DSI_studio for lesional MRI

[Marc Zanello]

Dear Mr Yeh,

Thanks a lot and congratulations on your tremendous work in DSI Studio.

I am a clinician and a newcomer to tractography so I will probably  ask silly questions but I can not find the answer easily from the website or the google group.

To give you the context, I am working on patients harboring periventricular nodular heterotopia, a brain malformation that can be subtle. They are known to harbor aberrant fiber tracts. The tractography have been made on a Siemens scan. Diffusion sequence was obtained with 30 directions in the vast majority of the patients.

1/How do I recognize " b0 image acquired at the opposite phase encoding"? If I am correct, the bo image would be T2 with few images. I have such sequences but I do not find the phase encoding direction? If DSI Studio do not produce " xxxxx_dMRI_dir258_PA_SBRef.nii.gz" file means that I do not have the bo image?

2/I want to only use DSI_studio and, if I am not mistaken, the use of FSL Topup is interesting but not a mandatory preprocessing tool. I do not have any bad slice according to DSI-Studio. How can I interpret the quality assessment made by DSI Studio (Fiber index)?

3/Nodular heterotopia can be quite small and difficult to see on raw DWI data. However, one patient suffers from shizencephaly and heterotopia: in this particular patient, it was easy to detect a flip along x-axis (even if it tooks me a while to understand how to correct it...). How can I be sure that all my patients do not have such a flip?

4/It is unclear to me if I have to choose DTI or GQI: using GQI and automatic fiber tracking, a lot of tracks have not been reconstructed in my first round of analysis whereas with DTI all tracks have been reconstructed. More broadly, if I obtain correct tractography with DTI and GQI, which one should I prefer in case of a patient with a brain lesion? I think GQI is more appropriate since it is a model free method?

5/It becomes clear to me that connectivity have to be perform with a very large amount of seeds. What is the correct way to determine the threshold? If I let it to a small value, I have a lot of noise in my matrix but I don't want to loose to much information.

6/Some teams have performed CSD-based tractography for patients suffering from heterotopia. It is possible to perform CSD based tractography with DSI Studio?

Thanks a lot for all your help,

Sincerely yours

M.ZANELLO

[Frank Yeh]

> 1/How do I recognize " b0 image acquired at the opposite phase encoding"? If I am correct, the bo image would be T2 with few images. I have such sequences but I do not find the phase encoding direction? If DSI Studio do not produce " xxxxx_dMRI_dir258_PA_SBRef.nii.gz" file means that I do not have the bo image?

It is likely that your sequence was not configured to acquire it.
To confirm, you may need to talk to the MR physicists who acquired the
data for you.

>
> 2/I want to only use DSI_studio and, if I am not mistaken, the use of FSL Topup is interesting but not a mandatory preprocessing tool. I do not have any bad slice according to DSI-Studio. How can I interpret the quality assessment made by DSI Studio (Fiber index)?
>

This video may help.
https://www.youtube.com/watch?v=stL4GMeTC1I&t=8s

You don't have reversed phase encoding b0, and thus FSL TOPUP cannot help here.

> 3/Nodular heterotopia can be quite small and difficult to see on raw DWI data. However, one patient suffers from shizencephaly and heterotopia: in this particular patient, it was easy to detect a flip along x-axis (even if it tooks me a while to understand how to correct it...). How can I be sure that all my patients do not have such a flip?

It is rare to have images flipped in x, and it is likely that you
don't have to do it.
Make sure you did not confuse neurology left-right convention to
radiology's convention.

>
> 4/It is unclear to me if I have to choose DTI or GQI: using GQI and automatic fiber tracking, a lot of tracks have not been reconstructed in my first round of analysis whereas with DTI all tracks have been reconstructed. More broadly, if I obtain correct tractography with DTI and GQI, which one should I prefer in case of a patient with a brain lesion? I think GQI is more appropriate since it is a model free method?

Both will generate very similar results in your case because you only
have 30 direction acquire in one b-value.

The differences will become more obvious only if you acquired 23+
b-values and hundred's of direction, just as recommended as here:
http://dsi-studio.labsolver.org/Manual/b-table-for-qbi-dsi-and-gqi-scans

>
> 5/It becomes clear to me that connectivity have to be perform with a very large amount of seeds. What is the correct way to determine the threshold? If I let it to a small value, I have a lot of noise in my matrix but I don't want to loose to much information.
>

I usually don't use connectivity matrix in my study, and I am sorry I
don't have a good insight.
My two cents is to check out a good study and see how they did it.

> 6/Some teams have performed CSD-based tractography for patients suffering from heterotopia. It is possible to perform CSD based tractography with DSI Studio?
>

Sure. Here's the instructions:
http://dsi-studio.labsolver.org/Manual/data-exchange-between-dsi-studio-and-mrtrix#TOC-Using-MRtrix-FOD-in-DSI-Studio

Best regards,
Frank

> Thanks a lot for all your help,
> Sincerely yours
> M.ZANELLO
>

> --
> 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/36a368d7-d8d5-4e18-9aee-1921feb925ccn%40googlegroups.com.

[Marc Zanello]

Thanks a lot for your quick reply! And for the instructive video.

Concerning coregistration, what are your recommendations? I perform another rigid body registration after the first one performs by DSI Studio.

My last question is on Track-Density Imaging: I do not understand the process described here = http://dsi-studio.labsolver.org/Manual/tract-specific-analysis

Thanks a lot.

Sincerely yours,

M.ZANELLO

[Frank Yeh]

>
> Concerning coregistration, what are your recommendations? I perform another rigid body registration after the first one performs by DSI Studio.

It depends. You may need to check if the alignment is good.

>
> My last question is on Track-Density Imaging: I do not understand the process described here = http://dsi-studio.labsolver.org/Manual/tract-specific-analysis

I have updated it, and hopefully it helps.

> To view this discussion on the web visit https://groups.google.com/d/msgid/dsi-studio/da7f4f46-aad4-4862-8e7e-ade72c65a0f1n%40googlegroups.com.