Parcellation without colors in mrview

1

Hi all

I’m trying to parcellate connectome using a my own atlas. I’ve followed the BATMAN tutorial for this. My parcellation is originally derived from HCP data and labels look as they should in wb_view. However when I open the final parcellation in mrview only subcortical regions are seen while the entire cortex is black (only subcortical structures are visible). What could have caused this?

Best
Magda

2

Hi @magda_brahe, and welcome!

What intensity values are present in the parcellation, and did you change the view options to include the full range of intensities?

Best,
Steven

3

Hi and thank you for your reply! I played around with the intensity but cortex remains black. I’m not allowed to attach anything but the first seven labels have the following color LUT values

1 LABEL_1 71 150 232 0
2 LABEL_2 62 100 77 0
3 LABEL_3 18 96 243 0
4 LABEL_4 95 240 15 0
5 LABEL_5 232 189 193 0
6 LABEL_6 223 46 193 0
7 LABEL_7 122 17 255 0

Labels as well as colors have all been created using wb_command (it’s also HCP data)

Thanks!
Magda

4

Hi @magda_brahe,

What kind of file is your parcellation? Maybe you can send me a google drive link so I can see if I can view it in my end.

Best,
Steven

5

Hi again. Since I followed the BATMAN tutorial with a production of several intermediate files I have it as dscalar, dlabel, .annot, .mif, .mgz… Which one should send?

When I run
labelconvert my_parc.mif no my_parc.txt my_parc.txt my_parc_parcels_nocoreg.mif

I get

labelconvert: [WARNING] Unexpected values detected in input image; suggest checking input image thoroughly

I’ve tried parcellating the data with HCP-MMP1 and that finishes succesfully.

Best
Magda

6

Hi @magda_brahe,

Mrview will only work with nifti and MIFs out of those file types, so you can send one of those.

Best,
Steven

7

Hi @magda_brahe ,

I had a similar issue before. The possible explanation for the “dark” cortex can be a few labels have a bizarre extremely high value. You can play around with intensity or use a limit. However, this means there is a problem with your labels. Of course, you had to be sure look-up tables were provided correctly.

In my case, labelconvert on the high-performance cluster was the reason -and I couldn’t find a fix- and it was resolved when I used labelconvert on my machine. You can also try with another MRtrix version (3.0.3?).

Best,
Amir