Longitudinal Tractography Mouse brains

[9ale...@gmail.com]

Dear Frank, 

Sorry for disturbing. I'm working on mouse brains and I'm interested in tracking longitudinal differences in this mouse as a preliminary study. 

1) I reconstructed a GQI with your suggested parameters once starting from the baseline brain image (compare SRC: I selected the follow-up image) and once starting from the follow-up image (compare SRC: I selected the baseline image). I did this so to test same outputs are confirmed. Unfortunately I get a map with a higher number of available tracks from differential tractography (visible already from a preview of the whole brain upon selecting a "Differential Tracking index") only when "dec qa" is selected in both differential reconstruction cases. Instead I expected a high-number of tracts with "dec qa" at baseline -minus- follow-up, confirmed by a high number of tracts with "inc qa" from follow-up -minus- baseline.

I hope you can help me with the attached files and also with the low R-squared values. 

Also, some other details:

2) if I want to create a Connectometry Database with mouse brains, which file should I add as Template (since Idon't know hot to add CIVM that I normally use for QSDR)?

3) at the reconstruction step, I flip the main image of the brain (the first one selected) to make it in the same orientation as the CIVM template, will these changes applied to the second image selected upon clicking "Compare SRC"? The same goes for when I edit the main image in order to adapt the mask to the white matter.

4) when working in GQI and/or attaching a file via "Compare SRC", what is the "Deformed" option?

Thank you in advance,

 

Alex

[Frank Yeh]

> 1) I reconstructed a GQI with your suggested parameters once starting from the baseline brain image (compare SRC: I selected the follow-up image) and once starting from the follow-up image (compare SRC: I selected the baseline image). I did this so to test same outputs are confirmed. Unfortunately I get a map with a higher number of available tracks from differential tractography (visible already from a preview of the whole brain upon selecting a "Differential Tracking index") only when "dec qa" is selected in both differential reconstruction cases. Instead I expected a high-number of tracts with "dec qa" at baseline -minus- follow-up, confirmed by a high number of tracts with "inc qa" from follow-up -minus- baseline.
> I hope you can help me with the attached files and also with the low R-squared values.

Could you also send the SRC file?

>
> 2) if I want to create a Connectometry Database with mouse brains, which file should I add as Template (since Idon't know hot to add CIVM that I normally use for QSDR)?

You can choose one animal's FIB file (reconstructed by QSDR) as the template.

> 3) at the reconstruction step, I flip the main image of the brain (the first one selected) to make it in the same orientation as the CIVM template, will these changes applied to the second image selected upon clicking "Compare SRC"? The same goes for when I edit the main image in order to adapt the mask to the white matter.

Good point. I will update DSI Studio to flip the SRC as well. I will
touch base again with you.

> 4) when working in GQI and/or attaching a file via "Compare SRC", what is the "Deformed" option?

This option will consider possible deformation between longitudinal
scans (e.g. tissue swelling). The new version will automatically
consider this and the checkbox is now removed.

Since this is the first time dT applied to animal studies, there may
be glitched and I will be happy to fix the problems for you (and sorry
for the hassle!).

Best,
Frank

>
> Thank you in advance,
>
> Alex
>

> --
> 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/a294d0ad-2a45-4672-a03e-52d4c0789c6bo%40googlegroups.com.

[Frank Yeh]

The low R2 is because the baseline images are flipped, but the follow-up is not.
I am going to fix DSI Studio so that the flip will apply to both image volume.
Frank

[9ale...@gmail.com]

Thank you for all the support!

1) Uploading. Hopefully the Rsquared will increase when flipping also the second image.

2) Ok, good to hear. I think it may even be good to create a template image of the whole group and then proceed. 

3) Regarding the mask editing on the main imag: would it be possible to see it also an the "compared image", as to make sure the mask covers both brains?

4) Very interesting point

Alex

[Frank Yeh]

I fixed DSI Studio. Please download today's version and see if it works for you.

The steps I tested is the following:

(1) Click [Step T2][Reconstruction] open Alex - baseline.src.gz
(2) At the top menu, select [Step T2][Edit][Image flip y]
(3) At the top menu, select [Step T2][Edit][Image flip z]
(4) At the top menu, select [Step T2][Edit][Image swap yz]
(5) Click [Step T2b(2)][Compare SRC] open Alex - followup.src.gz
(6) Set [Step T2b(1)]=GQI
(7) Set [Step T2b(1)][Diffusion sampling length ratio]=0.6
(8) Click [Step T2b][Run reconstruction]

The FIB file should have R90.

The settings for differential tractography:
(1) [Terminate if]=5000000 [Seeds]
(2) [Differential Tracking Index]="dec_fa"
(3) [Differential Tracking Threshold]=0.1

> 3) Regarding the mask editing on the main imag: would it be possible to see it also an the "compared image", as to make sure the mask covers both brains?

The follow-up scan will use the same mask as the baseline.

Best regards,
Frank

[9ale...@gmail.com]

Amazing, 

Thank you for helping me, now the R2 is really high.

Now, going back to my original question, for reproducibility purposes I tried again running both differential indeces on both reconstructions (baseline - followup, and follow-up - baseline). Unfortunately the number of tracts from inc qa in one reconstruction don't seem to be similar in number to the tracts from dec qa of the other reconstruction. Parameters used: whole brain seeding, minimal length 0.5, tracking threshold around 0.06, terminate if 100K seeds. 

Hope you can help me explain it or find other ways around. 

Thank you again, 

Best,

Alex

[Frank Yeh]

The tracking will use the fiber orientations from the first scan, and
the second scan is only used to provide the differences in FA.
Thus if there is a substantial change of fibers in the follow-up scan,
the tracking will not be good if it is used as the baseline.

> --
> 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/46de7c73-1610-46fe-ba1e-113499d98a91o%40googlegroups.com.

[9ale...@gmail.com]

It makes perfect sense, that's why it is important to have the best possible baseline image. 

Thank you again.

Best regards, 

Alex