Autoreject: Automated artifact rejection for MEG and EEG data

By Mainak Jas, Denis-Alexander Engemann, Yousra Bekhti, Federico Raimondo, Alexandre Gramfort. NeuroImage, 2016

526 citations recorded by the source corpus.

Abstract We present an automated algorithm for unified rejection and repair of bad trials in magnetoencephalography (MEG) and electroencephalography (EEG) signals. Our method capitalizes on cross‐validation in conjunction with a robust evaluation metric to estimate the optimal peak‐to‐peak threshold – a quantity commonly used for identifying bad trials in M/EEG. This approach is then extended to a more sophisticated algorithm which estimates this threshold for each sensor yielding trial‐wise bad sensors. Depending on the number of bad sensors, the trial is then repaired by interpolation or by excluding it from subsequent analysis. All steps of the algorithm are fully automated thus lending itself to the name Autoreject. In order to assess the practical significance of the algorithm, we conducted extensive validation and comparisons with state‐of‐the‐art methods on four public datasets containing MEG and EEG recordings from more than 200 subjects. The comparisons include purely qualitative efforts as well as quantitatively benchmarking against human supervised and semi‐automated preprocessing pipelines. The algorithm allowed us to automate the preprocessing of MEG data from the Human Connectome Project (HCP) going up to the computation of the evoked responses. The automated nature of our method minimizes the burden of human inspection, hence supporting scalability and reliability demanded by data analysis in modern neuroscience. Graphical abstract Symbol. No caption available. HighlightsA strategy for artifact rejection in M/EEG using peak‐to‐peak thresholds is proposedThe thresholds are estimated using cross‐validation with a robust error metricThe method detects and repairs outlier data segments for each sensorComparison with competing methods on 200 subjects with ground truth responses

Subfields: Cognitive Neuroscience, Radiology, Nuclear Medicine and Imaging

View Paper: https://www.semanticscholar.org/paper/9449821b8202cfe0797c724d9e770834b7e0518f

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Cognitive Neuroscience

In Radiology, Nuclear Medicine and Imaging

Citations in the corpus

Deep learning-based electroencephalography analysis: a systematic review (2019, background): “First, EEG has a low signalto-noise ratio (SNR) [23, 86], as the brain activity measured is often buried under multiple sources of environmental, physiological and activity-specific noise of similar or greater amplitude called ‘artifacts’.”
Data Augmentation for Deep-Learning-Based Electroencephalography. (2020, background): “First, it is an aggregate signal emanating from the aggregated neuronal activity of millions or more cells, which has been transduced through several layers of tissue, fluid, bone, etc. EEG also suffers from low signal-to-noise ratio (SNR) [1-4].”
Recurrence is required to capture the representational dynamics of the human visual system (2019, methodology): “Trials with excessive noise were discarded by means of the autoreject toolbox (46).”
Evaluation of Artifact Subspace Reconstruction for Automatic Artifact Components Removal in Multi-Channel EEG Recordings (2020, methodology): “To automate the artifact removal process, earlier methods have used channel-based statistical thresholding approaches to remove abnormal activities [7], [8] or adaptive filters with additional reference channels to regress out targeted artifacts [9], [10].”
Parameterizing neural power spectra into periodic and aperiodic components (2020, methodology): “The AutoReject procedure was used to estimate thresholds and automatically reject any trials with artifacts, as well as to interpolate bad channels 61 .”
Evaluation of Artifact Subspace Reconstruction for Automatic EEG Artifact Removal (2018, methodology): “To automate the artifact removal process, earlier methods used channel-based statistical-thresholding approaches to remove abnormal activities [5] or adaptive filters with additional reference channels to regress out targeted artifacts”
Altered network connectivity and global efficiency in tourette syndrome: insights into sensorimotor integration (2025, unknown): “The raw EEG data were preprocessed using MNE-Python (v0.”
Is Limited Participant Diversity Impeding EEG-based Machine Learning? (2025, unknown): “There is a rich literature on EEG preprocessing [39, 40, 41], although most methods were designed for classical EEG analysis and it is unclear how well they translate to ML-based EEG applications.”
Electrophysiological decoding captures the temporal trajectory of face categorization in infants (2024, unknown): “Artifact correction was conducted using the autoreject package (Jas et al., 2017, 2016).”
Resting-state EEG correlates of sustained attention in healthy ageing: Cross-sectional findings from the LEISURE study. (2024, methodology): “Epochs were cleaned of artefacts in all channels using the Python package Autoreject (AR; Jas et al., 2017, 2016).”
RELAX‐Jr: An Automated Pre‐Processing Pipeline for Developmental EEG Recordings (2024, unknown): “Manual artefact reduction also requires considerable operator expertise in order to accurately interpret the EEG signal, making it vulnerable to potential bias and inconsistency resulting from human influence (Fitzgibbon et al. 2007; Jas et al. 2017).”
Task‐specific topology of brain networks supporting working memory and inhibition (2024, methodology): “The EEG data were divided into 1-s consequent nonoverlapping segments for the resting state and 800 ms segments for the Flanker Segments containing artifacts were rejected using autoreject algorithm (Jas et al., 2017).”