Many thanks for your answer, Dr. Smith,
when i adjust the intensity level for both methodologies in fslview:
TFCE: Min=0.998 Max=1 and Cluster-based thresholding: Min=0.95 Max=1, the same areas are shown in terms of localisation. I understand by the results the analisy has been carried out correctly for both methods and the difference is the intensity level because TFCE has better sensibility than other methods. Is that correct?
I also have another question: When I include in Glm the age covariate there is not differences between both gorups. Does it suggest the only difference between groups is because of the age?
Thanks again!
Best regards,
Patricia.
Hi - this most likely means that TFCE is picking up a low-level pretty much global shift in values between your groups - which is statistically valid (as randomise always gives the right corrected p-values) but should not be over-interpreted in terms of _localising_ where the effect is - however you can still (e.g.) report where the local maxima in the TFCE output image are.
Cheers.
On 12 Mar 2010, at 14:37, SUBSCRIBE FSL Patricia Pires wrote:
> Dear FSL experts,
>
> I'm trying to analyze the differences between 2 groups (1 control group and
> 1 patient group) with tbss and I have some doubts regarding the outputs. I
> am wondering if you could please help me with the interpretation of the results.
>
> I'm comparing the results of randomise using cluster based thresholding (-c
> 1.7 option) and TFCE (--T2 option) with FSL 4.1.4 software according to the
> following procedure:
>
>
> - Cluster based thresholding:
> randomise -i all_FA_skeletonised.nii -m mean_FA_skeleton_mask.nii -o
> resultados_TBSS -d matriz.mat -t matriz.con -n 5000 -c 1.7 -V
>
> - TFCE:
> randomise -i all_FA_skeletonised.nii -m mean_FA_skeleton_mask.nii -o
> resultados_TBSS -d matriz.mat -t matriz.con -n 5000 --T2 -V
>
> The outcomes that I select are:
>
> Cluster based thresholding= _clustere_corrp_tstat1.nii.gz
> TFCE= _tfce_corrp_tstat1.nii.gz
>
> When I compare these images, they are very different! For the cluster based
> thresholding method (-c 1.7), I observe only significance (using Min=0.95
> and Max=1 in fslview) in some regions of the skeleton and in TFCE (--T2),
> almost the entire skeleton is significant.
>
> Can you please help me with the interpretation of these results?
>
> Best regards,
>
> Patricia
>
---------------------------------------------------------------------------
Stephen M. Smith, Professor of Biomedical Engineering
Associate Director, Oxford University FMRIB Centre
FMRIB, JR Hospital, Headington, Oxford OX3 9DU, UK
+44 (0) 1865 222726 (fax 222717)
[log in to unmask] http://www.fmrib.ox.ac.uk/~steve
---------------------------------------------------------------------------
|