Introducing RELAX (the Reduction of Electroencephalographic Artifacts): A fully automated pre-processing pipeline for cleaning EEG data - Part 1: Algorithm and Application to Oscillations

Abstract

Electroencephalographic (EEG) data is typically contaminated with non-neural artifacts which can confound the results of experiments. Artifact cleaning approaches are available, but often require time-consuming manual input and significant expertise. Advancements in artifact cleaning often only address a single artifact, are only compared against a small selection of pre-existing methods, and seldom assess whether a proposed advancement improves experimental outcomes. To address these issues, we developed RELAX (the Reduction of Electroencephalographic Artifacts), an automated EEG cleaning pipeline implemented within EEGLAB that reduces all artifact types. RELAX cleans continuous data using Multiple Wiener filtering [MWF] and/or wavelet enhanced independent component analysis [wICA] applied to artifacts identified by ICLabel [wICA_ICLabel]). Several versions of RELAX were tested using three datasets containing a mix of cognitive and resting recordings (N = 213, 60 and 23 respectively). RELAX was compared against six commonly used EEG cleaning approaches across a wide range of artifact cleaning quality metrics, including signal-to-error and artifact-to-residue ratios, measures of remaining blink and muscle activity, and the amount of variance explained by experimental manipulations after cleaning. RELAX with MWF and wICA_ICLabel showed amongst the best performance for cleaning blink and muscle artifacts while still preserving neural signal. RELAX with wICA_ICLabel (and no MWF) may perform better at detecting the effect of experimental manipulations on alpha oscillations in working memory tasks. The pipeline is easy to implement in MATLAB and freely available on GitHub. Given its high cleaning performance, objectivity, and ease of use, we recommend RELAX for data cleaning across EEG studies.

Right Anterior Theta Connectivity Predicts Autistic Social Traits in Neurotypical Children

“…In the present study, only the eyes-closed data were analysed so as to reduce the occurrence of electrooculography artifacts in the EEG record, as well as to remain consistent with Aykan et al [7] who also reported results obtained from eyes-closed resting-state data. The EEG data were pre-processed in MATLAB (R2020a, The Mathworks, Massachusetts, USA) using an automated cleaning pipeline (Reduction of Electrophysiological Artifacts [RELAX]) [16]. For further details on data pre-processing see [17].…”

Meditators probably show increased behaviour-monitoring related neural activity

“…Epochs with voltages exceeding +/-60 μV at any electrode were rejected, as were epochs containing improbable voltage distributions or kurtosis values >5SD from the mean in any single electrode or more than 3SD from the mean over all electrodes. Data were then baseline corrected by regressing out the average of the -400 to -100ms period from each epoch (timelocked to the response) using the fieldtrip function 'ft_regressconfounds' for each electrode and each participant separately, with the condition of each epoch (correct or error response) included in the regression model (but not rejected) to correct for potential voltage drift but still preserve any experimental effects (Bailey et al, 2022a(Bailey et al, , 2022b.…”

Introducing RELAX (the Reduction of Electroencephalographic Artifacts): A fully automated pre-processing pipeline for cleaning EEG data - Part 2: Application to Event-Related Potentials

“…To address these issues, we built the “RELAX” EEG cleaning pipeline (short for “Reduction of Electroencephalographic Artifacts”), in which we combined and adjusted pre-existing approaches to optimise EEG cleaning and ERP outcomes. The pipeline is explained in detail in our companion article (Bailey et al, 2022), but in brief, RELAX: 1) Removes bad electrodes and extreme outlying EEG periods that are unlikely to have recorded meaningful or recoverable brain activity using a combination of algorithms obtained from previous research so that the rejected data matched our expert judgement for all common extreme artifact types; 2) The pipeline then reduces artifacts with Multi-channel Wiener Filters (MWF) (Borowicz, 2018; Somers et al, 2018); and 3) RELAX further reduces any remaining artifacts using wavelet enhanced ICA (wICA) (Castellanos & Makarov, 2006) applied to artifact components identified by the machine learning algorithm ‘ICLabel’ (Pion-Tonachini et al, 2019). This combination approach effectively cleans blinks, muscle activity, horizontal eye movement, voltage drift, and atypical artifacts to produce more dependable ERP estimates that indicate increased signal-to-noise ratios.…”

“…A condition of use of the pipeline is that the version of the pipeline used is referred to as RELAX_[pipeline], for example "RELAX_MWF_wICA" or "RELAX_wICA_ICLabel", and that the current paper be cited, as well as the dependencies used. These dependencies are likely to include: EEGLAB (Delorme & Makeig, 2004), fieldtrip (Oostenveld et al, 2011), the MWF toolbox (Somers et al, 2019), fastICA (Hyvarinen, 1999), wICA (Castellanos & Makarov, 2006), ICLabel (Pion-Tonachini et al, 2019), and PREP (Bigdely-Shamlo et al, 2015) See our companion article for the application of RELAX to the study of oscillatory power (Bailey et al, 2022).…”

“…See our companion article for the application of RELAX to the study of oscillatory power (Bailey et al, 2022).…”