Hey fsl experts
I´ve been conducting diffusion analysis on data involving to patient groups. By during Dtifit and TBSS different questions have come to mind, which I can´t find answers to, so I hope you can help me.
1.1 The option “—sse” in dtifit gives an estimate of the residual sum of square. Normally SSE is the distance form each point to the regression line, in dtifit its the Sum of squared error of the diffusion tensor fit, but i´m having trouble figuring out what this really signifies?
1.2 in relation “—kurt” gives us the mean kurtosis to the fit, but is it relevant? do we expect the data to be gaussian distributed?
2. In TBSS i´m comparing two patient groups and have completed TBSS and randomise. Group 1. has significantly higher mode of anisotropy (MO), mean diffusivity, L1, L2, L3 than Group 2 [1 -1].
Oppositely is the FA significantly higher in group 2 compared to group 1 [-1 1].
I´m having difficulty interpreting these results, especially that FA is higher in group 2 when MO is significantly higher in group 1?
When displaying these results in Fsleyes there is a limited overlay between MO and FA. My question is what way I should interpret these results, and could I make contrasts that show if MO is significantly lower, as MO goes from -1 to 1, and currently i´m only looking for significantly higher MO between groups?
3. Testing asymmetry in TBSS if you invert the “left_minus_right skeleton” it’s of course giving the same values just with a different sign (-). But these values are not the same/representative to those in the “all_FA_symmetrised_skeletonised”, as when I look at these there is a higher variability between the values between hemispheres. So, by inverting it you are (in my understanding) not really checking the other side?
could this be done in another way, or is my understanding contra factual?
########################################################################
To unsubscribe from the FSL list, click the following link:
https://www.jiscmail.ac.uk/cgi-bin/webadmin?SUBED1=FSL&A=1
|