I'm trying to complete fiber tracking on some in-vivo spinal cords, but I'm running into a problem with bright spots on the images due to injury. When I construct the diffusion tensor maps, all I get in the output is a few bright spots. When I reconstruct the diffusion tensor maps on our scanner software, ParaVision, I still get FA values for the whole cord. I have "Automatic Brain Extraction" un-checked, and background removal threshold set to 0, so I'm not sure where this thresholding is coming in to play. If you could offer any advice on how to change the thresholding value, or on how to normalize my images, it would be greatly appreciated.

Thanks!

Hi,

Well, this is kind of weird. Brain extraction indeed works only for brain images so it should definitely be unchecked. The threshold is based on the B0 image of your dataset. The tensors (and FA) will be computed only on those pixels where the B0 value is higher than the threshold. So setting it to 0 should trigger an estimation on roughly all pixels.

This is why I find this a bit weird. If you are allowed to do so, I would be curious to see your data and how it behaves on our side. There might be some bug here that we did not see yet. If you don't want to post this data here, you can always send it to us at This email address is being protected from spambots. You need JavaScript enabled to view it.

Hi,
I was just trying to do the same thing myself. It looks like typing in the threshold doesn't do anything - it stays at 100 - whereas changing the threshold with the slider does affect the results.
Paul