A new ICA-based fingerprint method for the automatic removal of physiological artifacts from EEG recordings

Abstract

BackgroundEEG may be affected by artefacts hindering the analysis of brain signals. Data-driven methods like independent component analysis (ICA) are successful approaches to remove artefacts from the EEG. However, the ICA-based methods developed so far are often affected by limitations, such as: the need for visual inspection of the separated independent components (subjectivity problem) and, in some cases, for the independent and simultaneous recording of the inspected artefacts to identify the artefactual independent components; a potentially heavy manipulation of the EEG signals; the use of linear classification methods; the use of simulated artefacts to validate the methods; no testing in dry electrode or high-density EEG datasets; applications limited to specific conditions and electrode layouts.MethodsOur fingerprint method automatically identifies EEG ICs containing eyeblinks, eye movements, myogenic artefacts and cardiac interference by evaluating 14 temporal, spatial, spectral, and statistical features composing the IC fingerprint. Sixty-two real EEG datasets containing cued artefacts are recorded with wet and dry electrodes (128 wet and 97 dry channels). For each artefact, 10 nonlinear SVM classifiers are trained on fingerprints of expert-classified ICs. Training groups include randomly chosen wet and dry datasets decomposed in 80 ICs. The classifiers are tested on the IC-fingerprints of different datasets decomposed into 20, 50, or 80 ICs. The SVM performance is assessed in terms of accuracy, False Omission Rate (FOR), Hit Rate (HR), False Alarm Rate (FAR), and sensitivity (p). For each artefact, the quality of the artefact-free EEG reconstructed using the classification of the best SVM is assessed by visual inspection and SNR.ResultsThe best SVM classifier for each artefact type achieved average accuracy of 1 (eyeblink), 0.98 (cardiac interference), and 0.97 (eye movement and myogenic artefact). Average classification sensitivity (p) was 1 (eyeblink), 0.997 (myogenic artefact), 0.98 (eye movement), and 0.48 (cardiac interference). Average artefact reduction ranged from a maximum of 82% for eyeblinks to a minimum of 33% for cardiac interference, depending on the effectiveness of the proposed method and the amplitude of the removed artefact. The performance of the SVM classifiers did not depend on the electrode type, whereas it was better for lower decomposition levels (50 and 20 ICs).DiscussionApart from cardiac interference, SVM performance and average artefact reduction indicate that the fingerprint method has an excellent overall performance in the automatic detection of eyeblinks, eye movements and myogenic artefacts, which is comparable to that of existing methods. Being also independent from simultaneous artefact recording, electrode number, type and layout, and decomposition level, the proposed fingerprint method can have useful applications in clinical and experimental EEG settings.

ICLabel: An automated electroencephalographic independent component classifier, dataset, and website

“…Several other attempts to automatically solve the IC classification problem have been made publicly available. A recent and largely comprehensive summary of those methods can be found in the introduction of Tamburro et al (2018). For our purposes, we only consider and compare methods and their supporting algorithms that are (1) publicly available, (2) do not require any information beyond the ICA-decomposed EEG recordings and generally available meta-data such as electrode locations, and (3) have at minimum a category for Brain ICs as defined in Section 2.1.…”

Automatic Removal of Physiological Artifacts in EEG: The Optimized Fingerprint Method for Sports Science Applications

“…Retained datasets were pre-whitened by Principal Components Analysis (PCA; Delorme et al, 2007 ) and decomposed into 20, 50, or 80 ICs using the extended Infomax ICA algorithm (Bell and Sejnowski, 1995 ; Lee et al, 1999 ). These decomposition levels were selected to mimic the most common clinical and experimental EEG conditions: 21 electrodes are typically used in a clinical setting, whereas commercial EEG caps for research purposes generally mount from 32 to 128 electrodes or more (Tamburro et al, 2018 ). For each EEG dataset, we retained for further analysis only the sets of separated ICs that contained clearly identifiable and non-redundant ICs.…”

“…Fourteen different features were calculated for each IC from all datasets. A complete description of all features, including the calculation of each feature and the values of all parameters, is provided in Tamburro et al ( 2018 ). We briefly describe the calculation of each feature below.…”

“…We recently developed a system for the automatic classification and removal of the most ubiquitous physiological artifacts–eyeblinks, eye movements, and muscle and cardiac artifacts (Tamburro et al, 2018 ). Our method, the Fingerprint Method, was developed from both existing methods and novel approaches.…”

A Bayesian framework for unifying data cleaning, source separation and imaging of electroencephalographic signals

“…The practical use of ICA has been limited by its computational cost and the need for user intervention. Only recently, a real-time recursive ICA algorithm has been proposed (Hsu et al, 2016), as well as a number of automatic methods for minimizing the subjectivity of manual component selection (Tamburro et al, 2018;Pion-Tonachini et al, 2017;Radüntz et al, 2017). Despite these advances, turning ICA into a brain imaging modality requires that after source separation, we solve the inverse problem of localizing the set of identified brain components into the cortical space.…”