ROI statistics for FA

[colleen....@gmail.com]

Hi All,
I am working with diffusion kurtosis imaging data and would like to use DSI studio to get the FA values for specific ROIs, the corpus callosum, for each subject. I eddy corrected the subjects in FSL and then used DKE for tensor estimation which creates diffusivity and FA maps along with kurtosis parametric maps. It also creates diffusion and kurtosis tensors. I used DKE Fiber Tractography, in which I specified the fa.nii, diffusion tensor and kurtosis tensor, to create .fib files that can be read into DSI Studio. In the .fib file, the QA is replaced by the FA value. What I was thinking of doing was to use subject space JHU-White Matter-labels-1mm and get the FA statistics for the corpus callosum regions in that atlas. Does anyone have any ideas of why this is not a good idea or have a better way of going about this?
Thanks,
Colleen

[Fang-Cheng Yeh]

Hi Colleen,

JHU white matter is great but it has only a label for each voxel,
whereas the corpus callosum fibers are crossed with other major fiber
bundles. If you pull up JHU white matter regions in DSI Studio, you
may see that the CC region only covers the center portion at the
mid-sagittal plane. A more accurate way is doing a fiber tracking
fiber to define the entire corpus callosal regions, or use a
tractography atlas (recently included in DSI Studio, also available at
brain.labsolver.org) to better define the CC regions. You may also
check out other analysis approaches at
https://sites.google.com/a/labsolver.org/dsi-studio/Manual/how-to-analyze-diffusion-data

Best regards,
Frank

> --
> You received this message because you are subscribed to the Google Groups
> "DSI Studio" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to dsi-studio+...@googlegroups.com.
> For more options, visit https://groups.google.com/d/optout.