Neonatal dHCP Data and Shape Analysis

[Maryam Shapourjani]

Hello Dear Frank,

I'm currently working on the neonatal dHCP data, and I want to perform shape analysis on the association pathways of term neonates (analysis of the shape and metrics like length, volume, and area which are viewable in the statistics file).

I previously discussed this with you, and your suggestion for the reconstruction stage was to use the GQI method, because: "when the goal is to extract shape metrics, we should not use the QSDR method because it changes some of the shape features due to the transformation to another space."

Following the reconstruction stage is the tractography stage. In the tractography stage, I used automated tractography.

According to our previous conversation, in the tractography stage: "the adult's tractography atlas is normalized to the neonate's space, and this registration is linear + nonlinear."

Now, my question is:
If the goal is shape metrics and some shape metrics change due to transformation to another space:

• A. In the tractography stage, does the "linear + nonlinear registration, which occurs because the adult's tractography atlas is normalized to the neonate's space", not cause some of the shape features and shape metrics to change?

• B. If it does cause a change in some shape features and shape metrics, which shape metrics change and which ones do not?

• C. If it does not cause a change in some shape features and shape metrics, what is the reason?

• D. What is the difference between the "transformation to another space in the reconstruction stage using the QSDR method" and the "linear + nonlinear registration due to the adult's tractography atlas normalized to the neonate's space in the tractography stage", such that in the reconstruction stage with the QSDR method, some shape features and shape metrics change, but in the tractography stage, despite the linear + nonlinear registration, the shape features and shape metrics do not change?
   
  
  Sincerely, 
  Maryam Shapourjani