Estimation of single sweep steady-state visual evoked potentials by adaptive line enhancement

Multiscale noise suppression and feature frequency extraction in SSVEP based on underdamped second-order stochastic resonance

OBJECTIVE

As one of the commonly used control signals of brain-computer interface (BCI), steady-state visual evoked potential (SSVEP) exhibits advantages of stability, periodicity and minimal training requirements. However, SSVEP retains the non-linear, non-stationary and low signal-to-noise ratio (SNR) characteristics of EEG. The traditional SSVEP extraction methods regard noise as harmful information and highlight the useful signal by suppressing the noise. In the collected EEG, noise and SSVEP are usually coupled together, the useful signal is inevitably attenuated while the noise is suppressed. Also, an additional band-pass filter is needed to eliminate the multi-scale noise, which causes the edge effect.

APPROACH

To address this issue, a novel method based on underdamped second-order stochastic resonance (USSR) is proposed in this paper for SSVEP extraction.

MAIN RESULTS

A synergistic effect produced by noise, useful signal and the nonlinear system can force the energy of noise to be transferred into SSVEP, and hence amplifying the useful signal while suppressing multi-scale noise. The recognition performances of detection are compared with the widely-used canonical coefficient analysis (CCA) and multivariate synchronization index (MSI).

SIGNIFICANCE

The comparison results indicate that USSR exhibits increased accuracy and faster processing speed, which effectively improves the information transmission rate (ITR) of SSVEP-based BCI.

A dynamic selection method for reference electrode in SSVEP-based BCI

In SSVEP-based Brain-Computer Interface (BCI), it is very important to get an evoked EEG with a high signal to noise ratio (SNR). The SNR of SSVEP is fundamentally related to the characteristics of stimulus, such as its intensity and frequency, and it is also related to both the reference electrode and the active electrode. In the past, with SSVEP-based BCI, often the potential at ‘Cz’, the average potential at all electrodes or the average mastoid potential, were statically selected as the reference. In conjunction, a certain electrode in the occipital area was statically selected as the active electrode for all stimuli. This work proposed a dynamic selection method for the reference electrode, in which all electrodes can be looked upon as active electrodes, while an electrode which can result in the maximum sum relative-power of a specific frequency SSVEP can be confirmed dynamically and considered as the optimum reference electrode for that specific frequency stimulus. Comparing this dynamic selection method with previous methods, in which ‘Cz’, the average potential at all electrodes or the average mastoid potential were selected as the reference electrode, it is demonstrated that the SNR of SSVEP is improved significantly as is the accuracy of SSVEP detection.

SSVEP extraction based on the similarity of background EEG

Steady-state Visual Evoked Potential (SSVEP) outperforms the other types of ERPs for Brain-computer Interface (BCI), and thus it is widely employed. In order to apply SSVEP-based BCI to real life situations, it is important to improve the accuracy and transfer rate of the system. Aimed at this target, many SSVEP extraction methods have been proposed. All these methods are based directly on the properties of SSVEP, such as power and phase. In this study, we first filtered out the target frequencies from the original EEG to get a new signal and then computed the similarity between the original EEG and the new signal. Based on this similarity, SSVEP in the original EEG can be identified. This method is referred to as SOB (Similarity of Background). The SOB method is used to detect SSVEP in 1s-length and 3s-length EEG segments respectively. The accuracy of detection is compared with its peers computed by the widely-used Power Spectrum (PS) method and the Canonical Coefficient (CC) method. The comparison results illustrate that the SOB method can lead to a higher accuracy than the PS method and CC method when detecting a short period SSVEP signal.

Steady-state visually evoked potentials: focus on essential paradigms and future perspectives

After 40 years of investigation, steady-state visually evoked potentials (SSVEPs) have been shown to be useful for many paradigms in cognitive (visual attention, binocular rivalry, working memory, and brain rhythms) and clinical neuroscience (aging, neurodegenerative disorders, schizophrenia, ophthalmic pathologies, migraine, autism, depression, anxiety, stress, and epilepsy). Recently, in engineering, SSVEPs found a novel application for SSVEP-driven brain-computer interface (BCI) systems. Although some SSVEP properties are well documented, many questions are still hotly debated. We provide an overview of recent SSVEP studies in neuroscience (using implanted and scalp EEG, fMRI, or PET), with the perspective of modern theories about the visual pathway. We investigate the steady-state evoked activity, its properties, and the mechanisms behind SSVEP generation. Next, we describe the SSVEP-BCI paradigm and review recently developed SSVEP-based BCI systems. Lastly, we outline future research directions related to basic and applied aspects of SSVEPs.

A four command BCI system based on the SSVEP protocol

This paper discusses the development of a four command BCI system. This system is composed of a wearable electroencephalogram acquisition unit interfaced to a computer by a wireless Bluetooth (BT) connection. The implemented system relies on the steady-state visual evoked potential (SSVEP) protocol applied to a four selection system. In order to achieve the maximum reliability against false positives a five class classifier was used considering the idle state as an independent class. In order to maximize the usability of the system a two channel solution was tested and adopted. The BCI algorithm was based on a supervised multi-class classifier implemented by combining different binary regularized linear discriminant analysis (RLDA) classifiers. The biofeedback was evaluated by combining the resultant time signed distance with quality index related to the number of coherent identification obtained with the one-vs-all approach.

Design and implementation of a brain-computer interface with high transfer rates

This paper presents a brain-computer interface (BCI) that can help users to input phone numbers. The system is based on the steady-state visual evoked potential (SSVEP). Twelve buttons illuminated at different rates were displayed on a computer monitor. The buttons constituted a virtual telephone keypad, representing the ten digits 0-9, BACKSPACE, and ENTER. Users could input phone number by gazing at these buttons. The frequency-coded SSVEP was used to judge which button the user desired. Eight of the thirteen subjects succeeded in ringing the mobile phone using the system. The average transfer rate over all subjects was 27.15 bits/min. The attractive features of the system are noninvasive signal recording, little training required for use, and high information transfer rate. Approaches to improve the performance of the system are discussed.

Subspace averaging of steady-state visual evoked potentials

A new algorithm for doing signal averaging of steady-state visual evoked potentials (VEP's) is described. The subspace average is obtained by finding the orthogonal projection of the VEP measurement vector onto the signal subspace, which is based on a sinusoidal VEP signal model. The subspace average is seen to out-perform the conventional average using a new signal-to-noise-ratio-based performance measure on simulated and actual VEP data.

Subspace regularization method for the single-trial estimation of evoked potentials

A method for the single-trial estimation of the evoked potentials is proposed. The method is based on the so-called subspace regularization approach in which the second-order statistics of the set of the measurements is used to form a prior information model for the evoked potentials. The method is closely related to the Bayesian estimation. The performance of the proposed method is evaluated using realistic simulations. As a specific application the method is applied to the estimation of the target responses in the P300 test.

Analysis of pattern reversal visual evoked potentials (PRVEP's) by spline wavelets

In this study, the pattern-reversal visual evoked potentials (PRVEP's) collected from normal and demented subjects are investigated by applying the quadratic spline wavelet analysis. The data are decomposed into six octave frequency bands. For quantitative purposes, the wavelet coefficients in the residual waveform representing the delta-theta band activity (0-8 Hz) are explored to characterize the (N70-P100-N130) complex. Specifically, the coefficients corresponding to the location of N70, P100, and N130 peaks are investigated for their sign in order to test whether they represent a consistent (N70-P100-N130) complex in the averaged waveform. Waveforms with normal latency (N70-P100-N130) complex are observed to have positive second, negative third, and positive fourth coefficients in amplitude in their residual scale standing for the delta-theta (0-8 Hz) band activity. The method allows for the analysis of oscillatory-phase behavior of the normal and pathological PRVEP's in their delta-theta band based on a few quantitative measures consistent with the time-frequency occurrence of the major components of the evoked potential.

An adaptive filter for steady-state evoked responses

A 2-weight adaptive filter that determines the amplitude and phase of steady-state evoked potentials is presented. Reference signals are derived from the visual stimulator that are related to corresponding harmonics of the response and the filter weights are adjusted so as to minimize the squared estimation error between the reference and the recorded signal using the recursive least squares (RLS) method. The filter, which acts as an adaptive bandpass filter, is followed by a detector based on the T2circ statistic. The performance of the RLS adaptive filter was compared to that of the conventional Discrete Fourier Transform (DFT) and the filtered DFT of Tang and Norcia in a series of simulations with known sinusoids buried in Gaussian noise and in EEG noise. In the simulations, the RLS adaptive filter detected signals at about 3-4 times lower signal to noise ratios than did the DFT. The RLS filter also outperformed the filtered DFT. Qualitatively similar results were obtained with human visual evoked potential recordings. The adaptive RLS filter significantly outperforms both the DFT and filtered DFT and is much simpler to implement than the filtered DFT method of Tang and Norcia.

Rapid measurement of somatosensory evoked potential response to cerebral artery occlusion

The aim of the paper is to determine the speed of the neurological response to cerebral artery occlusion by monitoring transient changes in somatosensory evoked potentials (SEPs). SEPs, continuously monitored during temporary clipping of the middle cerebral artery (MCA) in anaesthetised cats, are analysed. The SEP signals are modelled by a quasi-periodic Fourier series, the coefficients of which are estimated with the aid of two adaptive least squares estimation algorithms. The energy levels at various harmonics throughout the protocol are obtained directly from the filter weights. Noise covariance is estimated from pre-stimulus recording, and the adaptation rate of the algorithm is adjusted sweep-by-sweep to accommodate transient changes in the pre-stimulus noise level. After the occlusion, a significant decrease (p < 0.05) in SEP amplitude is observed. The change in latency is not statistically significant (p approximately equal to 0.5). The spectral trends show a sudden decline in energy at all harmonics immediately following occlusion, although when the amplifier bandwidth is changed to 5-1500 Hz (from an initial setting of 30-1500 Hz), the fundamental frequency component of the SEP signal shows the greatest responsiveness to injury. The average time constant of the decline in amplitude resulting from MCA occlusion is only 10.6 +/- 4.0 s. It is concluded that rapid detection of cerebral artery occlusion and ischaemia may be feasible by continuously monitoring SEP signals and analysing transient changes in time and frequency domains.

Tracking of changes in latency and amplitude of the evoked potential by using adaptive LMS filters and exponential averagers

The adaptive LMS algorithm in combination with exponential averagers are compared to the use of exponential averagers only in tracking latency and amplitude changes in the evoked potential. The estimator is intended for use in applications where neurologic functions are monitored by detecting changes in the evoked potential. Two different structures of the estimator are evaluated and it is found that averaging before filtering is to be preferred. It is shown that the desired signal to the LMS-filter can have a rather low SNR with only mirror influence on the estimator performance. The estimator which combines an LMS-filter and an exponential averager was shown to detect changes in latency faster than the estimator which uses a nonfiltered average. The LMS-filter is shown to exhibit bias in the estimate of the evoked potential due to the fact that response and background spectra has overlapping frequency ranges. The bias seems not to affect the latency estimation while amplitude estimation was clearly affected. Simulations are performed with both white noise and EEG background.