Measuring GM edema using MSMT

1

Hi experts,

I would like to quantify amount of edema in gray matter and looking for the optimal method for it. Following this thread, I aimed at the CSF coefficient in GM, but its value in GM is commonly almost 0, and in many voxels even negative (with values about -1e-13, I guess this is a numerical error). Higher values are found at GM voxels near subarachnoidal CSF, but it is caused by partial voluming from CSF, so it seems that CSF coefficient is not useful to me.

I guess, that trying to quantify edema by identifying “pure water” signal will always suffer by hardly unsolvable partial voluming from CSF, regardless of the actual method used (NODDI, FW-DTI, etc.), I think.

I found somewhere the suggestion to compare GM coefficients instead, but it does not make much interpretation sense to me.

Do you have any recommendation how to quantify amount of edema in gray matter?

Thank you in advance for any suggestion,

Antonin Skoch

1 Like
2

Hi Antonin

I would suggest starting with interrogating the raw data. Need to know whether, within those regions where one is subjectively confident that edema is present, the diffusion-weighted signal “looks like” a linear combination of your GM-like and CSF-like response functions, or whether it instead only looks like a brighter version of the GM-like response function. These could all be shown as signal intensity as a function of b.

If the latter, then it’s possible that the GM ODF density is actually a somewhat sensible marker pragmatically, even if the interpretation of “there’s more GM” is not really true, it’s more of a T2 effect.

Also check to ensure that GM voxels affected by edema are not what is being chosen for response function construction.

Cheers
Rob