Hi all,
We forced the -odt float switch and checked headers for correct dimensions
etc. They were all correct/the same yet the issue continued. We resolved
it by splitting the large cohort into three groups which could be
"flsmaths added" together, then we added the three groups together in
Mango. Something to look out for.
Thanks for all your help,
SEAN HATTON B.Med.Sci (Hons) | PhD Candidate | Clinical Trials Manager
Brain & Mind Research Institute
On 3/11/12 7:00 AM, "Stephen Smith" <[log in to unmask]> wrote:
>If there is any possibility of integer overrun then output to:
>
>-odt float
>
>--------------------
>Stephen M. Smith, Professor of Biomedical Engineering
>Associate Director, Oxford University FMRIB Centre
>
>FMRIB, JR Hospital, Headington,
>Oxford. OX3 9 DU, UK
>+44 (0) 1865 222726 (fax 222717)
>[log in to unmask]
>http://www.fmrib.ox.ac.uk/~steve
>----------------------
>
>On 2 Nov 2012, at 20:54, Sean Hatton <[log in to unmask]> wrote:
>
>> After Probtrackx2, for each of the subject's tracts of interest:
>>
>> fslmaths subject1/fdt_paths -thrP 99 -bin subject1_thr_99_bin
>>
>> Then concatenate the thresholded binary images:
>>
>> fslmaths subject1_thr_99_bin -add subject1_thr_99_bin -add
>> subject2_thr_99_bin -add subject3_thr_99_bin -add subject4_thr_99_bin
>> group_ave
>>
>> Then visualise in FSLView thresholding out the bottom 50%. I have seen
>> this this technique in other papers (e.g.
>>10.1523/JNEUROSCI.1088-12.2012)
>> but not in large cohorts > n=100.
>>
>>
>>
>>
>> Very much appreciated,
>>
>> Sean
>>
>>
>>
>>
>>
>> On 3/11/12 5:57 AM, "Matt Glasser" <[log in to unmask]> wrote:
>>
>>> Perhaps you should post the code you are using to call fslmaths, as I
>>> certainly have gone beyond 20 images like this and not gotten any
>>> flattening.
>>>
>>> Peace,
>>>
>>> Matt.
>>>
>>> On 11/2/12 1:50 PM, "Sean Hatton" <[log in to unmask]> wrote:
>>>
>>>> Hi Saad,
>>>>
>>>> Yes, we are adding binarised tracts with flsmaths so if we combine 20
>>>> images the output image will have a threshold range of 0-19. We just
>>>>seem
>>>> to top out at one point and adding any additional images "flattens"
>>>>it to
>>>> a range of 0-1. Should I be able to add three groups of 25?
>>>>
>>>> Much appreciated,
>>>>
>>>> Sean
>>>> ________________________________________
>>>> From: FSL - FMRIB's Software Library [[log in to unmask]] on behalf of
>>>> Saad Jbabdi [[log in to unmask]]
>>>> Sent: Saturday, 3 November 2012 4:58 AM
>>>> To: [log in to unmask]
>>>> Subject: Re: [FSL] Concatenating tracts with fslmaths
>>>>
>>>> Hi - if by concatenating, you just mean adding the images, then you
>>>>can
>>>> do that with fslmaths. There is no limit as to how many images you can
>>>> add. The results should not be binarised, unless requested in fslmaths
>>>> with the "-bin" flag.
>>>> Are you sure it is not just a matter of setting the min-max
>>>>visualisation
>>>> settings correctly in fslview?
>>>> Note: if you want to know the upper and lower values of an image, you
>>>>can
>>>> use "fslstats -R"
>>>>
>>>> Cheers
>>>> Saad
>>>>
>>>> On 2 Nov 2012, at 10:06, Sean Hatton wrote:
>>>>
>>>>> Hi all,
>>>>>
>>>>> We are concatenating white matter tracts from probtrackx and having a
>>>>> strange problem. The tracts are binarised to a threshold of 99%.
>>>>>When we
>>>>> start concatenating the tracts (fslmaths image1 -add image2 etc
>>>>>output)
>>>>> all goes well until we hit around n=25, any more and the whole image
>>>>>is
>>>>> binarised. We have n=80 and we can create three concatenated images
>>>>>of
>>>>> n=~25, but then joining them together binarises them. Is there an
>>>>>upper
>>>>> limit to the number of images you can concatenate with fslmaths
>>>>>before
>>>>> it falls over?
>>>>>
>>>>> Cheers,
>>>>>
>>>>>
>>>>> Sean Hatton
>>>>> Brain & Mind Research Institute
>>>>> University of Sydney
>>>>
>>>> --
>>>> Saad Jbabdi
>>>> University of Oxford, FMRIB Centre
>>>>
>>>> JR Hospital, Headington, OX3 9DU, UK
>>>> (+44)1865-222466 (fax 717)
>>>> www.fmrib.ox.ac.uk/~saad
>>
|