It seems that you got the matching between the vector orientations and anatomy right. What is not still right (well it is not right or wrong. It's just a convention) is that usually, the first dimension of the data (the x element of your 3D data) is assigned red, second (y) green, and third (z) blue, which usually coincide to right-left, anterior-posterior, and superior-inferior. In your case, the order of the data matrix is right-left / superior-inferior / anterior-posterior. It's just the matter of the order of your 3D matrix.<br>
<br><div class="gmail_quote">On Wed, Jan 27, 2010 at 5:21 AM, mdfarid a r <span dir="ltr"><<a href="mailto:faridisme@gmail.com">faridisme@gmail.com</a>></span> wrote:<br><blockquote class="gmail_quote" style="border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;">
<div><br>Thank you Dr Mori and Dr Hangyi,</div>
<div> </div>
<div>Upon ur advice, I investigated the fibertracking and found that at all green regions, fibertracking was in the superior-inferior direction while at the blue regions, fibertracking was in the anterior-posterior direction. I have swapped the y and z columns of the gradient table and now my fibertracks of the cingulum have much better correspondence to the atlas.</div>
<div> </div>
<div>The consequence also is that, I swapped the green and blue colour around. so that my colormap and eigenvector map does not follow your book. I will try proceeding in this way and see if I experience any other problems with fibertracking.</div>
<div> </div>
<div>Thanks also for directing me to the page about the "issue of +/- definition of gradient tables". It helped me to understand the significance of the eigenvector directions in x,y,z direction.</div>
<div> </div>
<div>Regards,</div>
<div>Farid<br></div><div><div></div><div class="h5">
<div class="gmail_quote">On Wed, Jan 27, 2010 at 12:07 AM, susumu mori <span dir="ltr"><<a href="mailto:susumu@mri.jhu.edu" target="_blank">susumu@mri.jhu.edu</a>></span> wrote:<br>
<blockquote style="border-left: 1px solid rgb(204, 204, 204); margin: 0px 0px 0px 0.8ex; padding-left: 1ex;" class="gmail_quote">Farid,<br><br>There are several points I want to make, although at this point, I'm not 100% sure what went wrong;<br>
<br>> First of all, your tracking results do look odd.<br>> Second, I'm concerned about the non-conventional orientation of your image. The upper right must be axial and the lower right has to be the coronal. Also, your sagittal is 90 degree rotated. In DtiStudio, the X axis (horizontal) of the upper right and the lower right images are the anatomical right-left, which is assigned red. This part is correct. Then the Y (vertical) of the upper right image and the X axis (horizontal) of the lower left image is the anatomical anterior-posterior, which is assigned green. In your image, these axes have blue color and anatomical superior-inferior. In other words, the definition of the green and blue (anterior-posterior and superior-inferior) are switched. I'm not sure how it happened. If you used our software for image normalization, the consistency of the orientation definition (which is always the Radiology convention) is guaranteed, but if you used another software at some point, it's hard for us to track down the problem.<br>
>There is always the issue of +/- definition of gradient tables. Please refer description at "<a href="https://www.mristudio.org/wiki/user_manual/DTI_Studio_User_Manual/fiber_tracking" target="_blank">https://www.mristudio.org/wiki/user_manual/DTI_Studio_User_Manual/fiber_tracking</a>" that explains the issue. Before you apply any image transformation, you have to make sure that this +/- issue is correctly addressed. Practically, you need to confirm that you can get correct fiber tracking BEFORE you do any transformation (normalization) of a data; you want to make sure tracking is right in the native space.<br>
> Another issue is that you need to transform (normalize) the tensor to create a good population-averaged DTI data. The tensor transformation requires special algorithm for tensor reorientation.<br><br>MriStudio offers DiffeoMap (formerly known as Landmarker) that provides the state-of-the-art tensor normalization tools (<a href="https://www.mristudio.org/wiki/user_manual/landmarker" target="_blank">https://www.mristudio.org/wiki/user_manual/landmarker</a>). In this way, you don't have to worry much of the issues described above.<br>
<br>
<div class="gmail_quote">
<div>
<div></div>
<div>On Mon, Jan 25, 2010 at 7:38 PM, mdfarid a r <span dir="ltr"><<a href="mailto:faridisme@gmail.com" target="_blank">faridisme@gmail.com</a>></span> wrote:<br></div></div>
<blockquote style="border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;" class="gmail_quote">
<div>
<div></div>
<div>
<div>Dear DTIStudio users,</div>
<div> </div>
<div>I am interested in investigating the cingulum bundle but am having trouble fibertracking the cingulum bundle.</div>
<div> </div>
<div>Based on the ColorMap0 and Eigenvector0 values in the region of the dorsal cingulum bundle, I would expect fibertracking to occur in an anterior-posterior manner but it is not the case. </div>
<div> </div>
<div>1-ColorMap0.jpeg shows the ColorMap0. At the cingulum bundle region, the 2nd value is the greatest among the 3. (eg: 19,126,27 or 6,136,40)</div>
<div> </div>
<div>2-Evector0.jpeg shows the Eigenvector0. At the cingulum bundle region, the 2nd value is the greatest among the 3. (eg: -0.21,0.95,-0.20)</div>
<div> </div>
<div>3-fibertrack.jpeg is obtained by fibertracking using the binary image of the left cingulum mask from WMPMII </div>
<div>The threshold criteria is restrictive so that not too many fibers obstruct the visualisation of the cingulum (threshold criteria is 0.3 for minimum FA and angle of 50degrees). The resulting fibertrack is still unexpected because it does not traverse in anterior-posterior manner but in the dorsal-ventral manner. </div>
<div> </div>
<div>Fibertracking at 0.15 minimum FA and angle of 70degrees. results in 4-fibertrack.jpeg. I further select the cingulum bundles by using the ROI AND operation at different coronal slices.</div>
<div> </div>
<div>At coronal slice 101, I am still able to obtain reasonable fibertrack results. 5-fibertrack.jpeg, the fibertracks lie in the cingulum region that is expected and extends over around14-18 slices from 101 to 114.</div>
<div> </div>
<div>However, as I try to delineate the cingulum bundle at more anterior coronal slices the cingulum is more horizontal, the fibertracking result delineates in dorsal-ventral manner. In the 6-fibertrack.jpeg, the fibertracks obtained extend over only 7 slices. </div>
<div><br></div>
<div>I do not understand how this is so because the ColorMap0 and the Eigenvector0 seem to indicate that the principle diffusion direction is indeed in the anterior-posterior direction.</div>
<div><br></div>
<div>-----------</div>
<div>Some other information</div>
<div>The image I used is a group registered to the JHU-MNI atlas. The images are averaged across many subjects to increase signal to noise ratio. However, when i repeat fibertracking on single subjects, the fibertracking results are still more dorsal-ventral than anterior-posterior. </div>
<div>------------</div>
<div><br></div>
<div>Pls offer any suggestions. I really need the help. I can offer more additional information on my problem if required.</div>
<div><br></div>
<div>Thank you in advance. I would really appreciate it.</div>
<div>Regards,</div>
<div>Farid</div><br></div></div>
<div>_______________________________________________<br>Mristudio-users mailing list<br><a href="mailto:Mristudio-users@mristudio.org" target="_blank">Mristudio-users@mristudio.org</a><br><a href="http://lists.mristudio.org/mailman/listinfo/" target="_blank">http://lists.mristudio.org/mailman/listinfo/</a><br>
Unsubscribe, send a blank email to: <a href="mailto:Mristudio-users-unsubscribe@mristudio.org" target="_blank">Mristudio-users-unsubscribe@mristudio.org</a><br><br></div></blockquote></div><br><br>_______________________________________________<br>
Mristudio-users mailing list<br><a href="mailto:Mristudio-users@mristudio.org" target="_blank">Mristudio-users@mristudio.org</a><br><a href="http://lists.mristudio.org/mailman/listinfo/" target="_blank">http://lists.mristudio.org/mailman/listinfo/</a><br>
Unsubscribe, send a blank email to: <a href="mailto:Mristudio-users-unsubscribe@mristudio.org" target="_blank">Mristudio-users-unsubscribe@mristudio.org</a><br><br></blockquote></div><br>
</div></div><br>_______________________________________________<br>
Mristudio-users mailing list<br>
<a href="mailto:Mristudio-users@mristudio.org">Mristudio-users@mristudio.org</a><br>
<a href="http://lists.mristudio.org/mailman/listinfo/" target="_blank">http://lists.mristudio.org/mailman/listinfo/</a><br>
Unsubscribe, send a blank email to: <a href="mailto:Mristudio-users-unsubscribe@mristudio.org">Mristudio-users-unsubscribe@mristudio.org</a><br>
<br></blockquote></div><br>