What do the"bundles features" statistics describe

[houenou]

What do the statistics "Volume Ratio Image" (vr) , "Standard Deviation Image" (stdev) and "Diffusion-weighted T2 image" (dwt2) in the bundles-features statistics mean ?
Thanks

[Yann Cointepas]

Volume ratio (vr) is an anisotropy index. I do not remember who introduced it. Try to have a look at "Pierpaoli C, Basser PJ. Toward a quantitative assessment of diffusion anisotropy. Magn Reson Med. 1996;36:893-906.".

Standard Deviation Image (stdev) is the standard deviation of diagonal tensor values:

Code: Select all

stddev = sqrt( (Dxx - adc)^2 + (Dyy - adc)^2 + (Dzz - adc)^2 )

Diffusion-weighted T2 image (dwt2) is the product of diffusion-weighted values to the power 1/number-of-diffusion-weighted-values:

Code: Select all

dwt2 = dw[ 1 ]^(1/number-of-diffusion-weighted-values) *
       dw[ 2 ]^(1/number-of-diffusion-weighted-values) *
       ...
       dw[ number-of-diffusion-weighted-values ]^(1/number-of-diffusion-weighted-values)

[Narly]

Hello there,
You say that 'Diffusion-weighted T2 image (dwt2) is the product of diffusion-weighted values to the power 1/number-of-diffusion-weighted-values'. What does this mean in non-mathematical terms? In other words, how can a difference in the mean dwt2 value along a bundle between two populations be interpreted?
thanks!
Narly.

[Narly]

Hello there,
You say that 'Diffusion-weighted T2 image (dwt2) is the product of diffusion-weighted values to the power 1/number-of-diffusion-weighted-values'. What does this mean in non-mathematical terms? In other words, how can a difference in the mean dwt2 value along a bundle between two populations be interpreted?
thanks!
Narly.

[Jean-Francois Mangin]

Nobody really knows, it is still largely up to your interpretation... You can pick something among differences in axon packing, myelin sheet thickness, axon diameters... This is a key issue still largely open:-(

[Narly]

Hi Jean-Francois,
Thanks for your response. Can you point me to any references in which they discuss (or at least propose) the possible interpretation for dwt2 image values?
best,
Narly.

[Jean-Francois Mangin]

Sorry Narly,
I think I read your post too fast. I do not know what Yann is talking about...

[Narly]

Hi Jean-Francois,
So which measure is it that can be thought of as reflecting the things you mentioned (differences in axon packing, myelin sheet thickness, axon diameters), is it the FA?
thanks,
Narly.

[Jean-Francois Mangin]

FA is the most used, some people begin to use also the eigenvalues, but I am not really following this literature which is huge. I am afraid you have to get in touch with people that spent a lot of time thinking about this kind of issues. The ones you met in Orsay could be Gwen Douaud and Jessica Dubois, but other lab are full of other interesting people...