Methodology for generating hcp1065_avg_tracts_trk atlas

[Joseph Turner]

Hi Dr. Yeh,

I’m a huge fan of your work and DSI Studio.

I am curious about how the .trk files in hcp1065_avg_tracts_trk.zip (Source: https://github.com/data-others/atlas/releases/) were generated. I am trying to reproduce these files myself but haven't been able to match the output.

Here is the workflow I tried:

1. Loaded the HCP1065.1mm.fib.gz group average template. (I also tried using the 1.25mm group average template)

2. Performed whole-brain fiber tracking using default settings.

3. Ran tract recognition and clustering.

The results are not identical to the files in your release. I also noticed this atlas is distinct from the original human.tt.gz atlas provided in DSI Studio that I believe is used for automated tract recognition.

My guess is that rather than using the group average fib template, you might have performed fiber tracking for individual subjects. If that is the case, could you explain how you collapsed them down to a single average .trk file per tract?

Thanks for your help!

Cheers,
Joseph 

[Frank Yeh]

The approach is very different from mine, and even the identical methods were used, the results wont be identical. You may check out the original paper about the methodology.

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.
To view this discussion visit https://groups.google.com/d/msgid/dsi-studio/87c0c59e-f59f-4607-96a4-8b4ab8315305n%40googlegroups.com.

[Joseph Turner]

Hi Dr. Yeh,

Thanks for the quick reply.

I went back through the NatComms paper (10.1038/s41467-022-32595-4), and the methodology for generating the voxel-wise population-based probability maps for each tract (via the 1,065 individual subject entries) makes perfect sense.

However, I’m still scratching my head on one specific detail regarding the tractograms (.trk streamline files) included in the release. While the paper explains the probability maps (voxelwise) extensively, I couldn't find the specific recipe for generating the tract-specific streamlines that populate the final .trk files (like in the visualization in Supplementary Fig. 5).

My guess: Did you take the population probability maps you generated for each tract, apply them as ROIs/Masks onto the group average template, and then run the tracking constrained within those boundaries? No worries if it's not possible for me to identically reproduce your results, I'm just hoping to have a rough idea on how you approached this.

Thanks again for your time and the amazing software.

Best,

Joseph

[Frank Yeh]

Please check out the Method section of the paper. The steps are detailed there.

Best regards 

Frank

To view this discussion visit https://groups.google.com/d/msgid/dsi-studio/eca192b7-9c25-4553-b60d-369325ed32acn%40googlegroups.com.