They indicate the number of voxels needed for a cluster to be significant in the current map. They are determined based on the clusters you have in the data. If you had smaller clusters, then they could also be significant, but there is no exhaustive search to determine the exact threshold. For example, if your smallest cluster is 500 voxels and the corrected p-value is 0.0001. Then likely a cluster that is smaller would also be significant, but there is no smaller cluster to be significant, so SPM reports 500 as the cluster extent required.


Best Regards, Donald McLaren
=================
D.G. McLaren, Ph.D.
Research Fellow, Department of Neurology, Massachusetts General Hospital and
Harvard Medical School
Postdoctoral Research Fellow, GRECC, Bedford VA
Website: http://www.martinos.org/~mclaren
Office: (773) 406-2464
=====================
This e-mail contains CONFIDENTIAL INFORMATION which may contain PROTECTED
HEALTHCARE INFORMATION and may also be LEGALLY PRIVILEGED and which is
intended only for the use of the individual or entity named above. If the
reader of the e-mail is not the intended recipient or the employee or agent
responsible for delivering it to the intended recipient, you are hereby
notified that you are in possession of confidential and privileged
information. Any unauthorized use, disclosure, copying or the taking of any
action in reliance on the contents of this information is strictly
prohibited and may be unlawful. If you have received this e-mail
unintentionally, please immediately notify the sender via telephone at (773)
406-2464 or email.


On Fri, Jun 20, 2014 at 4:14 PM, Victoria Klimaj <[log in to unmask]> wrote:
Donald,

Do the numbers corresponding to cFWE and cFDR indicate voxel threshold, or cubic millimeter threshold? Mine are showing up as FWEc=629 and FDRc=103, which is much larger than I would expect a voxel threshold to be for my scans. 

Thanks!
-Victoria


On Fri, Jun 20, 2014 at 1:26 PM, MCLAREN, Donald <[log in to unmask]> wrote:
You can use SPM to determine the cFWE or cFDR extent thresholds. They are at the bottom of the results window.

You can also use 3dClustSim in AFNI to do a monte carlo simulation to determine the extent criteria.

Best Regards, Donald McLaren
=================
D.G. McLaren, Ph.D.
Research Fellow, Department of Neurology, Massachusetts General Hospital and
Harvard Medical School
Postdoctoral Research Fellow, GRECC, Bedford VA
Website: http://www.martinos.org/~mclaren
Office: (773) 406-2464
=====================
This e-mail contains CONFIDENTIAL INFORMATION which may contain PROTECTED
HEALTHCARE INFORMATION and may also be LEGALLY PRIVILEGED and which is
intended only for the use of the individual or entity named above. If the
reader of the e-mail is not the intended recipient or the employee or agent
responsible for delivering it to the intended recipient, you are hereby
notified that you are in possession of confidential and privileged
information. Any unauthorized use, disclosure, copying or the taking of any
action in reliance on the contents of this information is strictly
prohibited and may be unlawful. If you have received this e-mail
unintentionally, please immediately notify the sender via telephone at (773)
406-2464 or email.


On Fri, Jun 20, 2014 at 2:17 PM, Victoria Klimaj <[log in to unmask]> wrote:
Donald,

Thank you for the explanation! That is very helpful.

However, if we don't end up exploring TFCE in FSL, what cluster threshold would you advise for SPM given a voxel size of 3mm and smoothing kernel of 8mm?

-Victoria



On Fri, Jun 20, 2014 at 8:05 AM, MCLAREN, Donald <[log in to unmask]> wrote:
Please see inline responses below.


On Wed, Jun 18, 2014 at 6:46 PM, Victoria Klimaj <[log in to unmask]> wrote:
Hi SPM experts,

A quick question on thresholding:

For my data, I'm not using cluster-level thresholding because it doesn't seem to change the corrected T in SPM. If this is the case (unless I'm doing it wrong), there doesn't seem to be any justification for using cluster level thresholding and ignoring even a single significant voxel.

The voxel T-statistics and corrected voxel T-statistics are not influenced by clustering. You want to use cluster thresholding because you can make inferences on the clusters and people don't believe that single voxels are active because you likely started with larger voxels and have smoothed your data.
 

Why don't people use a hierarchical approach, simultaneously implementing different corrections for different sized clusters?

I'd take a look at TFCE in FSL, its close to what you are suggesting.
 
Using something like this, a single voxel would have the most stringent requirements, and each additional voxel would lower the required T (as determined by Monte Carlo simulation or some such thing?). I'm not a statistics expert, but it seems like this would be a very effective default procedure. Is there any way of utilizing this kind of thresholding in SPM?

In one of my earlier papers, we used several voxel height and extent thresholds and selected voxels that were included in any of them. Look for papers by Burton and McLaren. A few people raised questions about whether using multiple thresholds increased the false positive rate beyond 5% because of the multiple corrections. I don't have a good answer for that question. In the extreme case, where you have 1 voxel with a height requirement, then 2 voxels with a different height requirement, etc. This is an interesting concept that I've thought about myself; however, it would be time consuming to run all possible permutations. The idea behind TFCE is to account for the height an extent criteria at the same time and only do the thresholding once. I believe TFCE requires non-parametric testing though, so it has to be implemented through randomise. I think someone may have implemented in SPM as well, but I'm not entirely sure.
 


Thanks so much for any help!
-Victoria