Problem with whole brain tractography by using tckgen

1

Hi Experts,
I am following the FBA tutorial. After tckgen and tcksift steps, I find my tractography result is bad, the fibre tract does not fill whole brain. There are just few tracts in cerebellum. Could you please help me to figure out what should I do?

The below is the code what I used:
tckgen -angle 22.5 -maxlen 250 -minlen 10 -power 1.0 wmfod_template.mif.gz -seed_image voxel_mask.mif -mask voxel_mask.mif -select 20000000 tracks_20_million.tck -force -nthreads 8
tcksift tracks_20_million.tck wmfod_template.mif.gz tracks_2_million_sift.tck -term_number 2000000 -force -nthreads 8

The figure is the tracks_2_million_sift.tck:

33%20AM
Screen Shot 2018-08-03 at 10.40.33 AM.jpg927×701 191 KB

Thanks,
Elvira

2

Hi Elvira, Maybe check the masks or the wmfod to make sure the tracking space is complete.

3

35%20PM
Screen Shot 2018-08-03 at 12.03.35 PM.png824×655 99.6 KB
19%20PM
Screen Shot 2018-08-03 at 12.04.19 PM.png689×574 43.6 KB

Thank you very much. I have checked the wmfod_template.mif, it looks OK. But the voxel_mask.mif file is bad. Please see the attachment figures.

And I checked the whole steps, I found it is the problem of step 12–Compute the template mask (intersection of all subject masks in template space)

34%20PM
Screen Shot 2018-08-03 at 12.20.34 PM.png684×572 42.7 KB

Could you please tell me what should I do next?
Thanks,
Elvira

4

Yes it seems that the problem is the voxel_mask, How did you obtain that mask? Why is so small? Did you apply any threshold to the white matter mask? It seems that this mask is constraining your whole brain tractogram.

5

Hi Josue,

Yes, you are right. It is the problem of my mask. I checked my data, and found that one subject’s brain do not scan completely.

Thanks very much.

Best,
Elvira