Hi
As I mentioned in an earlier post, I am very new to diffusion imaging, and trying to explore this area. I am working through the BATMAN pipeline and just restarted because I think my input (either B0 or B1000) are incorrect. The following are the commands I ran as well as the output for residuals, noise, b0, and b1000:
Also in my previous iteration, my unringed image had revered colors compared to the image on the guide (namely, CSF was black compared to white). Any guidance on what I may be doing incorrectly?
Also in my previous iteration, my unringed image had revered colors compared to the image on the guide (namely, CSF was black compared to white). Any guidance on what I may be doing incorrectly?
I can see two possibilities here, depending on what you are referring to as the “unringed image”:
If in reference to the DWI data following mrdegibbs (in which case you’d be looking at a b=0 image), if the CSF were black because the image intensities were low, there’s no way you’d be able to produce the subsequent images that you’ve shown here. My best guess would be that you pressed the “U” key when you opened mrview, which inverted the colour map.
If in reference to the residuals between the DWI data before and after mrdegibbs, then the order in which you provide those two images to the mrcalc call will influence the sign of the resulting image, and hence whether specific areas are light or dark on a hot colourmap. Also, if you’re using your own data:
The residuals of that particular process will depend on the geometry of the anatomy and its alignment relative to the voxel grid;
Macroscopically homogeneous regions shouldn’t have a “colour” in the residuals of this process. The residuals will appear in an oscillating nature at transitions in image intensity. What you’re seeing in the residual map in the BATMAN tutorial is residuals extending from each of the edges of the lateral ventricles, that happen to interfere constructively in the core of the ventricles in that specific instance.
