MEG_Analysis

Collection of Python and Matlab Scripts to run the MEG data analysis for Mars2

Pipeline requires:

• MNE Python: https://mne.tools/stable/index.html
• FieldTrip: https://www.fieldtriptoolbox.org/
• VirtualTools: https://github.com/juliankeil/VirtualTools

Preprocessing: MNE Python

1. Step 1: Convert Raw MEG files to BIDS format:
   1. Run MNE_Raw2Bids.py to read in the raw MEG data and format for the automatic processing
2. Step 2: Run the automatic BIDS pipeline: Attention, need the fine MEG configuration files
   1. Go to the command line, go to the root folder of the BIDS data and run (watch out, this needs the fine configuration files for the MEG system!): mne_bids_pipeline --config=MNE_BIDS_config.py --steps=preprocessing
      1. Find Noisy or Flat Channels
      2. Maxwell Filter
      3. Bandpass and Notch Filter
      4. ICA
      5. Epoch
      6. Condition Selection
3. Step 3: Automatic rejection of bad channels and epochs
   1. Run MNE_BIDS_autoreject.py

TFR and Source Analysis: FieldTrip

1. Step 1: Read in MNE BIDS Data
   1. Run CamCan_Import.m to select MEG channels and prewhiten the MEG data to have the same scale between magnetometers and gradiometers
2. Step 2: Create Headmodel
   1. Read in the Headmodel in MRIcoGL to double check orientation
   2. Run CamCan_Headmodel.m to build headmodel and sourcemodel. Watch out, this requires some user input to re-orient the MRI and MEG data in space
3. Compute TFR in Sensor Space
   1. Run CamCan_TFR.m to compute TFR
   2. Stats and Plotting in the same script
4. Project data to virtual channels in source space
   1. Run CamCan_VirtChan.m for Virtual Sensor Projection
5. Compute TFR in Source Space
   1. Run CamCan_TFR_vc.m to compute TFR
   2. Stats and Plotting in the same script
6. Compute Connectivity in Source Space
   1. Run CamCan_Connectivity.m to compute Functional Connectivity
   2. Stats and Plotting in the same script