Rapid CommunicationContrasting single-trial ERPs between experimental manipulations: Improving differentiability by blind source separation
Introduction
In cognitive neuroscience, the event-related potential (ERP) is a frequently used tool for revealing different brain activation patterns associated with experimental manipulations. One is often interested in differences between ERP waveforms and their corresponding scalp distributions induced by different experimental conditions. As high-density EEG systems have become increasingly available, it is not immediately obvious how to best analyze such high dimensional data. One approach has been to focus on a subset of sensors where ERP differences have been observed under similar experimental conditions. This approach is often used for experimental situations where ERP waveform differences are expected to be relatively focal, i.e., present at a small subset of electrode sites. Under these circumstances, statistical tests of ERP differences have often been performed on waveforms at a small number of sensors. For example, to capture the differential activations of motor cortices prior to the generation of hand movements, motor readiness potentials have been indexed using two recording electrodes, one located over the left motor cortex (C3) and the other over the right (C4) (Coles, 1989). For a given task, the specific sensors chosen have often been shown to best differentiate between experimental conditions, in this case, left and right hand movement.
Implicit in this sensor-based ERP approach is the assumption that the EEG sensors displaying the maximal signal of interest (from here on referred to as the “best sensors”) offer reasonable estimates of the underlying brain sources of interest and reasonable sensitivity to detect ERP differences between experimental conditions. One advantage of using the “best sensor” ERP approach is that the concept is intuitive and the method requires no special tools and is thus accessible to the majority of ERP researchers. Another advantage is that preselection of a few “best sensors” allows for more focused statistical tests as an alternative to the omnibus low-powered ANOVA test on data from all electrode locations. Similar to principal component analysis (PCA) or independent component analysis (ICA), which can be viewed as applying spatial filters to the EEG data, this “best sensor”-based approach can also be viewed as applying a specific kind of spatial filter. While in the case of PCA or ICA, the spatial filters are derived mathematically, in the case of the “best sensor” approach, the spatial filters are selected intuitively or heuristically. The effect of the “best sensor”-based filters is to retain signals from only the selected EEG sensors and exclude signals from all others. Because the scalp recorded EEG at each sensor contains a mixture of signals from both extra-cranial noise sources and intra-cranial neuronal sources from different brain locations, such filtering via the “best sensor” maintains the mixed nature of signals. Thus, in principle, the event-related activation from specific functional brain regions cannot be optimally indexed by scalp EEG sensor data, even when the signals are from the sensors with the largest ERP amplitude.
A variation on the “best sensor”-based analysis is taking an average surrounding and including the “best sensors”. Although such averaging offers an opportunity to remove uncorrelated sensor noise, it does nothing to unmix or isolate signals from specific brain sources of interest from those originating within other brain regions. It is critical to note that this averaging process introduces a major source of subjectivity and uncertainty concerning how many surrounding sensors and which ones to include. These decisions can be difficult as the scalp voltage map may easily reveal activation at a large number of the sensors (e.g., Figs. 3C, D). Finally, averaging across the “best sensor” and its surrounding sensors would only serve to reduce, instead of increase, the resulting signal-to-noise ratio because, by definition, the maximum is always greater than the average. Therefore, the “best sensor” approach offers better signal-to-noise ratio than the “average sensor” approach.
Blind source separation (BSS) and ICA algorithms (Hyvarinen and Oja, 2000, Vigario et al., 2000, Stone, 2002, Makeig et al., 2004, Muller et al., 2004) are potential methods for separating the mixture of EEG signals into multiple components. Some of the components recovered from these algorithms correspond to noise sources, such as artifacts (Vigario, 1997, Jung et al., 1999, Tang et al., 2002b, Joyce et al., 2004), and others to neuronal sources of interest (Vigario et al., 2000, Tang et al., 2002b, Makeig et al., 2004, Muller et al., 2004, Tang et al., 2005a). Second-order blind identification (SOBI: Belouchrani et al., 1993, Cardoso and Souloumiac, 1996, Belouchrani et al., 1997) is one BSS algorithm that can isolate ocular artifacts and sensor noise from brain signals (Tang et al., 2000a, Tang et al., 2002b, Tang et al., 2005a, Joyce et al., 2004) and separate functionally distinct but sometimes correlated brain activity, for example, activity from left and right primary somatosensory (SI) cortices following median nerve stimulation (Tang et al., 2005a). Most importantly, it has been shown that the signal-to-noise ratio (SNR) of averaged somatosensory evoked potentials (SEPs) obtained from SOBI-recovered SI components is much higher than the SNR obtained from the “best sensors” over the SI hand region (Tang et al., 2005a). This suggests that patterns of neuronal activations associated with different sensory, motor, or cognitive activations–induced by experimental manipulation–should be rendered more distinguishable when using SOBI-recovered components than when using the EEG “best sensor” signals directly.
We tested this hypothesis in a simple task domain where we attempted to differentiate single-trial ERPs induced by different stimulation conditions. We collected EEG data during an intermixed sequence of three different median nerve stimulations (left, right, and bilateral) and attempted to distinguish or classify, according to the single-trial SEPs, which of the three stimulations had been delivered during a given trial. If SOBI-recovered components isolate the neuronal responses from underlying sources better than the “best sensor” data, then one would expect that single-trial ERPs from the SOBI neuronal components should be classified with greater accuracy than those directly from the “best sensors”. To quantify the differentiability of single-trial ERPs, we used back-propagation neural networks (BPNNs) to classify the EEG data from each trial according to the stimulus condition presented. BPNNs are commonly used for pattern classification and have been previously used to classify single-trial ERPs (e.g., Pfurtscheller et al., 1996, Pfurtscheller et al., 1997). Here, we evaluated comparative performance between network classifiers trained with SOBI-recovered SI component data (SOBI component network) and those trained with the “best sensor” data (“best sensor” network). Specifically, we determined the percentage of single-trial SEPs that were correctly classified when the two different types of ERP representations were used.
It is important to point out that the goal of the present study was not to achieve optimal performance for classifying single-trial ERPs or to determine the best possible method for classification. The aim of the present study was to demonstrate the potential usefulness of SOBI preprocessing for ERP research relative to the use of the “best sensors” when contrasting the effects of experimental manipulations on brain activations at the level of single-trial. We tested a specific prediction that the SEPs of the SOBI-recovered components can serve as better indices for distinguishing among different patterns of brain activations associated with different sensory stimulations than the “best sensors” selected using information from all sensors. A variety of advanced signal processing methods, either alone or in combination, may result in further classification improvement.
Section snippets
Methods
The present study comprised the following steps: (1) collection of continuous non-averaged EEG data for single-trial classification; (2) application of SOBI to continuous non-averaged EEG data; (3) identification of components corresponding to left and right SI activations; (4) neural network classification of trial types using two kinds of input data: the SOBI component data derived in steps 2 and 3 and the “best sensor” data. It is important to point out that in the single-trial analysis
SOBI-recovered left and right SI components
The spatial and temporal aspects of previously validated SI components (see, Tang et al., 2005a) from a typical subject are shown in Fig. 2. The CSD maps and the corresponding equivalent current dipole models (Fig. 2, top) revealed the spatial origin of the two identified components, one corresponding to the left and the other to the right SI. The differential SEPs to contra- and ipsilateral stimulations–selective response to contralateral stimulation (Fig. 2, bottom)–further confirmed the
Discussion
The goal of the present study was to evaluate whether single-trial ERPs from previously validated SOBI neuronal components were more distinguishable among different experimental manipulations than the single-trial ERPs from the comparable “best sensors”. Using the performance of back-propagation neural networks as a quantitative measure of the differentiability of single-trials, we showed that single-trial ERPs from appropriately selected SOBI-recovered components were more distinguishable than
Summary
We demonstrated that SEPs from SOBI-recovered components can serve as better inputs to neural network classifiers to achieve improved single-trial ERP classification over the “best sensor” ERPs. This finding is consistent with our hypothesis that patterns of neuronal activations associated with different sensory, motor, or cognitive activations – i.e., the experimental manipulations – can be made more distinguishable when represented as SOBI component ERPs than as the “best sensor” ERPs over
Acknowledgments
This project was funded by a DARPA grant from the Augmented Cognition Program (ONR: N00014-02-1-0348) and the MIND institute (#2021) to ACT. We thank: Drs. K. Friston, J. Kilner, L. Otten, J.Y. Liu, S. Sands, and C. Saron for critical comments on early versions of the manuscript, C.J. McKinney for assistance during data collection, two anonymous reviewers for their helpful comments, and the Institute of Cognitive Neuroscience, University College London, particularly Drs. Ray Dolan and Leun
References (63)
- et al.
What aspect of the fMRI BOLD signal best reflects the underlying electrophysiology in human somatosensory cortex?
Clin. Neurophysiol.
(2003) - et al.
Scalp electrode impedance, infection risk, and EEG data quality
Clin. Neurophysiol.
(2001) - et al.
Beyond averaging: II. Single-trial classification of exogenous event-related potentials using stepwise discriminant analysis
Electroencephalogr. Clin. Neurophysiol.
(1980) - et al.
Independent component analysis: algorithms and applications
Neural Netw.
(2000) - et al.
Mining event-related brain dynamics
Trends Cogn. Sci.
(2004) - et al.
Designing optimal spatial filters for single-trial EEG classification in a movement task
Clin. Neurophysiol.
(1999) - et al.
Linear spatial integration for single-trial detection in encephalography
NeuroImage
(2002) - et al.
On-line EEG classification during externally-paced hand movements using a neural network-based classifier
Electroencephalogr. Clin. Neurophysiol.
(1996) - et al.
EEG-based discrimination between imagination of right and left hand movement
Electroencephalogr. Clin. Neurophysiol.
(1997) - et al.
Functional MRI of human primary somatosensory and motor cortex during median nerve stimulation
Clin. Neurophysiol.
(1999)
Beyond averaging: the use of discriminant functions to recognize event related potentials elicited by single auditory stimuli
Electroencephalogr. Clin. Neurophysiol.
Independent component analysis: an introduction
Trends Cogn. Sci.
Blind source separation of multichannel neuromagnetic responses
Neurocomputing
Independent components of magnetoencephalography: single-trial response onset times
NeuroImage
Validation of SOBI components from high density EEG
NeuroImage
Dipole source localization and fMRI of simultaneously recorded data applied to somatosensory categorization
NeuroImage
Extraction of ocular artefacts from EEG using independent component analysis
Electroencephalogr. Clin. Neurophysiol.
Independence: a new criterion for the analysis of the electromagnetic fields in the global brain?
Neural Netw.
Brain–computer interfaces for communication and control
Clin. Neurophysiol.
Human cortical potentials evoked by stimulation of the median nerve: I. Cytoarchitectonic areas generating short-latency activity
J. Neurophysiol.
Potentials evoked in human and monkey cerebral cortex by stimulation of the median nerve. A review of scalp and intracranial recordings
Brain
Multivariate autoregressive models for classification of spontaneous electroencephalographic signals during mental tasks
IEEE Trans. Biomed. Eng.
An information-maximization approach to blind separation and blind deconvolution
Neural Comput.
Second-order blind separation of correlated sources
A blind source separation technique using second-order statistics
IEEE Trans. Signal Process.
BCI Competition 2003—Data sets Ib and IIb: feature extraction from event-related brain potentials with the continuous wavelet transform and the t-value scalogram
IEEE Trans. Biomed. Eng.
Jacobi angles for simultaneous diagonalization
SIAM J. Matrix Anal. App.
Modern mind-brain reading: psychophysiology, physiology, and cognition
Psychophysiology
A multivariate approach to the analysis of average evoked potentials
IEEE Trans. Biomed. Eng.
The mental prosthesis: assessing the speed of a P300-based brain–computer interface
IEEE Trans. Rehabil. Eng.
Comparison of linear, nonlinear, and feature selection methods for EEG signal classification
IEEE Trans. Neural Syst. Rehabil. Eng.
Cited by (36)
A resting-state network for novelty: Similar involvement of a global network under rest and task conditions
2022, Psychiatry Research - NeuroimagingMulti-ganglion ANN based feature learning with application to P300-BCI signal classification
2015, Biomedical Signal Processing and ControlCitation Excerpt :Researchers have proposed the applications of different advanced spatial, temporal and spectrum feature extraction methods to BCI systems. These methods include principal component analysis (PCA) [18], common spatial pattern (CSP) [19–22], wavelet transform method [23], blind source separation [24] etc. They can reduce the dimensionality of BCI signals’ feature space for the classification procedure.
Bayesian estimation of ERP components from multicondition and multichannel EEG
2014, NeuroImageCitation Excerpt :In particular, principal component analysis (PCA) was used in combination with a dipole source localization algorithm to extract ERP components by assuming that the spatial patterns of the ERP components were orthogonal (Donchin, 1966). The second-order blind identification (SOBI) algorithm was applied to single-trial ERP analysis (Tang et al., 2006). Information maximization (Infomax), which is a popular algorithm for independent component analysis (ICA), was used to separate the statistically independent EEG components; the single-trial latency and the amplitude variance were subsequently examined in the components identified as evoked responses (Jung et al., 2001; Makeig et al., 2002).
One-unit second-order blind identification with reference for short transient signals
2013, Information SciencesCitation Excerpt :The available constrained BSS algorithms are based on either a higher order statistic (HOS), typically the fourth-order [9,10], or, a second-order statistic (SOS). Compared with an HOS [11–15], a SOS has the following advantages: (1) non-stationary signals can be separated using only a SOS [3,16], (2) SOS performs better than HOS on data with short sample points [17,18], and (3) no limitation on the number of Gaussian sources [17]. Further, the second-order separability involving reference signals has been shown feasible, even for non-linear models [17].