AD: Preclinical Preprocessing

1. ln CSPM, click the ‘Preclinical’ button

Open

2. In the ‘Setup’ tab, make sure ‘Rat’ is selected (with a scale factor of 8), paste your ‘projects’ folder path into the ‘Project folder’ box and the folder path of your ‘stats' folder into the ‘Stats’ box. In the boxes under 'Rat (unscaled)’, paste in the folder path for the unscaled SIGMA rat brain template, GM, WM and CSF (C:\Matlab\spm12\toolbox\CSPM\preclinical\rat\…).

Open

3. In the ‘Main’ tab, you should now see all of your subject folder names in the left hand panel.

4. Tick the ‘DTI’, ‘fMRI' and ‘T2map’ boxes and click the ‘Preprocess to Coreg’ button. This will create ‘coreg’ folders in each subject’s ‘DTI’, ‘fMRI’ and ‘T2map' folders, with the relevant file in each folder, and a text document file named 'coreg_not_done’.

5. Realign i.e. coregister the DTI and fMRI images in the ‘coreg’ folders, for each subject, as detailed in steps 10-12 of the confluence page ‘File/Folder Setup and Preparation for Preprocessing’. Write down the pitch, roll, yaw and crosshair position values in the the text document file named ‘coreg_not done’ and rename this file to ‘coreg_*date*', and make sure to apply these realignment parameters to all the DTI files in the raw folder and all the files in the ‘tensors’ and ‘b0_images’ folders. There is no need to realign the ‘T2map’ files as they were realgined i.e. coregistered in steps 9-14 from the confluence page ‘File/Folder Setup and Preparation for Preprocessing’. If at this point, you notice that your DTI images are slightly distorted (or remain slightly distorted despite distortion correction), follow the below steps:

   1. Copy over each subject’s DTI folder into a temporary folder as a back up.

   2. Open the subject’s b0 DTI image as well as it’s T2map and use the crosshairs and yoking tool to compare the two images and confirm that the DTI is in the same space as the T2map. For example:

   3. In this example we can see that the back of the cerebellum in the DTI (bottom image) is squashed i.e. it does not line up with the back of the cerebellum in the T2map (upper image), and will therefore need to be manually stretched.

   4. In the SPM12 GUI, click ‘Display’ and open up the DTI image.

   5. Adjust the resize x, y and z parameters as required to correct the distortion (values between 1 and 2 will stretch the image, whereas values between 0 and 1 will squash the image). Be sure to write down the resize x, y and z values you used in the ‘coreg text file’.

   6. Once you are happy with how you have stretched/squashed the image, click ‘Set Origin’, followed by ‘Reorient’.

   7. Then open the DTI image you have just reoriented and compare it to it’s corresponding T2map:

      Here we can see that the back of the cerebellum in the DTI image now matches the position of the back of the cerebellum in the T2map image. Also compare the top and bottom of the brain etc. If it still does not match, delete the DTI image you have reoriented and copy over the original b0 image from the back up folder you made in step a, and adjust the resize values accordingly. Another way to compare the T2map and DTI images is by selecting ‘Check Reg’ in the SPM12 GUI and selecting the two images.

   8. Once you are happy with the resize values you have used, apply these same resize values to all the files in the ‘raw’ folder and all the files in the ‘tensors’ and ‘b0_images’ folders (a total of 131 files), but make sure to enter the ‘resize y’ value you used into the ‘resize z’ value box and the ‘resize z’ value into the ‘resize y’ value box (because for some unknown reason ‘resize y' stretches vertically instead of horizontally and vice versa for 'resize z’ i.e. their functions are swapped when applying to the raw, tensor and b0 image files). The only way to avoid this, is to delete the b0 image and replace it with the original b0 image in native space and then enter both the realigning and manual resizing parameters, and then apply these to the b0 image as well as all the files in the 'raw’, ‘tensors’ and ‘b0_images’ folders, at the same time.

6. Now, go back to the ‘Preclinical’ GUI and unselect the ‘T2map’ box and select the ‘VBM’ box i.e. make sure ‘DTI’, ‘fMRI’ and ‘VBM’ boxes are ticked and that the ‘T2map’ box is unticked.

7. Then click into the ‘fMRI’, ‘VBM’ and ‘DTI’ tabs and set the parameters you desire. E.g.

If you are unsure what value to enter in the ‘Normalised voxel size’ box, open one of your raw T2map images in MRIcron, click ‘Window’ → ‘Information’ and look at the values in the ‘Spacing’ column.

8. Click the ‘Preprocess after Coreg’ button. This step will take a few hours to complete, depending on how many subjects you have.

9. Once the ‘Preprocess after Coreg’ step is finished, all of the files necessary for analysis will be in your ‘stats’ folder.