Gabor,
It has been pointed out to me that this technique, ie, determining the validity of clusters thresholded at one z-value by comparing them with clusters at a different z-value, is unlikely to be valid. If you intended to publish such results you would need to back them up with compeling evidence for the validity of the statistics represented by this method.
Regards,
Dave
------Original Message------
From: Gabor Perlaki
Sender: FSL - FMRIB's Software Library
To: [log in to unmask]
Subject: Re: [FSL] postprocessing
Sent: 8 Nov 2009 12:29
This is exactly what I want. Do you know some command which can I use to
keep count of the 2.3 clusters and split them into separate images?
On Sat, 7 Nov 2009 08:37:22 +0000, Dave Flitney <[log in to unmask]> wrote:
>Does something like this do what you want?
>
>First split your 2.3 clusters into separate images.
>Multiply each 2.3 cluster image by the combined 3.1 clusters image and
>discard 2.3 cluster image if result empty, i.e., no common voxels.
>Re-combine all surviving 2.3 cluster images.
>
>Dave Flitney, IT Manager
>University of Oxford, FMRIB Centre
>JR Hospital, Oxford OX3 9DU
>T: 01865 222713 F: 01865 222717
>
>On 7 Nov 2009, at 03:48, Gabor Perlaki <[log in to unmask]>
>wrote:
>
>> I think the problem is not solvable with a simple multiplication of
>> masks. I
>> would like to keep the clusters from the thresholding 2.3 which
>> contains at
>> least one voxel which is active in thresholding 3.1 as well. But I
>> would
>> like only these clusters from thresholding 2.3. I don't need
>> clusters which
>> are active in thresholding 2.3, but don't contain any voxels which are
>> active in thresholding 3.1. So I want to do a new map from clusters of
>> thresholding 2.3, which have seed with z>3.1.
>>
Sent using BlackBerry® from Orange
|