chriscline / aarateppipeline

Advanced artifact removal for automated TMS-EEG processing

AARATEP Pipeline

This repo contains the code for the TMS-EEG preprocessing pipeline originally described in:

C.C. Cline, M.V. Lucas, Y. Sun, M. Menezes, A. Etkin. "Advanced Artifact Removal for Automated TMS-EEG Data Processing," 2021 10th International IEEE/EMBS Conference on Neural Engineering (NER), 2021, doi: 10.1109/NER49283.2021.9441147.

In brief, this pipeline consists of the following stages:

• Epoching
• Artifact interpolation (with custom autoregressive blending)
• Downsampling
• Baseline correction
• High-pass filtering
• Bad-channel identification
• Early eye-related IC rejection (added in v2.0.0)
• SOUND
• Decay component removal
• Artifact interpolation
• Line noise filtering
• ICA
• IC rejection with ICLabel and additional TMS-specific rejection rules
• Low-pass filtering
• Average rereferencing

Usage

This code assumes you have EEGLab installed at ~/Documents/MATLAB/eeglab on Windows or Mac, or ~/matlab/eeglab on Linux. It also assumes you have installed the ICLabel and TESA extensions in EEGLab.

Assuming you have downloaded this whole repo to a folder called AARATEPPipeline, add dependencies to your MATLAB path with

addpath('AARATEPPipeline');
addpath('AARATEPPipeline/Common');
addpath('AARATEPPipeline/Common/EEGAnalysisCode');

Load your data as an EEGLab struct EEG. For example:

[EEG, misc] = c_TMSEEG_prepareForPreprocessing(...
    'inputFilePath', 'MyStudy/rawdata/RecordingName.vhdr',...
    'epochTimespan', [-1 2]);

Then call the main preprocessing pipeline script:

EEG = c_TMSEEG_Preprocess_AARATEPPipeline(EEG,...
    'pulseEvent', misc.pulseEvent,...
    'epochTimespan', misc.epochTimespan,...
    'outputDir', 'MyStudy/derivatives/RecordingName',...
    'outputFilePrefix', 'RecordingName')

See individual scripts for additional available parameters.