[Mristudio-users] Fiber Tracking

[susumu]

1.       How do I select a portion of a selected fiber in DTI studio?   
 
> Use the Cut operation.  After you click CUT, it waits for two inputs. The
first ROI will be treated as "OR" and the result looks like "OR". Proceed
for the second ROI. After the input of the second ROI, only the projection
between the two ROIs are constructed.
 
> Use the "Statistics -> Profile" function. If the trajectory of the
reconstructed fiber is rather straight (not U-shape), you can use this
function. It generates a table in which you can see pixel values averaged
over axial, coronal, and sagittal slices. In this way, you can retrieve
values at each slice level.
 
> Use RoiEditor: Save the result as a "binary" format and read it to
RoiEditor together with the image you want to quantify (e.g. FA map). With
some creative mind, you can do several interesting things using functions in
RoiEditor. For example, with a region growing tool, you can re-define the
fiber coordinates within RoiEditor (called "object") and apply it as an ROI
to any co-registered images such as FA. You can manually modify the object.
You can also manually remove the fiber coordinate in the binary map at one
slice level. Then you can do region growing to define a new object and the
region growing will stop at the slice level you modified.
 
 2.       How do I modify the fiber-data file within DTI studio or using
other softwares such as MATLAB? 
 
> In the FAQ section of www.mristudio.org, there is a section explaining the
format of the DtiStudio fiber files. If you saved your fiber in this format,
you can reload and re-edit them in DtiStudio. Once you convert the data into
the binary format (1/0 for each pixel), RoiEditor is the way to go to modify
them. 

 

3.       How can I read fiber-tracking results saved in Analyze format back
into DTI studio (it only seems to work for dat format)? 

  

> If I'm not mistaken, fiber-tracking results in Analyze format is
pixel-by-pixel "image data" (for example, 256x256x50 matrix data). For
example, a binary image in which pixels that contain the fiber is 1 and
everything else is 0. The original "loss-less" fiber-tracking file should be
streamline information, not the pixel-by-pixel information. For exmaple,
such file should contain information about how many fibers are reconstructed
and then coordinate information of each line. Once this type of streamline
information is converted to an image format, the tracking information is
degenerated and streamline can not be reconstructed (and further editing is
impossible). So, yes, you need the dat format to prevent any information
loss.

 

4.       Where can I access information such as #fibers for a particular
voxel? 

 

> That's a good question. I believe that there is a version that can save an
image for "the number of fibers at each pixel". 
> Hangyi, the version with this function is ready?

 

5.       Are the statistical calculations given by the Statistics Profile
weighted by fibers/voxel? 

 

> Yes, the current version is weighted by the fibers/voxel, but we have a
newer version in which you can choose weighting / not weighting.
 
> I'll check the status of the latest version. I think we will release
Version 3, non-beta very soon.

 

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.mristudio.org/pipermail/mristudio-users/attachments/20090119/617cc7f0/attachment-0001.html