Hi Steve,
Thank you for your answers.
I have a follow-up question, which is likely more elementary than my first inquiry.
In terms of the subject-specific RSN maps and group masks, do we need to threshold either of these subject-specific or group RSN maps before using fslmeants to obtain a value of network strength? If the answer is yes, then how do we determine the ideal threshold value?
The reason I'm asking is because I have a feeling that if thresholding is required, then it can become quite subjective as to what cut-off value we should be using.
I appreciate any insight you may have.
Regards,
Sam
On Mon, 11 May 2015 05:30:20 +0100, Stephen Smith <[log in to unmask]> wrote:
>Hi
>
>> On 10 May 2015, at 23:36, Sam Choi <[log in to unmask]> wrote:
>>
>> Hi FSLers,
>>
>> I apologize in advance for my elementary question.
>>
>> I ran MELODIC and dual regression with my resting state data set. Now, I'm interested in measuring the strength of the functional connectivity within a resting state network. Unfortunately, I haven't really seen any post on this forum that discusses how to obtain this strength value. Would the following steps be correct?
>>
>> 1) Use fslsplit on our IC of interest from dual regression stage 2 (e.g., dr_stage2_ic[#ICA].nii.gz) in order to get a subject-specific RSN map corresponding to our IC of interest.
>>
>> 2) Take the melodic_IC.nii.gz file and run fslsplit to get the group mean RSN map corresponding to our IC of interest. Then, binarize the group RSN map since we just want a mask of the group RSN.
>>
>> 3) Finally, run fslmeants (i.e., fslmeants -i dr_stage2_ic[#ICA]_subjX.nii.gz -o strength.txt -m melodic_IC#_bin.nii.gz) for the mean value as a measure of network strength.
>
>yes - all of the above makes sense - in effect you're averaging the subject's RSN strength map within the group-derived mask from the same RSN.
>
>The units are by default the same as the input data scaling, given that by default the model in stage two is normalised.
>
>Given the fact that standard preprocessing normalises the whole average brain intensity to a constant (10000 I believe), then the units are proportional to a % signal change.
>You could get a more explicit version of % signal change by also averaging, within the same mask, the mean_func image.
>
>Cheers.
>
>
>>
>> As a related question, are there no units to the strength value? And, are there a range of values for functional connectivity strength?
>>
>> I appreciate any advice you may have. Thanks for your help!
>>
>> Sam
>
>
>---------------------------------------------------------------------------
>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 <http://www.fmrib.ox.ac.uk/~steve>
>---------------------------------------------------------------------------
>
>Stop the cultural destruction of Tibet <http://smithinks.net/>
>
>
>
>
>
|