Article

Technical Note Enhanced detection of artifacts in EEG data using higher-order statistics and independent component analysis

Computational Neurobiology Laboratory, Salk Institute for Biological Studies, 10010 N. Torrey Pines Road, La Jolla, CA 92107, USA.
NeuroImage (Impact Factor: 6.36). 03/2007; 34(4):1443-9. DOI: 10.1016/j.neuroimage.2006.11.004
Source: PubMed

ABSTRACT

Detecting artifacts produced in EEG data by muscle activity, eye blinks and electrical noise is a common and important problem in EEG research. It is now widely accepted that independent component analysis (ICA) may be a useful tool for isolating artifacts and/or cortical processes from electroencephalographic (EEG) data. We present results of simulations demonstrating that ICA decomposition, here tested using three popular ICA algorithms, Infomax, SOBI, and FastICA, can allow more sensitive automated detection of small non-brain artifacts than applying the same detection methods directly to the scalp channel data. We tested the upper bound performance of five methods for detecting various types of artifacts by separately optimizing and then applying them to artifact-free EEG data into which we had added simulated artifacts of several types, ranging in size from thirty times smaller (-50 dB) to the size of the EEG data themselves (0 dB). Of the methods tested, those involving spectral thresholding were most sensitive. Except for muscle artifact detection where we found no gain of using ICA, all methods proved more sensitive when applied to the ICA-decomposed data than applied to the raw scalp data: the mean performance for ICA was higher and situated at about two standard deviations away from the performance distribution obtained on raw data. We note that ICA decomposition also allows simple subtraction of artifacts accounted for by single independent components, and/or separate and direct examination of the decomposed non-artifact processes themselves.

Download full-text

Full-text

Available from: Arnaud Delorme
  • Source
    • "If the waveform under analysis has many similar values, or there are two extremes between which the values alternate, the kurtosis will be highly negative. Artefacts produced by direct currents or alternating currents show high negative kurtosis; those produced due to movement such as eye blinks return high positive kurtosis (Delorme et al., 2007). Delorme and his group found that kurtosis was partly successful in the detection of large discontinuities in the data, and trend artefacts; however it was found that other methods for artefact detection performed better than kurtosis measures. "
    [Show description] [Hide description]
    DESCRIPTION: MSci project (2014/2015) including the development of algorithms and a GUI to analyse brain EEG activity in epilepsy. The algorithms investigated were Rank Vector Entropy, Cross Correlation, Variance, Skewness, Kurtosis and Continuous Wavelet Transforms. This research was conducted with Caitlin O'Brien but this paper written by Emma Parker (nee Warren).
    Full-text · Research · Jan 2016
  • Source
    • "Two recent studies have indicated that movement artifact can lead to high levels of spectral power, especially at very low and very high frequencies, during double support (Castermans et al., 2014; Kline et al., 2015). ICA has proven very effective for separating eye and muscle artifacts from EEG electrocortical signals during seated or standing tasks (Jung et al., 2000; Delorme et al., 2007). Algorithms that model independent components as equivalent current dipoles, such as DIPFIT, have also been shown to be able to accurately localize the resultant neural sources (Oostenveld and Oostendorp, 2002). "
    [Show abstract] [Hide abstract]
    ABSTRACT: There has been a recent surge in the use of electroencephalography (EEG) as a tool for mobile brain imaging due to its portability and fine time resolution. When EEG is combined with independent component analysis (ICA) and source localization techniques, it can model electrocortical activity as arising from temporally independent signals located in spatially distinct cortical areas. However, for mobile tasks, it is not clear how movement artifacts influence ICA and source localization. We devised a novel method to collect pure movement artifact data (devoid of any electrophysiological signals) with a 256-channel EEG system. We first blocked true electrocortical activity using a silicone swim cap. Over the silicone layer, we placed a simulated scalp with electrical properties similar to real human scalp. We collected EEG movement artifact signals from ten healthy, young subjects wearing this setup as they walked on a treadmill at speeds from 0.4-1.6 m/s. We performed ICA and dipole fitting on the EEG movement artifact data to quantify how accurately these methods would identify the artifact signals as non-neural. ICA and dipole fitting accurately localized 99% of the independent components in non-neural locations or lacked dipolar characteristics. The remaining 1% of sources had locations within the brain volume and low residual variances, but had topographical maps, power spectra, time courses, and event related spectral perturbations typical of non-neural sources. Caution should be exercised when interpreting ICA for data that includes semi-periodic artifacts including artifact arising from human walking. Alternative methods are needed for the identification and separation of movement artifact in mobile EEG signals, especially methods that can be performed in real time. Separating true brain signals from motion artifact could clear the way for EEG brain computer interfaces for assistance during mobile activities, such as walking.
    Full-text · Article · Dec 2015 · Frontiers in Human Neuroscience
  • Source
    • "A simulation study was performed to examine the performance of EEMD-ICA in comparison with two dominant approaches to artifact rejection, i.e., the classic ICA and AWICA. The classical ICA separated artifacts and concentrated them into the corresponding independent components for rejection afterwards [3]. AWICA first used wavelet transform to decompose raw data of each channel into multiple frequency sub-bands, and it then applied ICA to the decomposed data to separate the artifacts for rejection at the end [10]. "
    [Show abstract] [Hide abstract]
    ABSTRACT: As neural data are generally noisy, artifact rejection is crucial for data preprocessing. It has long been a grand research challenge for an approach which is able: 1) to remove the artifacts and 2) to avoid loss or disruption of the structural information at the same time, thus the risk of introducing bias to data interpretation may be minimized. In this study, an approach (namely EEMD-ICA) was proposed to first decompose multivariate neural data that are possibly noisy into intrinsic mode functions (IMFs) using ensemble empirical mode decomposition (EEMD). Independent component analysis (ICA) was then applied to the IMFs to separate the artifactual components. The approach was tested against the classical ICA and the automatic wavelet ICA (AWICA) methods, which were dominant methods for artifact rejection. In order to evaluate the effectiveness of the proposed approach in handling neural data possibly with intensive noises, experiments on artifact removal were performed using semi-simulated data mixed with a variety of noises. Experimental results indicate that the proposed approach continuously outperforms the counterparts in terms of both normalized mean square error (NMSE) and Structure SIMilarity (SSIM). The superiority becomes even greater with the decrease of SNR in all cases, e.g., SSIM of the EEMD-ICA can almost double that of AWICA and triple that of ICA. To further examine the potentials of the approach in sophisticated applications, the approach together with the counterparts were used to preprocess a real-life epileptic EEG with absence seizure. Experiments were carried out with the focus on characterizing the dynamics of the data after artifact rejection, i.e., distinguishing seizure-free, pre-seizure and seizure states. Using multi-scale permutation entropy to extract feature and linear discriminant analysis for classification, the EEMD-ICA performed the best for classifying the states (87.4%, about 4.1% and 8.7% higher than that of AWICA and ICA respectively).
    Full-text · Article · Dec 2015 · IEEE Transactions on Neural Systems and Rehabilitation Engineering
Show more