.e. missing 1 out of 11 regions) with the ROI findings, this could (maybe) be seen as an indication for robust findings? And the absence of findings after correction may indicate distributed results, but that both suggestions are very tentative because of the different approaches?
Or would you rather say, 'Let's not do that, it's too different and therefore quite useless'?
With great gratefulness, Best, Siri
Van: freesurfer-bounces@nmr.mgh.harvard.edu [freesurfer-bounces@nmr.mgh.harvard.edu] namens Douglas N Greve [greve@nmr.mgh.harvard.edu] Verzonden: woensdag 24 mei 2017 0:05 Aan: freesurfer@nmr.mgh.harvard.edu Onderwerp: Re: [Freesurfer] Question - multiple comparisons in FS versus SPSS
The ROI vs voxel-wise differences can come from a lot of places. When you do an ROI analysis, you get a lot of averaging so the noise can be expected to be much less than voxel-wise. If your effect is largely constrained to an ROI, then the ROI result will be more powerful. But they are really very different analyses with very different power.
On 05/23/2017 11:43 AM, Noordermeer, S.D.S. wrote:
Dear Dr. Mak,
Thank you for your time and response! I looked into the link you suggested. Unfortunately, it's not exactly what I need, although I realize my question may have been stated incorrectly.
Part of the rationale behind creating the map(s) is to be able to see if the findings we report on regional-level are highly focal within that region or rather more widely distributed / traversing regions. When overlaying the p-value over the entire region, that would indeed help me in terms of a visualization, but not in terms of useful information. Sorry for that misunderstanding.
Would you (or anybody else) know if that is possible?
Or is the only solution what I've done: re-analyze in Freesurfer and study the uncorrected images?
And maybe not even use those images in the manuscript, but merely state that the findings seem distributed? Since no findings remain after FDR correction in Freesurfer, which indicates that per voxel small differences are present which add up for the region to a larger (significant) difference?
Thank you again!
Best,
Siri
*From:*Elijah [mailto:elijahmak@gmail.com] *On Behalf Of *Elijah Mak *Sent:* dinsdag 23 mei 2017 11:52 *To:* Freesurfer support list; Noordermeer, S.D.S. *Subject:* Re: [Freesurfer] Question - multiple comparisons in FS versus SPSS
Hi Siri,
Are you trying to overlay the p value associated with each region on the fsaverage surface? This may help:
http://www.mail-archive.com/freesurfer@nmr.mgh.harvard.edu/msg50465.h tml
Best Wishes,
Elijah
*Dr. Elijah Mak, Research Associate*
Department of Psychiatry, Old Age Psychiatry Group
University of Cambridge Trinity College, CB21TQ, UK
http://www.neuroscience.cam.ac.uk/directory/profile.php?fkm24
On 23 May 2017 at 10:46:38, Noordermeer, S.D.S. (sds.noordermeer@vu.nl mailto:sds.noordermeer@vu.nl) wrote:
Dear Freesurfer experts, I haven't had the pleasure of receiving a response (yet), and was wondering whether that is because more information is needed? In addition to the general theoretical question, the second part of the question is as important to me: Is there is any way, known to you experts, for me to visualize my SPSS results in a statistical-map-way? If anyone could spare some time to help me out I would greatly appreciate it!! Best, Siri *From:*freesurfer-bounces@nmr.mgh.harvard.edu <mailto:freesurfer-bounces@nmr.mgh.harvard.edu> [mailto:freesurfer-bounces@nmr.mgh.harvard.edu] *On Behalf Of *Noordermeer, S.D.S. *Sent:* zaterdag 13 mei 2017 22:22 *To:* freesurfer@nmr.mgh.harvard.edu <mailto:freesurfer@nmr.mgh.harvard.edu> *Subject:* [Freesurfer] Question - multiple comparisons in FS versus SPSS Hi there! I was hoping you could help me with the following 'problem' I am experiencing. Short background: I compared volumetric brain characteristics as output from FreeSurfer parcellations and segmentations between 3 groups, by extracting the volumes per anatomical label from FreeSurfer and analyzing these in SPSS. In SPSS I did a mixed-model analysis, and performed subsequent Benjamini-Hochberg FDR correction for multiple comparisons. At the moment, I want to make brain maps (because this is a nice way to visualize the data and a reviewer suggested it). However... I believe I end up with only being able to use the uncorrected maps? When I perform an FDR-correction (or MonteCarlo, or clusterwise) in FreeSurfer all the group differences disappear. Unfortunately there is nobody that I can consult on this matter in my lab, thus I really hope someone here can help me? Personally, I think that the difference in the presence (SPSS) versus absence (FS) of groups effects is due to the correction method. I believe that in FreeSurfer I apply a whole brain multiple comparison correction, while in SPSS I merely apply a ROI-based FDR-correction I in SPSS (because I look at volume per label). However, I am not sure if this is indeed (theoretically) sound? Could anyone advise me? Is there a way I could visualise my own (SPSS) results? If more info is needed please do say so, and I will provide it! Thank you very much for your time and reply! Best, Siri _______________________________________________ Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu <mailto: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
-- 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
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@nmr.mgh.harvard.edu
-
Douglas Greve -
Douglas N Greve
-
Elijah Mak -
Noordermeer, S.D.S.