Hi,
You can split the files (in time, not space/slices) by using fslroi.
All the best,
Mark
> On 27 Nov 2014, at 16:15, Mayte Parada, Ms. <[log in to unmask]> wrote:
>
> Splitting the files makes complete sense, and I was attempting to do that but I could not figure out how to with FSL. Is it a matter of clipping slices or can the files but split another way before I run them through tensor ICA.
>
>
>
>> On Nov 27, 2014, at 10:55 AM, Mark Jenkinson <[log in to unmask]> wrote:
>>
>> Hi,
>>
>> If you’ve got two different movies then you’ll need two separate tensor ICA analyses (one for each movie). This will require you to split up the scans into pieces where each piece corresponds to one movie. However, combining the results afterwards will not be straightforward and it really depends on what you want to do. Comparing different groups with the same movie type is not too difficult, and forming a difference within each subject between the strongest “movie component” and then comparing between groups would also be possible. The one thing you cannot easily do is put both movie types into one analysis (without forming the difference beforehand). This is for the same reasons as the general restrictions for repeated-measures analyses in randomise.
>>
>> All the best,
>> Mark
>>
>>
>>
>>
>>
>>
>>> On 25 Nov 2014, at 02:03, Mayte Parada, Ms. <[log in to unmask]> wrote:
>>>
>>> Thanks Mark,
>>>
>>> I have a couple of files from 2 subjects that were run using movies as stimuli however, there were two scans and in scan 1, participants watched a neutral movie for 7 min followed by an erotic movie for another 7 min…in scan 2 it was vice versa. How do I tell melodic when one movoe starts and stops so that I can compare activation by movie type? In FEAT this was easily done inputing .txt files as custom but in Melodic (since its model free) this isn’t an option. I’m a screwed here with these scans or can I somehow deal with this?
>>>
>>> --
>>> Mayte Parada Ph.D.
>>> ____________________________
>>> Laboratory for the Biopsychosocial Study of Sexuality
>>> McGill University
>>> 1205 Docteur Penfield
>>> Montreal QC
>>> Canada H3A 1B1
>>> [log in to unmask]
>>> (514) 398-5323
>>>
>>>> On Oct 30, 2014, at 5:10 PM, Mark Jenkinson <[log in to unmask]> wrote:
>>>>
>>>> Hi,
>>>>
>>>> If all the subjects watch the same movie, so that this stimulus is time-locked across subjects, then you can use tensor ICA to do the analysis.
>>>>
>>>> All the best,
>>>> Mark
>>>>
>>>>
>>>>> On 27 Oct 2014, at 14:08, Mayte Parada <[log in to unmask]> wrote:
>>>>>
>>>>> Hello FSLers,
>>>>>
>>>>> There have been some recent papers that cover brain activation in response to long movie clips. This seems fairly new but with my testing paradigm, movie clips are ideal so I am wondering if anyone has had experience using FSL to analyze data from movie clips using Melodic ICA and if it was fairly straight forward? The papers I've found have used other analysis tools.
>>>>>
>>>>> The clips I'm hoping to use could be about 7-10 min long and I have a second physiological measure (like body temperature) that I eventually want to correlate with brain activity. I expect body temp to go increase steadily throughout the clip. I plan to have two movies with different content in them. Last, the participants will be subjectively rating the movies continuously.
>>>>>
>>>>> If anyone has any experience with this I would love to chat about their experiences, difficulties etc. I would like to collect some data and know that I will be able to get something from it and not have to go back to the drawing board.
>>>>>
>>>>> M.
>>
|