Hi Mark,
One more question; I would much prefer to have FSL calculate the FWE corrected probability. I have the correct masks and know the spatial smoothness of the data (10 mm FWHM). How would I go about using FSL to do cluster statistics on the tstat maps that were provided to me?
Many thanks,
Diederick
On Jun 6, 2012, at 4:28 PM, Diederick Stoffers wrote:
> Thanks Mark! Didn't grasp the --osize option from cluster's info, that will save me some scripting.
>
> The spm's and cluster sizes were provided to me by an expert AFNI user, not something I would normally use in my FSL-based analysis.
>
> Cheers,
>
> Diederick
>
> On Jun 6, 2012, at 3:41 PM, Mark Jenkinson wrote:
>
>> Dear Diederick,
>>
>> We don't normally specify the size as a threshold, but rather use the FWE corrected probability.
>> It is related to size though, and so if you have the size values (and they are validly calculated)
>> then you can do it that way around. In that case you would use the --osize option (which puts
>> a value into each voxel in a cluster that is equal to the cluster's size) and then use fslmaths
>> with the -thr set to your cluster size threshold.
>>
>> All the best,
>> Mark
>>
>>
>> On 6 Jun 2012, at 20:41, Diederick Stoffers wrote:
>>
>>> Hi all,
>>>
>>> I have a collection of Z-maps and a table with cluster sizes needed to survive FWE-correction at different combinations of uncorrected and corrected thresholds. To my surprise it seems there is no easy way to output masks for a set cluster size threshold; the only I see is to combine the output from the cluster command, awk and fslmaths to get these masks. Am I correct or is there an easier way?
>>>
>>> Cheers,
>>>
>>> Diederick
>
|