Sorry, for the delay...

1. You can use mri_surfcluster with the --fdr option.

2. I'm not sure I understand. Are you really setting the FDR to .88? Usually it is some thing like .05

doug

On 02/10/2014 03:28 PM, pfotiad@nmr.mgh.harvard.edu wrote:

Hi FS community,

I was wondering if anyone had a chance to give this a look.

Thanks in advance, Panos

...

Hi FS community,

I am doing a cortical thickness comparison between a diseased and a healthy group. After running the mri_glmfit command for both hemispheres, I project the uncorrected significance map and there are a number of significant clusters showing (if you consider the p value to be at most 0.05). However, instead of using the clusterwise Correction for Multiple Comparisons, I wanted to use FDR to see what the results would look like, and I had a few questions regarding the whole process:

1. Are Tksurfer and QDEC the only tools in FS that I can apply the FDR

correction

...

with, or are there any other ways?

2. In Tksurfer, I tried using the "Set Threshold Using FDR" tool, and when

I set the rate from 0.88 to 0.87 it bumps up the minimum threshold from 1.37 to 3.24 hence resulting in no significant clusters for rates less than 0.88. Does that sound normal?

Thanks in advance! Panos _______________________________________________ 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

Hi Doug and Martin,

Thanks both for your replies!

Doug, I would normally set the FDR to something close to .05 but when this value didn't give me any significant clusters for my analysis, I was trying to find which value was the min one that gave significant clusters just out of curiosity, and this value happened to be 0.88. But I wasn't going to actually use this value for my analysis. It just surprised that for such a high fdr threshold, the min -log(p) value increased so much, hence excluding every uncorrected significant cluster for any fdr threshold less than 0.88.

Best, Panos

Hi Panos,

also the LME tools (matlab) have tools for FDR correction:

/lme_mass_FDR/ (Original Benjamini and Hochberg FDR procedure)

/lme_mass_FDR2/ (A more powerful two-stage FDR procedure:

"Adaptive Linear Step-up Procedures that control the False Discovery Rate" Benjamini, Krieger, Yekutieli )

Best, Martin

On 02/10/2014 03:28 PM, pfotiad@nmr.mgh.harvard.edu wrote:

...

Hi FS community,

I was wondering if anyone had a chance to give this a look.

Thanks in advance, Panos

...

Hi FS community,

I am doing a cortical thickness comparison between a diseased and a healthy group. After running the mri_glmfit command for both hemispheres, I project the uncorrected significance map and there are a number of significant clusters showing (if you consider the p value to be at most 0.05). However, instead of using the clusterwise Correction for Multiple Comparisons, I wanted to use FDR to see what the results would look like, and I had a few questions regarding the whole process:

1. Are Tksurfer and QDEC the only tools in FS that I can apply the FDR

correction

...

with, or are there any other ways?

2. In Tksurfer, I tried using the "Set Threshold Using FDR" tool, and

when I set the rate from 0.88 to 0.87 it bumps up the minimum threshold from 1.37 to 3.24 hence resulting in no significant clusters for rates less than 0.88. Does that sound normal?

Thanks in advance! Panos _______________________________________________ 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

-- Martin Reuter, Ph.D. Assistant in Neuroscience - Massachusetts General Hospital Instructor in Neurology - Harvard Medical School MGH / HMS / MIT

A.A.Martinos Center for Biomedical Imaging 149 Thirteenth Street, Suite 2301 Charlestown, MA 02129

Phone: +1-617-724-5652 Email: mreuter@nmr.mgh.harvard.edu reuter@mit.edu Web : http://reuter.mit.edu

I see. Thanks for your input! Panos

This kind of thing can happen, especially when there is not much activity in the map doug

On 02/11/2014 12:02 PM, pfotiad@nmr.mgh.harvard.edu wrote:

...

Hi Doug and Martin,

Thanks both for your replies!

Doug, I would normally set the FDR to something close to .05 but when this value didn't give me any significant clusters for my analysis, I was trying to find which value was the min one that gave significant clusters just out of curiosity, and this value happened to be 0.88. But I wasn't going to actually use this value for my analysis. It just surprised that for such a high fdr threshold, the min -log(p) value increased so much, hence excluding every uncorrected significant cluster for any fdr threshold less than 0.88.

Best, Panos

...

Hi Panos,

also the LME tools (matlab) have tools for FDR correction:

/lme_mass_FDR/ (Original Benjamini and Hochberg FDR procedure)

/lme_mass_FDR2/ (A more powerful two-stage FDR procedure:

"Adaptive Linear Step-up Procedures that control the False Discovery Rate" Benjamini, Krieger, Yekutieli )

Best, Martin

On 02/10/2014 03:28 PM, pfotiad@nmr.mgh.harvard.edu wrote:

...

Hi FS community,

I was wondering if anyone had a chance to give this a look.

Thanks in advance, Panos

...

Hi FS community,

I am doing a cortical thickness comparison between a diseased and a healthy group. After running the mri_glmfit command for both hemispheres, I project the uncorrected significance map and there are a number of significant clusters showing (if you consider the p value to be at most 0.05). However, instead of using the clusterwise Correction for Multiple Comparisons, I wanted to use FDR to see what the results would look like, and I had a few questions regarding the whole process:

1. Are Tksurfer and QDEC the only tools in FS that I can apply the

FDR

correction

...

with, or are there any other ways?

2. In Tksurfer, I tried using the "Set Threshold Using FDR" tool, and

when I set the rate from 0.88 to 0.87 it bumps up the minimum threshold from 1.37 to 3.24 hence resulting in no significant clusters for rates less than 0.88. Does that sound normal?

Thanks in advance! Panos _______________________________________________ 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

-- Martin Reuter, Ph.D. Assistant in Neuroscience - Massachusetts General Hospital Instructor in Neurology - Harvard Medical School MGH / HMS / MIT

A.A.Martinos Center for Biomedical Imaging 149 Thirteenth Street, Suite 2301 Charlestown, MA 02129

Phone: +1-617-724-5652 Email: mreuter@nmr.mgh.harvard.edu reuter@mit.edu Web : http://reuter.mit.edu

---

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