dsistudio autotrack not saving out all bundles

[Hamsi Radhakrishnan]

Hi Frank!!

I'm using the latest version of QSIprep (0.21.4) which is using the latest version of DSIstudio to run autotrack- and it looks like some of the bundles (CorpusCallosum, Cingulum, SLF, CST) aren't being saved out in the final output directory. Some of the missing ones do seem to be present in the work directory, but not in subsegments (i.e., just "CorpusCallosum" and not "CorpusCallosumForcepsMinor" etc), and they're not ported over to the main output directory. There are no errors thrown in the logs. The call from qsiprep is 

dsi_studio --action=atk --export_trk=1 --source=<fibfile> --thread_count=3 --output=<workdir> --tolerance=22,26,30 --track_id=Fasciculus,Cingulum,Aslant,Corticos,Thalamic_R,Reticular,Optic,Fornix,Corpus --track_voxel_ratio=2.00 --trk_format=trk.gz --yield_rate=0.0000010000

Thank you so much for your help!

[Frank Yeh]

The new DSI Studio supports the hierarchical segmentation of tracts.
Thus, when only specifying only 'Cingulum', then it will map all its
subcomponents together.

To get each subcomponent individually, its full name has to be
supplied instead of a partial name.

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 on the web visit https://groups.google.com/d/msgid/dsi-studio/8fef9a6c-4879-47e6-887d-b3abce2ad428n%40googlegroups.com.

[Hamsi Radhakrishnan]

Thank you for your quick response, Frank!!

- Despite the command having "Cingulum" in its list, and being in the temporary work directory, no cingulum bundle is found in the final output directory.

- However, for other bundles like the "Aslant" tract, I see the outputs split as ParietalAslant, FrontalAslant etc.

- The log file claims that the other sub-tracts are being attempted as well with the stdout:

nipype.interface INFO:

AutoTrack attempted to reconstruct the following bundles:, * Association_ArcuateFasciculusL, * Association_ArcuateFasciculusR, * Association_CingulumL_SuperiorLongitudinalFasciculus1, * Association_CingulumR_SuperiorLongitudinalFasciculus1, * Association_InferiorFrontoOccipitalFasciculusL, * Association_InferiorFrontoOccipitalFasciculusR, * Association_InferiorLongitudinalFasciculusL, * Association_InferiorLongitudinalFasciculusR, * Association_MiddleLongitudinalFasciculusL, * Association_MiddleLongitudinalFasciculusR, * Association_SuperiorLongitudinalFasciculusL_2, * Association_SuperiorLongitudinalFasciculusL_3, * Association_SuperiorLongitudinalFasciculusR_2, * Association_SuperiorLongitudinalFasciculusR_3, * Association_UncinateFasciculusL, * Association_UncinateFasciculusR, * Association_VerticalOccipitalFasciculusL, * Association_VerticalOccipitalFasciculusR, * ProjectionBasalGanglia_FasciculusLenticularisL, * ProjectionBasalGanglia_FasciculusLenticularisR, * ProjectionBasalGanglia_FasciculusSubthalamicusL, * ProjectionBasalGanglia_FasciculusSubthalamicusR, * Association_CingulumL_FrontalParahippocampal, * Association_CingulumL_FrontalParietal, * Association_CingulumL_SuperiorLongitudinalFasciculus1, * Association_CingulumL_ParahippocampalParietal, * Association_CingulumL_Parahippocampal, * Association_CingulumL_Parolfactory, * Association_CingulumR_FrontalParahippocampal, * Association_CingulumR_FrontalParietal, * Association_CingulumR_SuperiorLongitudinalFasciculus1, * Association_CingulumR_ParahippocampalParietal, * Association_CingulumR_Parahippocampal, * Association_CingulumR_Parolfactory, * Association_FrontalAslantTractL, * Association_FrontalAslantTractR, * Association_ParietalAslantTractL, * Association_ParietalAslantTractR, * ProjectionBasalGanglia_CorticostriatalTractL_Anterior, * ProjectionBasalGanglia_CorticostriatalTractL_Posterior, * ProjectionBasalGanglia_CorticostriatalTractL_Superior, * ProjectionBasalGanglia_CorticostriatalTractR_Anterior, * ProjectionBasalGanglia_CorticostriatalTractR_Posterior, * ProjectionBasalGanglia_CorticostriatalTractR_Superior, * ProjectionBrainstem_CorticospinalTractL, * ProjectionBrainstem_CorticospinalTractR, * ProjectionBrainstem_ReticularTractL, * ProjectionBrainstem_ReticularTractR, * ProjectionBasalGanglia_OpticRadiationL, * ProjectionBasalGanglia_OpticRadiationR, * ProjectionBasalGanglia_FornixL, * ProjectionBasalGanglia_FornixR, * Commissure_CorpusCallosum_ForcepsMinor, * Commissure_CorpusCallosum_Body, * Commissure_CorpusCallosum_Tapetum, * Commissure_CorpusCallosum_ForcepsMajor

Happy to send over more info if helpful!!

[Frank Yeh]

I did a quick test using dsi_studio --action=atk --source=*.fib.gz
--track_id=Cingulum,Corpus

The output folders have the following:

Association_CingulumL
Association_CingulumR
Commissure_CorpusCallosum

There is no subcomponent output such as
Association_CingulumL_FrontalParahippocampal, unless I specify them.
Frontal Aslant and Parietal Aslant are separated tracts without
hierarchical relationship.

Hope this helps. If there is remaining questions, please don't
hesitate to follow up with me.
Frank

[Hamsi Radhakrishnan]

Thank you for testing Frank! I don't see those segments either in our outputs, but might be an issue with qsiprep porting things over. Appreciate your help!

[Frank Yeh]

Glad you reached out to me.
If there is still any issue, or any suggestions, please feel free to
let me know.
Frank

[Matthew Cieslak]

Thanks Frank! Is there a comprehensive list of all the possible subcomponents you can get from autotrack?

[Frank Yeh]

Yes, the complete tree structure is here:
https://github.com/frankyeh/DSI-Studio-atlas/blob/main/ICBM152_adult/ICBM152_adult.tt.gz.txt

The part before the first underscore is the pathway system.
The one after is the tract name.
Following the second underscore is the subcomponents of the tracts.

> --
> 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 on the web visit https://groups.google.com/d/msgid/dsi-studio/00ff94ee-12d9-4004-98a2-dfb1ec961dben%40googlegroups.com.

[Steven Meisler]

Hi Frank,

Sorry to revive an old thread, but I was trying to find the same kind of tract name list on the current GitHub repo, and found one here (https://github.com/data-others/atlas/blob/main/human/HCP842_tractography.txt). I could not find the ICBM152 template on the current main branch.

This newer HCP842 atlas did not have Corpus Callosum sub-structures like the previous ICBM152 one did. There also appear to be other bundles (such as the thalamic radiation) that are not present in the more HCP842 tractography file. Do you know if it is possible to get callosal and other substructures using the current tractography atlas / version of DSI Studio?

Best,

Steven

[Frank Yeh]

Hi Steven,

The current tractography atlas uses a different tract hierarchy from the older atlas. AutoTrack can only reconstruct the tract labels and subcomponents that are defined in the selected tractography atlas.

The old and new versions do not translate to each other easily because the atlas definitions and tract hierarchy are different. If you need the older Corpus Callosum substructures or thalamic radiation subdivisions, the practical option is to use the older ICBM152 atlas. If you want to use the atlas, then the analysis should follow the tract definitions available in current version.

Best,

Frank

On Thu, 21 May 2026 11:10:58 -0700 (PDT), Steven Meisler smeis...@gmail.com wrote:

Hi Frank,

Sorry to revive an old thread, but I was trying to find the same kind of tract name list on the current GitHub repo, and found one here (https://github.com/data-others/atlas/blob/main/human/HCP842_tractography.txt). I could not find the ICBM152 template on the current main branch.

This newer HCP842 atlas did not have Corpus Callosum sub-structures like the previous ICBM152 one did. There also appear to be other bundles (such as the thalamic radiation) that are not present in the more HCP842 tractography file. Do you know if it is possible to get callosal and other substructures using the current tractography atlas / version of DSI Studio?

Best,

Steven

On Monday, May 13, 2024 at 3:18:40 PM UTC-4 Frank Yeh wrote:

Yes, the complete tree structure is here:

https://github.com/frankyeh/DSI-Studio-atlas/blob/main/ICBM152_adult/ICBM152_adult.tt.gz.txt

The part before the first underscore is the pathway system.

The one after is the tract name.

Following the second underscore is the subcomponents of the tracts.

On Mon, May 13, 2024 at 3:02 PM Matthew Cieslak mattc...@gmail.com wrote:

Thanks Frank! Is there a comprehensive list of all the possible subcomponents you can get from autotrack?

On Monday, May 13, 2024 at 2:09:06 PM UTC-4 Frank Yeh wrote:

Glad you reached out to me.

If there is still any issue, or any suggestions, please feel free to

let me know.

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 on the web visit https://groups.google.com/d/msgid/dsi-studio/00ff94ee-12d9-4004-98a2-dfb1ec961dben%40googlegroups.com.

–

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/259ac9dd-3811-4033-9e79-dc771ab83fban%40googlegroups.com.