Hi,
That is still quite a large FOV.
If this effectively covers all, or most, of the simulated FOV then there will be no timing benefit to excluding the other areas of the input brain image. It is only when you exclude things within the simulated (EPI) FOV that you get some benefit, as POSSUM ignores areas of the input brain image (object) that are outside the simulated FOV. Maybe that explains the situation.
All the best,
Mark
On 23 May 2014, at 10:00, Joerg Pfannmoeller <[log in to unmask]> wrote:
> Hello,
>
> I applied fslroi to the Possum-Data input object. The CLI command was:
>
> fslroi brain.nii.gz brain_crop.nii.gz 15 165 15 200 10 155
>
> Afterwards I replaced the input object by brain_crop.nii.gz. Concerning the test simulation:
>
> Activation, but no artifacts were included.
>
> So long pfannmoe
>
>
>
>
> On Thu, 22 May 2014 20:26:30 +0000
> Mark Jenkinson <[log in to unmask]> wrote:
>
>> Hi,
>>
>> This isn't what I get - for me it makes a noticeable difference to the run-time.
>> What exactly are you changing?
>>
>> All the best,
>> Mark
>>
>>
>> On 22 May 2014, at 08:36, Joerg Pfannmoeller <[log in to unmask]>
>> wrote:
>>
>>> Hello,
>>>
>>> thank you a lot for your answer. I just would like to make this thread complete:
>>>
>>> After a test I found that the computation time stays the same regardless wheter the full brain or partial brain is used.
>>>
>>> Sincerely yours pfannmoe
>>>
>>>
>>> On Wed, 21 May 2014 11:00:37 +0000
>>> Mark Jenkinson <[log in to unmask]> wrote:
>>>
>>>> Hi,
>>>>
>>>> The simulation is still valid in this situation.
>>>> It is the equivalent of putting a smaller piece of tissue (e.g. a postmortem sample) into the scanner.
>>>> The simulation that you get is the same as if the specified region is the only object in the scanner.
>>>>
>>>> All the best,
>>>> Mark
>>>>
>>>>
>>>> On 17 May 2014, at 10:27, Joerg Pfannmoeller <[log in to unmask]> wrote:
>>>>
>>>>> Dear Experts,
>>>>>
>>>>> concerning simulation of fMRI data, in which only the central region of the brain around the central sulcus is included and large parts of the frontal and parietal parts of the brain are not included. The runtime estimation of Possum decreases if those regions are removed from the input object. Does this really decrease the runtime and is the simulation still valid in this case?
>>>>>
>>>>> Regards pfannmoe
>
>
> --
> Joerg Pfannmoeller <[log in to unmask]>
|