3. Registering several MRI images and creating a template

3.1. Registering several raw MRI images to common space

The anats_to_common pipeline implements the multi-level, iterative scheme to create a fine anatomical template from individual anatomical MRI scans. First, head images are centered on their respective brain mask centroid while intensity biais corrected is performed. A first rough template is obtained by rigid-body aligning extracted brain to a digitized version of a previous histological atlas and applying the transform to the original heads. Thereafter, sucessive averaging and registration process is iterated while increasing the number of degrees of freedom of the estimated transform. The target template updated at each step allowing the creation of high quality templates.

In sammba-MRI, you can call the function by typing

from sammba import registration
anats_to_common(Images_to_register, Saved_Registered_dir, 400, caching=True)

Note that most of registration steps relies on the freely available AFNI software. For skull-stripping, the open source RATS tool is also available by setting the parameter use_rats_tool to True.

An example of use this function is available called Template creation

3.2. Registering several MRI images to template space

Matching of individual MRI image to standard template is critical to perform group-wise analysis. Multimodal images preprocessing can be performed in the template space through the TemplateRegistrator class. User can fit each anatomical images to a a set of modality (functional or perfusion MRI). This ready-to-use pipelines align first the functional or perfusion volume to the anatomical images through a rigid body registration. Then, a per-slice basis registration is performed allowing correction of EPI distortion. Finally, the rigid body transform and the per-slice warps are combined and applied to the initial volume to minimize interpolation errors.

In sammba-MRI, the class can be call by typing

from sammba import registration
registrator = TemplateRegistrator(template, brain_volume, caching=True)
registrator.fit_anat(anat_files)
registrator.fit_modality(moadality_images, name_of_the_modality

An example of use this function is available.

Note that TemplateRegistrator class encapsulate several functions that are available as “stand-alone” in sammba-MRI:

  • anats_to_template : Allow to registering several MRI images to an existing template

    anats_to_template(images_to_register, Template_name, Saved_Registered_dir, 400, caching=True)

  • fmri_sessions_to_template : Allow to register several functional and anatomic MRI images to an existing template

    fmri_sessions_to_template(session, Atlas_name,Saved_Registered_dir, maxlev=7, t_r=1.0, brain_volume=400)

  • FMRISession : Encapsulate the data for performing registration of several functional MRI images to an existing template and EPI correction

    FMRISession(Anats_to_register, Funcs_to_register, Saved_Registered_dir)

  • coregister_fmri_session : process the data for performing registration of several functional MRI images to an existing template and EPI correction

    coregister_fmri_session(session, t_r=1.0, Saved_Registered_dir, brain_volume=400, slice_timing=True)