BrainHack-Physio

• Physiological Noise Correction using the PhysIO Toolbox - Examples for the BrainHack 23, Toronto
• Project Proposal, with links, references: https://brainhack.org/global2023/project/project_71/
• Chat to us on MatterMost: https://mattermost.brainhack.org/brainhack/channels/physio

Purpose

This small repository sets up a simple preprocessing (realignment, smoothing) and single subject fMRI analysis using the modeled physiological noise.

Installation

1. You will need to install SPM12
2. You will need to install the PhysIO Toolbox
   • The PhysIO/code folder should end up in spm12/toolbox/PhysIO
3. You will need data from the following publication (Brainstem fMRI, native Siemens physio logfiles):
   • Paper (for background info): https://doi.org/10.1016/j.celrep.2023.113405
   • Physiological and behavioral logfiles on OSF: https://osf.io/td5kp/
   • fMRI data on OpenNeuro/DataLad: https://github.com/OpenNeuroDatasets/ds004808/

Getting Started

1. Create the following folder structure:

   project/
           data/
                sub-46/             (copy from OpenNeuro.org fmri) and osf.io (physio))
                       func/        (functional data from OpenNeuro)
                       anat/        (structural data from OpenNeuro)
                       physio/      (physiological logfiles from osf.io)
           code/brainhack-physio/   (this repository)
   

2. Open the main script brainhack_physio_main in Matlab
3. Follow the instructions there to adapt paths to your environment
4. For each subject to try, download copy the fMRI and physiological recording data
5. Run brainhack_physio_main (play button) in Matlab.
   • For starters, the interactive mode (isInteraxtive = true) is recommended
   • In this mode, each step is displayed in the SPM Batch Editor, and you have to press the play button (and wait) to execute
   • After the step finishes, press Enter in the command window

The final step executed by this script will be the visualization of the physiological regressors. For the F-map of all RETROICOR (physiological phase) and movement regressors, this should look like the following screenshot for a single subject (sub-46):