Hi,
If you are seeing a lot of variation within the run then I can understand why you would try this. It isn't the optimal model for fixing this problem (a future version of eddy should be better) but at the moment it is probably the best we have. In that case you will need to save out the motion correction matrices with the -mats option and then run a script to call the script avscale on each matrix (with the --allparams option) and then save these results.
All the best,
Mark
> On 5 Jul 2017, at 17:09, Rosanne L. Rademaker <[log in to unmask]> wrote:
>
> Dear Mark,
>
> Thanks for the response! The reason I’m using 12 dof is because we’re using multi band, and even after reconstructing and unwarping there is considerable warping clearly visible even within a single run. Moreover, I am concatenating my data across 3 days of scanning, and I do not want parts of the brain moving in and out of voxels. The 12 dof MCFLIRT is doing a good job of correcting all of this, I just have not been able to find any quantifiable output of how much it is actually correcting.
>
> -Rosanne
>
>
>
>
> On Jul 5, 2017, at 1:31, Mark Jenkinson <[log in to unmask]> wrote:
>
>> Dear Rosanne,
>>
>> Is there any reason that you are using 12 DOF for motion correction?
>> This does not make a lot of sense typically (as the brain is only translating and rotating, not scaling or skewing due to motion). That is the reason that the mcflirt tool only outputs 6 DOF.
>>
>> All the best,
>> Mark
>>
>>
>>> On 15 Jun 2017, at 19:35, Rosanne Rademaker <[log in to unmask]> wrote:
>>>
>>> Hi all,
>>>
>>> I am currently running MCFLIRT for motion correction with 12 dof. I want to regress out my motion parameters so have included the -plots flag. However, I noticed that only 6 parameters (x-y-z and roll-pitch-yaw) are saved out. How can I save out the other 6 parameters?
>>>
>>> Thanks,
>>> -Rosanne
|