Time | Topic | Presenter |
---|---|---|
9:00 - 9:15 | Bagel Config + Coffee Download + Computer Setup | |
9:15 - 9:30 | Course Intro + MEGcore Intro | MEGcore |
9:30 - 10:30 | Intro to MEG and general overview of source localization | Fred |
10:30 - 11:00 | MEG Hardware and Signal Generation / Collection | Stephen |
11:00 - 11:15 | Break | |
11:15 - 12:15 | Stimuli, trigger processing, epochs, evoked data, bad chans | Tom + Anna |
bad epochs, filtering, abberant signals | ||
12:15 - 1:30 | Lab 1 + Lunch | |
1:30 - 2:30 | Filtering Data, Frequency Analysis, Hilbert, Brain Rythms, and Artifacts | Allison |
2:30 - 3:00 | Lab 2 | |
3:00 - 4:00 | MRI Processing, placing fiducials, coreg, source model (volume + surface) | Anna + Jeff |
BEM, Forward Model | ||
4:00 - 4:30 | Lab 3 |
Time | Topic | Presenter |
---|---|---|
9:30-10:30 | Source Localization (Dipoles, multiple dipoles, MNE, dSPM, Beamformer) | Fred + Jeff |
10:30 - 11:30 | Lab 4 | |
11:30 - 12:00 | Beamformer Specifics - Covariance, filtering, data rank, regularization | Tom |
12:00 - 1:30 | Lab 5 + Lunch | |
1:30 - 2:00 | Single subject to group data | Jeff |
2:00 - 2:30 | Lab 6 | |
2:30 - 3:30 | Statistics (parametric / log transform / resampling stats / clusters ) | Fred + Allison |
3:30 - 4:00 | Code | |
4:00 - 4:20 | Course Review | Allison |
4:20 - 5:00 | Additional Techniques: OPMs, Connectivity | Amaia + Lucrezia |
Additional Techniques: Decoding, Dynamic Causal Modeling | Shruti + Jess |
Log into biowulf: ssh -Y [email protected]
#You can type tmux before starting sinteractive to have a persistent session between disconnecting wifi
#Allocate resources for processing
sinteractive --mem=16G --cpus-per-task=12 --gres=lscratch:10 --tunnel --time=08:00:00
You will see a line like the below. Follow the instructions (start a new terminal into biowulf), then return to original terminal for the rest of the commands.
Copy notebooks to your local folder. Change directories first if you don't want the code/data in your home folder.
module use --append /data/MEGmodules/modulefiles #You can add this to your .bashrc for convenience
module load meg_workshop
get_code #Copy the code to your current directory
get_data #Copy and untar the data to your /data/${USER}/meg_data_workshop
cd MEG_workshop_2023
./start_notebook_Day1.sh #Start the notebook for Day1 - allows for time series scrolling
#OR use for Day2 material -- ./start_notebook_Day2.sh - Visualize the 3D brain renderings
Enter this into the address bar of your web browser localhost:<PORT>
NOTE:If you get something about a token: Copy it from the commandline
The following software is required to run all parts of the coding sections: afni + freesurfer + git + miniconda(/conda)
To run the majority of the code: miniconda/conda + git are required
Miniconda will provide the minimum features for the installation:
https://docs.conda.io/projects/miniconda/en/latest/
In a terminal, find your Download folder (typically cd /home/<USERNAME>/Downloads
or cd /Users/<USERNAME>/Downloads
).
chmod +x Miniconda3-latest.....sh #Make this file executable - Fill in the rest of name (it will be Linux / Mac/ or Windows)
./Miniconda3-latest...sh #Run the installer. Open a new terminal after finishing the installation directions
Mamba is not required, but will install faster than conda (functionally they are the same)
To install mamba - conda install --channel=conda-forge --name=base mamba
git clone https://github.com/nih-megcore/MEG_workshop_2023.git
cd MEG_workshop_2023
make install
#Clone this repository - If you don't have git, just download the zip file from the green button at the top of page
git clone https://github.com/nih-megcore/MEG_workshop_2023.git
#Install MNE - Substitute conda for mamba if any errors
mamba create --override-channels --channel=conda-forge --name=MEG_workshop mne pip jupyterlab -y
conda activate MEG_workshop
pip install h5io pymatreader
#Install the Workshop files
cd MEG_workshop_2023
pip install -e . #Install this code
pip install git+https://github.com/nih-megcore/nih_to_mne.git #Install some auxilliary NIH code
Install dataset - if you want to create from scratch. Use the link provided by email if you want to download
mamba create -n datalad -c conda-forge datalad gdown -y
conda activate datalad
This will pull the NIMH_hv OpenNeuro repository and associated files.
This will download the freesurfer processed MRI files as well
./extras/datalad_pull.sh
Install freesurfer: https://surfer.nmr.mgh.harvard.edu/fswiki/rel7downloads
Install Afni: https://afni.nimh.nih.gov/pub/dist/doc/htmldoc/background_install/main_toc.html
Run ./check_requirements.sh
to check for freesurfer / afni / mne / jupyter installation.
Input data is in BIDS format from the NIH HV protocol.
Data should be located in /data/${USER}/meg_workshop_data
This will automatically be performed on biowulf with the get_data
command.
Derivatives data will be in /data/${USER}/meg_workshop_data/{Day1,Day2}/${bids_id}/ses-01/meg/
Day2 derivatives will have pre-calculated bem, fwd, trans, src files