Hi,
Ad 1: I suspect that you used the (standard) 4mm resampling resolution? If so then all the results are in MNI152 space, albeit in 4mm isotropic voxel size. If you want to show these in the space of the 2mm MNI152 space youll need to resample using flirt.
Ad2: neither ICA nor DR really loo at FOV or anything else, they just work with whatever is provided. You should certainly check that all registrations worked. ICA and Dr then will provided results in the lowest-common-denominator space, ie.. in the space where all subjects have data.
hth
Christian
On 20 Apr 2014, at 19:58, Broulidakis M.J. <[log in to unmask]> wrote:
> Hi Christian,
>
> Thank you so much for your help! As you suggested I didn¹t input the
> complete path and just started from the directory I was in when I booted
> up melodic.
>
>
> I have two more questions and I would be grateful if you or anyone could
> give me a little more advice. Due to some errors during data collection a
> lot of my resting state data has (to varying degrees) the vertex of the
> brain missing.
>
>
> I have completed group ICA and I am trying to view the file melodic_IC in
> the standard space. However despite the functional data being registered
> (I assume successfully) to the standard MNI 2mm brain as part of group ICA
> I cannot seem to overlay the two (perhaps because of the missing data and
> the two files not having the same base dimensions?) So my questions are,
> 1) is it possible to view the file melodic_IC in the standard space given
> some of the functional data is missing? And 2) would having a small chunk
> of functional data missing interfere with processes like duel regression?
> (Fortunately, the areas missing do not encompass any of my regions of
> interest).
>
>
> Many thanks and once again happy Easter,
>
>
> John
>
>
> --------------------
>
> John Broulidakis
> Developmental Brain-Behaviour Laboratory
> University of Southampton
>
>
>
>
>
> On 19/04/2014 09:15, "Christian F. Beckmann" <[log in to unmask]>
> wrote:
>
>> Hi
>>
>> So C/subj_011/SE9/rest¹ is the complete path to your input file? If so
>> then I suggest pre-processing all data separately and running melodic
>> from the command line instead then it¹s easier to figure out if the
>> pre-processing or the melodic I/O is not happy
>> hth
>> Christian
>>
>>
>> On 17 Apr 2014, at 16:28, Broulidakis M.J. <[log in to unmask]> wrote:
>>
>>> Dear FSL experts,
>>>
>>> I desperately need your help! I am trying to preform group ICA on
>>> resting state data using Melodic and have been running into a serious
>>> difficulty. While melodic seems to work fine for single subject analysis
>>> or even group ICA on only a limited number of participants it is unable
>>> to run my entire data set (n=73).
>>>
>>> At the very start of the analysis the progress report appears to freeze
>>> at:
>>>
>>> /bin/cp /tmp/feat_5cCO9X.fsf design.fsf
>>>
>>> mkdir .files;cp /Applications/fsl/doc/fsl.css .files;ln -s
>>> /Applications/fsl/doc/images .files/images
>>>
>>> Melodic progress report for the first subject generates this error
>>> message:
>>>
>>> /Applications/fsl/bin/fslmaths C/subj_011/SE9/rest
>>> prefiltered_func_data -odt float
>>> Cannot open volume C/subj_011/SE9/rest for reading!
>>>
>>> FATAL ERROR ENCOUNTERED
>>> Cannot open volume C/subj_011/SE9/rest for reading!
>>> while executing
>>> "if { [ catch {
>>>
>>> for { set argindex 1 } { $argindex < $argc } { incr argindex 1 } {
>>> switch -- [ lindex $argv $argindex ] {
>>>
>>> -I {
>>> incr argin..."
>>> (file "/Applications/fsl/bin/feat" line 119)
>>>
>>> I have no idea what it means by ³cannot open volume for reading²
>>> because, as mentioned previously, the dataset (including subj_011) works
>>> fine for many other types of analysis. If anyone else has had a similar
>>> problem or may suspect what the matter is please let me know.
>>>
>>> Thanks in advance and a happy Easter,
>>>
>>> John
>>>
>>> --------------------
>>>
>>> John Broulidakis
>>> Developmental Brain-Behaviour Laboratory
>>> University of Southampton
>>
|