Hello everyone, just starting off by saying how grateful I am that this community exists and that people are so responsive and helpful.
I’m trying to run population_template on my dataset, with dwi masks that have been combined with lesion masks (the lesion masks are subtracted from the dwi masks). I also included the “-nanmask” in order to prevent lesions from contributing to registration/aggregation.
I’m currently running into the error :
[population_template: [WARNING] large scaling ([0.56244526742489, 0.671282921623549, 0.408687896687572]) in linear_transforms_06/01.txt
population_template: [WARNING] large scaling ([0.634096412641267, 0.64672576713187, 0.394523090842308]) in linear_transforms_06/01.txt
population_template: [WARNING] you might want to manually repeat mrregister with different parameters and overwrite the transformation file:
linear_transforms_06/01.txt
press enter to continue population_template]
I was inspecting the “population_template-tmp” folder, and when I examine “initial_template_c0.mif” at the point where the command stopped, it looks like the following screenshot :
The log.txt file stops at :
[near_template05_c0.mif -mask1 /Users/benzhang/Desktop/MotorLearning_wmfods/template/combined_masks/001_combined_mask.mif -affine_scale 0.3 -affine_niter 500 -affine_lmax 2 -type affine -datatype float32 -affine linear_transforms_06/01.txt -affine_init_translation mass -affine_init_rotation search
transformcalc linear_transforms_06/01.txt decompose linear_transforms_06/01.txtdecomp
posix.remove(‘linear_transforms_06/01.txtdecomp’) ]
which seems to be my first participant, so from my understanding it looks like even the first participant is having an issue and it’s not that only one of the wmFOD’s are super far off.
I’m curious if the initial template screenshot is telling of anything regarding something I’m doing wrong? It almost seems like it’s only including the lesion and not the actual brain (although I’m not too sure how the initial template aspect operates)
Thanks so much for the help of the community! Cheers.