An fMRI preprocessing pipeline

This is an automatic fMRI spatial preprocessing pipeline based on the SPM12 and CAT12 that can be used for both task-based and resting-state fMRI data. It consists of the following steps:

• Susceptibility distortion correction
  • VDM calculation based on the fieldmap phase and amplitude data.
• Realignment and Unwarp
  • Simultaneous realignment of functional volumes and applying VDM.
• Segmentation of the structural data
  • It also creates a bias-corrected T1 image, normalizes it and all the extracted tissue masks to the MNI space, and outputs the forward and inverse transformation parameters.
• Coregistration of the functional data into the T1 image.
  • It aligns the functional data to the bias-corrected T1 image.
• Normalization to the MNI space
  • It uses the forward transformation parameters of the segmentation step, to normalize the functional data into the MNI space.
• Smoothing

I tried to make this pipeline as automatic as possible. You just need to set some directories and acquisition parameters, run the pipeline, and sit and drink your coffee!

Useful info:

• I used CAT12 for the segmentation as it gives more precise results than the SPM.
• Make sure that before running the pipeline, you have downloaded the Artifact Detection Tools (ART) and have added its directory to the Matlab paths. Also make sure that you have downloaded the CAT12 package and have put it in the SPM toolbox folder.
• The pipeline considers that your data is organized in the BIDS format as follows:

  Data_dir -->
              sub-XXX
              sub-YYY
              .
              .
              .
              sub-ZZZ -->
                         ses-xxx
                         ses-yyy
                         .
                         .
                         .
                         ses-zzz -->
                                    anat -->
                                            sub-ZZZ_ses-zzz_T1w.json
                                            sub-ZZZ_ses-zzz_T1w.nii
                                    fmap -->
                                            sub-ZZZ_ses-zzz_task-<TaskName>_magnitude1.json
                                            sub-ZZZ_ses-zzz_task-<TaskName>_magnitude1.nii
                                            sub-ZZZ_ses-zzz_task-<TaskName>_phasediff.json
                                            sub-ZZZ_ses-zzz_task-<TaskName>_phasediff.nii
                                    func -->
                                            sub-ZZZ_ses-zzz_task-<TaskName>_bold.json
                                            sub-ZZZ_ses-zzz_task-<TaskName>_bold.nii
  

• To run the pipeline, open the preprocessing.m file, set the requested directories and parameters, and run the file.
• If you need to change the hyperparmeters of different preprocessing steps, open the func_PreprocBatch.m file, and set the parameters with your own values accordingly.
• I have tested the pipeline on macOS Sonoma Version 14.5 and it works without any error.

Do not hesitate to send your feedback to improve this pipeline!