6.2.1. Functional and anatomical coregistration

Standard functional preprocessing and registration of functional image to the anatomical.

6.2.1.1. Retrieve data

from sammba import data_fetchers

retest = data_fetchers.fetch_zurich_test_retest(subjects=[0],
                                                correct_headers=True)

retest contains paths to images and data description

anat_filename = retest.anat[0]
func_filename = retest.func[0]
print(func_filename)

Out:

/home/salma/nilearn_data/zurich_retest/baseline/1366/rsfMRI_corrected.nii.gz

We use the Coregistrator, which coregisters the anatomical to a given modality

from sammba.registration import Coregistrator

coregistrator = Coregistrator(output_dir='animal_1366', brain_volume=400,
                              use_rats_tool=False, caching=True)
print(coregistrator)

Out:

Coregistrator(brain_volume=400, caching=True, clipping_fraction=0.2,
              output_dir='animal_1366', use_rats_tool=False, verbose=True)

Coregistrator comes with a parameter clipping_fraction=.2 which sometimes needs to be changed to get a good brain mask. You can check how this parameter impacts the brain segmentation

from sammba.segmentation import brain_extraction_report

print(brain_extraction_report(anat_filename, brain_volume=400,
                              clipping_fractions=[.1, .2, .9, None],
                              use_rats_tool=False))

Out:

AP length    RL width   IS height    symmetry      volume

fraction 0.10        13.50        9.70        6.10        0.90      373.50
fraction 0.20        12.90        9.70        6.20        0.91      384.19
fraction 0.90        19.00       14.00        7.00        1.00     1862.00
  no fraction        19.00       14.00        7.00        1.00     1862.00

6.2.1.2. Anatomical to functional registration

coregistrator.fit_anat(anat_filename)
coregistrator.fit_modality(func_filename, 'func', t_r=1.,
                           prior_rigid_body_registration=True)

The paths to the registered functional and anatomical images are accessible through the coregistrator attributes

registered_func_filename = coregistrator.undistorted_func_
registered_anat_filename = coregistrator.anat_in_func_space_

6.2.1.3. Check out the results

from nilearn import plotting, image

display = plotting.plot_epi(image.mean_img(registered_func_filename),
                            title='coreg anat edges on top of mean coreg EPI')
display.add_edges(registered_anat_filename)
plotting.show()

Total running time of the script: ( 0 minutes 50.045 seconds)