Codes for DTI analysis of the CSG clinical reports

Data Preparation

1. From the DICOM files extrct the structural image and save as T1.mif. Open a terminal in the dicom folder and type:

mrconvert  dicom_folder T1.mif 

From the list check the number dedicated to the t1_mp2rage_sag_p2_iso_FLAWS_fast2_INV2 sequence, type it and press Enter. you can check the image using mrview:

mrview T1.mif 

2. From the dicom files, concatenate the three shells of the dwi images and save it as dwi.mif. For that, in the terminal type:

mrcat . . . dwi.mif

Then it asks for the number dedicated to the first shell sequence. Write the dedicated number to the ep2d_diff_mddw_30_p2_s4_b700_AP_DO_NOT_READJUST and press Enter. Then it will ask for the numbers dedicated to the second and third shells accordingly. For the second shell, write the number dedicated to the ep2d_diff_mddw_64_p2_s4_b1000_AP_DO_NOT_READJUST and for the third shell, write the dedicated number to the ep2d_diff_mddw_64_p2_s4_b2000_AP_DO_NOT_READJUST sequences. After that, you can check the image using mrview:

mrview dwi.mif

3. Correct the French special characters in the extracted files headers. For this purpose, we should open the mif files using any text reader software, correct for (not delete!!!!) the French special characters and save the files. As these files are big, the suggestion for the Linux users is to open the file in the terminal using mc package. In the terminal, write mc and in the new environment, for example open the dwi.mif file. In the CSG data, usually the problem is in the e letter with accents. These letters in the mc environment are shown with highlighted dots. You can replace these dots with a simple e letter and save the file. Do the same thing for the T1.mif file.

Running the Tractography Scripts

1. Go to the folder that contains the scripts and open a terminal in that folder.
2. In the terminal, run the script related to the tractography:

bash DTI_Pipeline_Clinical.sh 

3. It then asks for the directory of the data. Copy and paste the directory that contains dwi.mif and T1.mif files and press Enter. If everything is done well, and the data are without problem, all the other steps will be done autonmatically and all the results will be saved in the same directory that the data exist. The final tractogram will be saved as Allbrain.tck file. The whole piplien will take around 1 to 2 hours.

4. After pipeline is finished, check the tracts using mrview. For that,

   • open mrview and show the structural image:

   mrview T1.mif 

   • In the Tool tab select Tractography.
   • Click the Open Tractogram icon and select the tractogram file and open it.
   • You will see the white matter tracts on the T1 image. Then you can change the view options in the View tab. For example, Ortho view will show the tractograms on the T1 in Sagital, Coronal, and Axial views.
   • To visualize the tractogram as a whole brain, first in the View tab, select Volume render. Then in the Tool tab, select View options. And finally in the opened tab, click on the Hide main image icon. In the Tool tab, Select the Tractography and there you can adjust different parameters of the tracts such as their thickness, color code, opacity, etc.

Visualizing tracts using Trackvis and generating report images

NOTE: Although we can generate report images using the mrview, CSG people prefer images of the trackvis software. So, we should generate desired images using this software!

1. As trackvis cannot read .tck files, we should convert the tractogram file to a .trk format. In the scripts folder, there is a python script called tck2trk.py which does that. Simply open a terminal in the script folder and type:

python tck2trk.py 

It will ask you the directory which contains the Allbrain.tck file. Copy and paste the directory in the terminal and press Enter. It will automatically read the Allbrain.tck file and save the Allbrain.trk file in the same directory.

2. In the scripts folder, open the dtk folder and run the trackvis file to open the software.

3. Simply drag and drop the Allbrain.trk file into the software and the whole brain tracts will be shown.

4. In the right panel, in the Property box,

   • set the Annotation to off.
   • set Render to Tube
   • Under the Render, set Radius to 0.2

5. Manually put the tractogram in the different directional views (Righ2Left, Left2Right, Front2Back, Back2Front, and Top2Down) and take a screenshot (In linux you can use Shutter software for taking screenshots). Save each screenshots with proper names as follows:

   • Right_Left.png
   • Left_Right.png
   • Front_Back.png
   • Back_Front.png
   • Top_Down.png

and put these files in a specific folder (for example the directory that all the other results also exist).

NOTE: If you think rotating the tractogram is lagging heavily, change the the Render to the line, rotate it easily, and change it back to tube before taking the screenshot.

6. The next step is to put these images next to the images related to the control subject. Some control images and tractograms have been provided in the Subj_ctrlF (Female Control) and Subj_ctrlM (Male Control) folders.
   • In the scripts folder, open a terminal and run the pic_result.py file:

   python pic_result.py

   • First it will ask you the directory where you put the subject tractogram images, and then it asks you the directory of a suitable control subject. Then it will automatically create the images you need for the report and will save them in the same directory that the subject images exist.