BrainVISA

Hi,

We've been trying the Cortical Surface Parameterization pipeline on one left hemisphere with no success. Both texture files "<subject>_L_gyri.tex" and "<subject>_L_gyri_regul.tex" displayed only one or two sulci (in the insula). As inputs, we used the "default_session_manual" sulcal graph parameters; as for parameters, we used "name" as sulcus_identification, "nomenclature/surfaceanalysis/constraint_correspondance.txt" as the file_correspondant_constraint, and
"nomenclature/translation/gyri.trl" as the translation parameter. Are there other parameters to tweak in order to make it work?

Then, we tried the 2012 release of the same pipeline with no success either. In this case, we did not find any gyral parcellisation output. Is it an improved pipeline compared to the previous one? Specifically, what is the name of the gyral parcellisation file?

Cheers,
Francois

Hi François,

The older parameterization pipeline has stability problems, and does not work on all data...
The newer one is still unfinished I think.

Denis

Hello François. The 2012 pipeline is not ready yet, no point using it. We're almost there, though. The old one.... I thought it worked, it has been used many times now. So I don't know what would make it not work anymore... I will check it out, but most our energy is dedicated to setting up the new one, really.

Olivier

Thank you for your help. Because I cannot get the old pipeline to work, and that the 2012 pipeline is not ready yet, I think we will choose Arnaud Cachia's gyral parcellation (the other parcellation tool within Brainvisa). Our idea is to use this other gyral parcellation as a starting point. Then, we will use Arnaud Le Trotter's manual drawing tool (for sulci and gyri) to improve the parcellation. This is not too much work since the parcellation will be applied to only one infant brain.

OK. For only one brain, you could almost do the whole parcellation directly with SurfPaint...

Olivier