I'm not sure why the align would have worked versus not aligned unless you
didn't cut off neck or the registration was different. I would guess that
if you didn't cut off as much neck or input the full image and a weighting
mask to first_flirt to omit the neck the brainstem should segment fine.
If that is the case, the error was because the registered brainstem model
extended beyond the lower extremity of the image.
Cheers,
Brian
>
> Hi Brain,
>
> Originally, I ran 'run_first_all' with non-aligned data (Nifti files
> converted from DICOM using Mricron).
> Mricron reorient MRI (similar to MNI orientation) and cut the neck for
> users (the final image that I used, co*.nii).
> The registration seems to be fine (by checking *_std_sub.nii with
> MNI152_T1_1mm) and all other ROIs were correctly segmented.
> But, brain stem was not segmented.
>
> When I used a resliced AC-PC aligned (AC was set as the origin) MRI, there
> was no error.
>
> Thank you,
>
> Buyean
>
>
>
>
>
>
>
> -----Original Message-----
> From: Brian Patenaude <[log in to unmask]>
> To: [log in to unmask]
> Sent: Tue, Aug 18, 2009 12:00 pm
> Subject: Re: [FSL] Strange FIRST (FSL 4.1.4) error - missing output files
>
>
>
>
>
>
>
>
>
>
> Hi,
>
> I'm coming in a bit late on this thread. first_flirt, the registration
> portion of run_first_all was run on the aligned image, right? My best
> guess is that the seg fault would come from sampling outside the image.
> Did the alignment cut the brain stem quite high?
>
> In general, I would suggest running first on the non-aligned data (i.e.
> not resampled). Then, if you want the output in a different space,
> transform the surfaces to that space.
>
> Cheers,
>
> Brian
>
>
>
>>
>> Hi Mark,
>>
>> The fault was resolved when I pre-processed MRIs with SPM8 Display.
>> All I did was ACPC alignment, setting the AC as origin, resliced the
>> MRIs
>> with 4th degree beta spline.
>> So far, I ran FIRST with three resliced MRIs without the error.
>>
>> By the way, please let me know if this preprocessing can introduce any
>> artifacts in FIRST analysis.
>>
>> If you still want to see the MRIs (before and after ACPC alignment), let
>> me know how to upload it.
>>
>> Thank you,
>>
>> Buyean
>>
>>
>>
>>
>>
>>
>>
>> -----Original Message-----
>> From: Mark Jenkinson <[log in to unmask]>
>> To: [log in to unmask]
>> Sent: Mon, Aug 17, 2009 3:12 pm
>> Subject: Re: [FSL] Strange FIRST (FSL 4.1.4) error - missing output
>> files
>>
>>
>>
>>
>>
>>
>>
>>
>>
>> Hi,?
>> ?
>>
>> The error message here is a Segmentation fault which is?
>>
>> quite serious. Is this repeatable??
>>
>> If so, can you upload the data so that we can see what is?
>>
>> causing it??
>> ?
>>
>> All the best,?
>>
>> ? Mark?
>> ?
>>
>>
>> On 14 Aug 2009, at 19:01, Buyean Lee wrote:?
>> ?
>>
>>> There are about 17 error files, but only this one has 168 bytes (the >
>>> rest have o bytes).?
>>
>>>?
>>
>>> Thank you,?
>>
>>>?
>>
>>> Buyean?
>>
>>>?
>>
>>>?
>>
>>> -----Original Message-----?
>>
>>> From: Mark Jenkinson <[log in to unmask]>?
>>
>>> To: [log in to unmask]?
>>
>>> Sent: Fri, Aug 14, 2009 10:42 am?
>>
>>> Subject: Re: [FSL] Strange FIRST (FSL 4.1.4) error - missing output >
>>> files?
>>
>>>?
>>
>>> Hi,?
>>
>>>?
>>
>>> The logs should be in a directory *.logs and you want?
>>
>>> to look at all files *.e* as these contain the error messages.?
>>
>>>?
>>
>>> All the best,?
>>
>>> Mark?
>>
>>>?
>>
>>> On 14 Aug 2009, at 16:59, Buyean Lee wrote:?
>>
>>>?
>>
>>> > Hi Mark,?
>>
>>> >?
>>
>>> > Where can I find the error logs??
>>
>>> >?
>>
>>> > Buyean?
>>
>>> >?
>>
>>> > -----Original Message-----?
>>
>>> > From: Mark Jenkinson <[log in to unmask]>?
>>
>>> > To: [log in to unmask]?
>>
>>> > Sent: Fri, Aug 14, 2009 6:11 am?
>>
>>> > Subject: Re: [FSL] Strange FIRST (FSL 4.1.4) error - missing > output
>>> > files?
>>
>>> >?
>>
>>> > Hi,?
>>
>>> >?
>>
>>> > Can you send us the details in the error logs??
>>
>>> > We need to see why the brainstem segmentation is failing.?
>>
>>> >?
>>
>>> > All the best,?
>>
>>> > Mark?
>>
>>> >?
>>
>>> > On 13 Aug 2009, at 21:33, Buyean Lee wrote:?
>>
>>> >?
>>
>>> > > Hi,?
>>
>>> > >?
>>
>>> > > Strangely, FIRST ('run_first_all') does not generate the output > >
>>> > files (both *_all_fast_firstseg.nii and *_all_fast_origsegs.nii) >
>>> of > > a few MRIs (5/92 MRIs).?
>>
>>> > > I believe this is due to the fact that FIRST failed to generate > >
>>> > *_BrStem.* files of those 5 MRIs (this is why it can't run >
>>> fslmerge > > in the 'run_first_all' script).?
>>
>>> > >?
>>
>>> > >?
>>
>>> > > Buyean?
>>
>>> <first.e21133.17>?
>>
>>
>>
>>
>>
>>
>
>
>
>
>
>