Tractography atlas & validity/Data access/Difference in QC/Difference in CLI & GUI

[Maryam Shapourjani]

Hello Dear Frank,

I have a few questions:

1. I am currently working on the dHCP neonate dataset and intend to investigate the shape-related parameters of neonates.
   To avoid nonlinear registration and preserve the subjects’ native anatomical features, I performed the reconstruction using the GQI method so that I can analyze the subjects in their native space without introducing shape alterations.
   During the tractography step, I’m using automated tractography, and I have a few questions:
a. At this stage, is the neonate tractography atlas normalized to the subject’s space?
b. If registration takes place at this stage, is it linear or nonlinear?
c. If nonlinear registration occurs, does that mean the neonates’ shape parameters are affected?
d. If nonlinear registration is indeed applied, how can we interpret the assumption that shape parameters remain unchanged during automated tractography?

2. After performing tractography and extracting different pathways (for example, the AF pathway):
a. Are there any reference tracts available that I can use to compare my tractography results with and verify the accuracy of the extracted pathways?
b. How can one ensure the reliability or validity of the tractography results?

3. Regarding the tractography atlas:
a. Is the neonate tractography atlas based on a single neonate or multiple neonates?
b. Is the adult tractography atlas based on a single adult or multiple adults?
c. How can I access more detailed information about both the neonate and adult tractography atlases?

4. The dHCP neonate data available in the Fiber Data Hub have undergone quality control, and the QC file is also provided there.
   It is stated that 34 scans failed the QC process and were completely excluded, meaning these scans are not available in the Fiber Data Hub (no files from these 34 scans are included).
    Is it possible to access these 34 subjects (in src and fib formats) and also their QC information?

5. I already had the raw dHCP neonate data (the nii, bval, and bvec files) from the third release.
When I perform QC on these raw data, the results I obtain differ from the QC results provided in the Fiber Data Hub.
   For example, in the QC I performed on the raw data, the dimension is shown as 100 × 100 × 64, which is different from the dimension reported in the QC file available in the Fiber Data Hub, and the NDC values are also different.
   I would appreciate it if you could guide me to understand why this is happening.

6. My last question is about the difference between the GUI and the CLI.
   Are there any parameters (for example, fiber tracking parameters) that are available and adjustable in the CLI but not in the GUI?
Or vice versa — parameters that exist in the GUI but are not available in the CLI?

Sincerely,
Maryam Shapourjani

[Frank Yeh]

   During the tractography step, I’m using automated tractography, and I have a few questions:
a. At this stage, is the neonate tractography atlas normalized to the subject’s space?

Not exactly, it is adult's tractography atlas to the neonate's space

 

b. If registration takes place at this stage, is it linear or nonlinear?

linear+nonlinear

 

c. If nonlinear registration occurs, does that mean the neonates’ shape parameters are affected?

The answer depends on what " neonates’ shape parameters " are.
 

d. If nonlinear registration is indeed applied, how can we interpret the assumption that shape parameters remain unchanged during automated tractography?

The same as above. Please specify "shape parameters"

2. After performing tractography and extracting different pathways (for example, the AF pathway):
a. Are there any reference tracts available that I can use to compare my tractography results with and verify the accuracy of the extracted pathways?

HCP1065 atlas: https://www.nature.com/articles/s41467-022-32595-4
HCP842: linkinghub.elsevier.com/retrieve/pii/S1053-8119(18)30432-4
 

b. How can one ensure the reliability or validity of the tractography results?

Test-retest reliability: https://www.sciencedirect.com/science/article/pii/S1053811920308156?via%3Dihub
validity: see neuroanatomy textbooks, dissection results...etc. 

 

3. Regarding the tractography atlas:
a. Is the neonate tractography atlas based on a single neonate or multiple neonates?

It is based on the adult atlas.
 

b. Is the adult tractography atlas based on a single adult or multiple adults?

atlas from a population averaged template: see HCP1065 and HCP842
 

c. How can I access more detailed information about both the neonate and adult tractography atlases?

There is no neonate tractography atlas in DSI Studio. For adults, references are given above.
 

4. The dHCP neonate data available in the Fiber Data Hub have undergone quality control, and the QC file is also provided there.
   It is stated that 34 scans failed the QC process and were completely excluded, meaning these scans are not available in the Fiber Data Hub (no files from these 34 scans are included).
    Is it possible to access these 34 subjects (in src and fib formats) and also their QC information?

Sorry they are not available, but you may check out the subject ID to see which are missing and get them from the source.

5. I already had the raw dHCP neonate data (the nii, bval, and bvec files) from the third release.
When I perform QC on these raw data, the results I obtain differ from the QC results provided in the Fiber Data Hub.
   For example, in the QC I performed on the raw data, the dimension is shown as 100 × 100 × 64, which is different from the dimension reported in the QC file available in the Fiber Data Hub, and the NDC values are also different.
   I would appreciate it if you could guide me to understand why this is happening.

You may open the FIB files and check the methods section at Step T3 Fiber Tracking.

The top menu at Step T3 Fiber Tracking also has "FIB protocol" to tell the differences between the steps used in making the FIB file. 

6. My last question is about the difference between the GUI and the CLI.
   Are there any parameters (for example, fiber tracking parameters) that are available and adjustable in the CLI but not in the GUI?
Or vice versa — parameters that exist in the GUI but are not available in the CLI?

There are some parameters that are obsolete and not shown in GUI.

You may check the console window and DSI Studio will still list them to identify them.

Best regards,

Frank

 

Sincerely,
Maryam Shapourjani

--
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/51161786-88d1-47eb-81d8-de1a3f6da95cn%40googlegroups.com.

[Maryam Shapourjani]

Hello dear Frank,

I am sorry for my late reply.

Thank you for the guidance.

Regarding question 1 (sections c and d) which I asked, you mentioned that I should specify which of the shape parameters I am considering.

The shape parameters that I want to extract and examine from the neonates include the following:

number of tracts
mean length(mm)
median length(mm)
span(mm)
curl
elongation
total volume(mm^3)
1st quarter volume(mm^3)
2nd and 3rd quarter volume(mm^3)
4th quarter volume(mm^3)
total surface area(mm^2)
total radius of end regions(mm)
total area of end regions(mm^2)
irregularity
area of end region 1(mm^2)
radius of end region 1(mm)
volume of end branches 1
area of end region 2(mm^2)
radius of end region 2(mm)
volume of end branches 2

Sincerely,
Maryam Shapourjani

[Frank Yeh]

You may check out tutorial videos at practicum.labsolver.org about how
to get tract statistics.

Best regards,
Frank

> To view this discussion visit https://groups.google.com/d/msgid/dsi-studio/7030cdf6-7d81-4377-b444-e095a7903defn%40googlegroups.com.