[Mristudio-users] question about segmentation of CST

Thu Feb 14 08:17:15 EST 2013

Hi Deepa,

Before using the automated method introduced by the Yajing's paper, please
perform the manual-based analysis (paper attached) to make sure your data
are ready for tractography.

The most common problem is, as you indicated, the sign of the fiber
vectors. If you are using Siemens DICOM (MOSAIC), I recommend you to obtain
gradient vector information from the header using the "Get gradient from
DICOM file header" function. If you use this information, no "flipping" is
necessary. If the fiber angles are correct, you should find many
streamlines that connect the brainstem and the motor cortex.

Susumu

On Wed, Feb 13, 2013 at 7:09 PM, Deepa Krishnaswamy <
deepa.krishnaswamy1 at gmail.com> wrote:

> Hi,
> Yes I do understand that you would expect a lot less fibers. I am trying
> to recreate a similar looking CST as is shown in the paper 'Atlas-guided
> tract reconstruction for automated and comprehensive examination of the
> white matter anatomy.' I have been using the sample Siemens DICOM dataset,
> FA of 0.2 and an angle constraint of 40 degrees, as these were values given
> in the paper. This does not yield many fibers at all; in fact most do not
> reach the motor cortex. I've tried experimenting with the FA threshold and
> the angle constraint, but this does not seem to yield any better fibers. Is
> there anything else I can modify to yield more CST fibers? The only other
> issue I have thought of is that the flipping of the eigenvectors was
> incorrect, but I am pretty sure that there must be a Y and Z flip.
> Thanks,
>
> Deepa
>
> On Thu, Feb 7, 2013 at 10:48 PM, susumu mori <susumu at mri.jhu.edu> wrote:
>
>> Hi Deepa,
>>
>> if you specify the second ROI with "1", the fibers have to penetrate both
>> ROIs. So usually, you expect much less fibers.
>>
>> Susumu
>>
>>
>> On Thu, Feb 7, 2013 at 1:21 PM, Deepa Krishnaswamy <
>> deepa.krishnaswamy1 at gmail.com> wrote:
>>
>>> Hi,
>>> I'm using the Siemens DICOM sample data set. I believe that I have to
>>> choose a y and z flip, as that combination yields the correct CST fibers
>>> when I put an ROI at the cerebral peduncle.
>>> I see reasonable tracts when I use two ROIs with operation 0. There just
>>> seems to be a problem when using operation 1 for the second ROI, as there
>>> are a very small amount of fibers selected.
>>> Thanks,
>>>
>>> Deepa
>>>
>>> On Sat, Feb 2, 2013 at 4:51 PM, susumu mori <susumu at mri.jhu.edu> wrote:
>>>
>>>> One potential reason is that your vector angles are not right.
>>>> First, please make sure that your DTI data have correct fiber
>>>> orientations by using manual interactive ROI drawing and fiber tracking by
>>>> DtiStudio. For example, you can define one ROI at the cerebral peduncle and
>>>> make sure the CST trajectory reaches the motor cortex.
>>>>
>>>> If you confirm that everything is working, then it could be the problem
>>>> of your ROI file.
>>>> Do you see reasonable tracts when you use two ROIs with both operation:
>>>> 0?
>>>> If so, then you can see what happens with Operation: 1 for the second
>>>> ROI.
>>>>
>>>> On Wed, Jan 30, 2013 at 12:04 PM, Deepa Krishnaswamy <
>>>> deepa.krishnaswamy1 at gmail.com> wrote:
>>>>
>>>>> Hi,
>>>>> I have used the pipeline at
>>>>> https://www.mristudio.org/wiki/user_manual/diffeomap/GetStarted_DualChannelNormalizationPipeline%3Ato transform the sample Siemens DICOM data to the JHU_MNI_SS atlas. I am
>>>>> currently trying to segment the CST using the ROI operation rules listed in
>>>>> Table 1 of 'Atlas-guided tract reconstruction for automated and
>>>>> comprehensive examination of the white matter anatomy.' I've written my own
>>>>> ROI map file that specifies the ROIs and the operations. In Table 1, there
>>>>> are two ROIs listed, two additional AND ROIs, and a list of NOT ROIs. The
>>>>> syntax I have been using for part of the map file is below:
>>>>>
>>>>> BinaryFile: pathname for ROI 1
>>>>> Operation: 0
>>>>> etc.
>>>>>
>>>>> BinaryFile: pathname for ROI 2
>>>>> Operation: 1
>>>>> etc.
>>>>>
>>>>> Binary File: pathname for first additional AND ROI
>>>>> Operation: 1
>>>>> etc.
>>>>>
>>>>> Binary File: pathname for second additional AND ROI
>>>>> Operation: 1
>>>>> etc.
>>>>>
>>>>> I do the same for all of the NOT ROIs.
>>>>>
>>>>> When I do this, there are no tracts visible. Even if I take out all
>>>>> the NOT ROIs, there are still no tracts. I am able to see tracts only if
>>>>> the operations for the first four ROIs are set to zero (OR). I know that
>>>>> this is incorrect, because the AND operation must be used for those four
>>>>> ROIs.
>>>>>
>>>>> Has anybody encountered a similar problem?
>>>>> Thanks,
>>>>>
>>>>> Deepa Krishnaswamy
>>>>>
>>>>> _______________________________________________
>>>>> mristudio-users mailing list
>>>>> mristudio-users at mristudio.org
>>>>> http://lists.mristudio.org/mailman/listinfo/
>>>>> Unsubscribe, send a blank email to:
>>>>> mristudio-users-unsubscribe at mristudio.org
>>>>>
>>>>>
>>>>
>>>> _______________________________________________
>>>> mristudio-users mailing list
>>>> mristudio-users at mristudio.org
>>>> http://lists.mristudio.org/mailman/listinfo/
>>>> Unsubscribe, send a blank email to:
>>>> mristudio-users-unsubscribe at mristudio.org
>>>>
>>>>
>>>
>>> _______________________________________________
>>> mristudio-users mailing list
>>> mristudio-users at mristudio.org
>>> http://lists.mristudio.org/mailman/listinfo/
>>> Unsubscribe, send a blank email to:
>>> mristudio-users-unsubscribe at mristudio.org
>>>
>>>
>>
>> _______________________________________________
>> mristudio-users mailing list
>> mristudio-users at mristudio.org
>> http://lists.mristudio.org/mailman/listinfo/
>> Unsubscribe, send a blank email to:
>> mristudio-users-unsubscribe at mristudio.org
>>
>>
>
> _______________________________________________
> mristudio-users mailing list
> mristudio-users at mristudio.org
> http://lists.mristudio.org/mailman/listinfo/
> Unsubscribe, send a blank email to:
> mristudio-users-unsubscribe at mristudio.org
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.mristudio.org/pipermail/mristudio-users/attachments/20130214/9994773d/attachment-0001.html 
-------------- next part --------------
A non-text attachment was scrubbed...
Name: WakanaReproducibility.pdf
Type: application/pdf
Size: 7272835 bytes
Desc: not available
Url : http://lists.mristudio.org/pipermail/mristudio-users/attachments/20130214/9994773d/attachment-0001.pdf 