Seeking Automation Method for Post-FSL BET and FLIRT 4D fMRI Image Verification

, ,
1

Hello,

I am new to this field and currently using FSLeyes to manually verify the quality of 4D fMRI images after processing them with BET and FLIRT. However, this manual verification process is quite time-consuming. I am looking for a way to automate this verification process.

Specifically, I am interested in:

  1. Steps to automate the processing of 4D fMRI images using BET and FLIRT.
  2. Automated scripts or tools that can help verify the processed images.

My questions are as follows:

  • Is there a way to automate the verification of 4D fMRI images after processing them with BET and FLIRT using FSL or any other tools?
  • Are there any recommended scripts or examples that can assist in this automation?
  • Additionally, could you explain how this process is generally handled by more experienced practitioners?

I would greatly appreciate any references, advice, or examples to help solve this issue.

Thank you.

2

Hi @e.lee and welcome to neurostars!

You might find it useful to make a python script that loops through your images and makes brain plots with something like Nilearn. Doing this in a jupyter notebook could be a good place to manually inspect your images with a higher throughput. 7. Plotting brain images - Nilearn

Best,
Steven

1 Like
3

Hi @e.lee, you can also use FSLeyes from the command-line to generate screenshots. For example, this command could be used to generate a summary of brain extraction:

fsleyes render -of bet.png -slightbox -zr 0 1  struct.nii.gz struct_brain.nii.gz -cm hot

And a command like this could be used to generate a summary of functional->structural registration (the values passed to -t may need to be adjusted for different datasets):

fsleyes render -of reg.png -slightbox -zr 0 1 reg/highres.nii.gz reg/example_func2highres.nii.gz -ot mask -o -t 2500 7500 -mc 1 0 0

Details on all of the command-line options are printed when you run fsleyes --fullhelp.

1 Like