A two-dimensional approach for lossless EEG compression

In this paper, we study various lossless compression techniques for electroencephalograph (EEG) signals. We discuss a computationally simple pre-processing technique, where EEG signal is arranged in the form of a matrix (2-D) before compression. We discuss a two-stage coder to compress the EEG matrix, with a lossy coding layer (SPIHT) and residual coding layer (arithmetic coding). This coder is optimally tuned to utilize the source memory and the i.i.d. nature of the residual. We also investigate and compare EEG compression with other schemes such as JPEG2000 image compression standard, predictive coding based shorten, and simple entropy coding. The compression algorithms are tested with University of Bonn database and Physiobank Motor/Mental Imagery database. 2-D based compression schemes yielded higher lossless compression compared to the standard vector-based compression, predictive and entropy coding schemes. The use of pre-processing technique resulted in 6% improvement, and the two-stage coder yielded a further improvement of 3% in compression performance.     

An Approach to Quantifying the Multi‐Channel EEG Spatial‐Temporal Feature

Investigation of spatial characteristics in the multi‐channel EEG (electroencephalogram) has been of great importance to the fundamental and clinical medicine. Conventional methods are mostly based on the coherence and correlation analysis in time and frequency domain. The authors present the method of quantifying the global waveform complexity in EEG by analyzing the high‐dimensional state‐space trajectory constructed from the multi‐channel EEG. The resulting average complexity index ( ) is well distinct between the background (4.0 ∼ 4.5) and event (3.0 ∼ 3.5 for the focal‐sharp‐wave transient) in the 5‐channel case study. In addition, the running measurement for different electrode arrays, for a fixed number of state‐space dimension (n = 5), reveals the EEG spatial characteristic. An array composed of nearby channels with similar EEG activities tends to have a small . On the other hand, the computed increases as the array consists of electrode sites with spatially incorrelated EEG activities. Thus the method is capable of quantifying the spatial‐temporal feature of the multi‐channel EEG.     

Visual unit,EEG and sustained potential shift responses to biologically significant stimuli in the brain of toads (Bufo bufo)

Summary Visual unit activity, EEG changes and sustained potential shifts (SPS) were recorded from the toad tectum whilst the animal was presented with a visual stimulus. Telencephalic EEGs were also recorded.On the surface of the tectum, retinal unit activity preceded a sustained negative shift in potential and an increase in the amplitude and dominant frequency of the EEG. In deeper layers of the tectum, T5 units with configurational selectivity for wormlike stimuli were found. The activity of these units followed a pronounced SPS and EEG change.Visual unit activity was most pronounced during the negative-going phase of the synchronised EEG, when there was also a small decrease in amplitude of neuronal spikes. Similarities between the latencies and durations of EEGs and SPSs, and their response decrements, on repeated stimulus presentation, implies a close relationship between them not shared by the visual units studied. The specific activity of tectal units is discussed in relation to the correlated EEG and SPS changes, which may form part of an adaptive sensitizing mechanism.Abbreviations EEG electroencephalogram - ERF excitatory receptive field - SPS sustained potential shift - T4, T5 tectal neurons     

Modulation of attention in healthy children using a course of EEG-feedback sessions

We studied changes in the power spectra of EEG in the course of sessions of feedback by EEG characteristics (neurofeedback sessions) and estimated the effects of neurofeedback on psychological and EEG correlates of voluntary attention. Indices of the latter were estimated using Bourdon’s test (a correcture test) and Schulte’s tables. Twenty-nine reasonably healthy 10-to 13-year-old children took part in the study; they were divided into two groups, an experimental group (n = 12) and a control group (n = 15). The results obtained support the statement on noticeable changes in the functional state of the brain both immediately in the course of a neurofeedback session and after a course of such trainings. Changes in the ratios of the spectral powers of the beta₁ vs theta rhythms and the low-frequency beta vs theta rhythms were found in EEG recorded from the sensorimotor zone of the right hemisphere (C4). The observed changes in the spectral characteristics of EEG induced by the course of neurofeedback sessions were accompanied by the improvement of a few indices of voluntary attention. Neirofiziologiya/Neurophysiology, Vol. 38, Nos. 5/6, pp. 458–465, September–December, 2006.     

Deterministic chaos and the first positive Lyapunov exponent: a nonlinear analysis of the human electroencephalogram during sleep

Under selected conditions, nonlinear dynamical systems, which can be described by deterministic models, are able to generate so-called deterministic chaos. In this case the dynamics show a sensitive dependence on initial conditions, which means that different states of a system, being arbitrarily close initially, will become macroscopically separated for sufficiently long times. In this sense, the unpredictability of the EEG might be a basic phenomenon of its chaotic character. Recent investigations of the dimensionality of EEG attractors in phase space have led to the assumption that the EEG can be regarded as a deterministic process which should not be mistaken for simple noise. The calculation of dimensionality estimates the degrees of freedom of a signal. Nevertheless, it is difficult to decide from this kind of analysis whether a process is quasiperiodic or chaotic. Therefore, we performed a new analysis by calculating the first positive Lyapunov exponent L ₁ from sleep EEG data. Lyapunov exponents measure the mean exponential expansion or contraction of a flow in phase space. L ₁ is zero for periodic as well as quasiperiodic processes, but positive in the case of chaotic processes expressing the sensitive dependence on initial conditions. We calculated L ₁ for sleep EEG segments of 15 healthy men corresponding to the sleep stages I, II, III, IV, and REM (according to Rechtschaffen and Kales). Our investigations support the assumption that EEG signals are neither quasiperiodic waves nor a simple noise. Moreover, we found statistically significant differences between the values of L ₁ for different sleep stages. All together, this kind of analysis yields a useful extension of the characterization of EEG signals in terms of nonlinear dynamical system theory.     

Multifractality of decomposed EEG during imaginary and real visual-motor tracking

We test the possible multifractal properties of dominant EEG frequency components, when a subject tracks a path on a map, either only by eyes (imaginary movement – IM) or by visual-motor tracking of discretely moving spot in regular (RM) and Brownian time-step (BM) (real tracking of moving spot). We check the hypotheses that the fractal properties of filtered EEG (1) change with respect to the law of spot movement; (2) differ among filtered EEG components and scalp sites; (3) differ among real and imaginary tracking. Sixteen right-handed subjects begin to perform IM, next – real spot tracking (RM and BM) following a moving spot on streets of a citymap displayed on a computer screen, by push forward/backward a joystick. Multichannel long-lasting EEG is band-pass filtered for theta, alpha, beta and gamma oscillations. The Wavelet-Transform-Modulus-Maxima-Method is applied to reveal multifractality [local fractal dimensions D _max(h)] among task conditions, frequency bands and sites. Non-parametric statistical estimation of the fractal measures h _{D max} is finally applied. Multifractality is established for all experimental conditions, EEG components and sites as follows among filtered components – anticorrelation (h _Dmax < 0.5) in beta and gamma, and long-range correlation (h _Dmax > 0.5) for theta and alpha oscillations; among tasks – for RM and BM, h _Dmax differ significantly whereas IM resembles mostly RM; among sites – no significant difference for local fractal properties is established. The results suggest that for both imaginary and real visual-motor tracking a line, multifractal scaling, specific for lower and higher EEG oscillations, is a very stable intrinsic one for the activity of large brain areas. The external events (task conditions) insert weak effect on the scaling.     

Topographic analysis of dimension estimates of EEG and filtered rhythms in epileptic patients with complex partial seizures

Nonlinear dynamic properties were analyzed on the EEG and filtered rhythms recorded from healthy subjects and epileptic patients with complex partial seizures. Estimates of correlation dimensions of control EEG, interictal EEG and ictal EEG were calculated. The values were demonstrated on topograms. The delta (0.5–4 Hz), theta (4–8 Hz), alpha (8–13 Hz), beta (13–30 Hz) and gamma (30–40 Hz) components were obtained and considered as signals from the cortex. Corresponding surrogate data was produced. Firstly, the influence of sampling parameters on the calculation was tested. The dimension estimates of the signals from the frontal, temporal, parietal and occipital regions were computed and compared with the results of surrogate data. In the control subjects, the estimates between the EEG and surrogate data did not differ (P > 0.05). The interictal EEG from the frontal region and occipital region, as well as its theta component from the frontal region, and temporal region, showed obviously low dimensions (P < 0.01). The ictal EEG exhibited significantly low-dimension estimates across the scalp. All filtered rhythms from the temporal region yielded lower results than those of the surrogate data (P < 0.01). The dimension estimates of the EEG and filtered components markedly changed when the neurological state varied. For each neurological state, the dimension estimates were not uniform among the EEG and frequency components. The signal with a different frequency range and in a different neurological state showed a different dimension estimate. Furthermore, the theta and alpha components demonstrated the same estimates not only within each neurological state, but also among the different states. These results indicate that the theta and alpha components may be caused by similar dynamic processes. We conclude that the brain function underlying the ictal EEG has a simple mechanism. Several heterogeneous dynamic systems play important roles in the generation of EEG. Received: 10 December 1999 / Accepted in revised form: 8 May 2000     

On the synchrony of steady state visual evoked potentials and oscillatory burst events

In this paper, we investigate the large-scale synchrony of EEG oscillatory bursts, during stimulation by a flickering square of light. Whereas most studies focus on averaged raw EEG responses, this study considers oscillatory events within EEG of single trials, which leads to various new insights. We recorded EEG signals before, during and after stimulation by a flickering square of light in medium (16 Hz) and high frequency (32 Hz) ranges. Similar oscillatory bursts, to those observed in spontaneous EEG, can be found in single-trial synchrony of steady state visual evoked potentials (SSVEP). These bursts are extracted from the EEG of single trials using bump modeling. Stochastic event synchrony method is applied to those events, which quantifies synchronies of oscillatory bursts on a large-scale basis. Those oscillatory patterns have a significantly higher degree of co-occurrence during SSVEP, uncorrelated with ongoing signal synchrony. It means that EEG oscillatory patterns are presumably an outcome of brain activity, rather than a mere side effect of ongoing EEG. They undergo a consistent reorganization during visual stimulation, preferentially along the visual pathway, depending on magno or parvo stimulations. Flickering stimuli may induce some cognitive side-effects depending on the stimulation frequency.

Multi-Variate EEG Analysis as a Novel Tool to Examine Brain Responses to Naturalistic Music Stimuli

Note onsets in music are acoustic landmarks providing auditory cues that underlie the perception of more complex phenomena such as beat, rhythm, and meter. For naturalistic ongoing sounds a detailed view on the neural representation of onset structure is hard to obtain, since, typically, stimulus-related EEG signatures are derived by averaging a high number of identical stimulus presentations. Here, we propose a novel multivariate regression-based method extracting onset-related brain responses from the ongoing EEG. We analyse EEG recordings of nine subjects who passively listened to stimuli from various sound categories encompassing simple tone sequences, full-length romantic piano pieces and natural (non-music) soundscapes. The regression approach reduces the 61-channel EEG to one time course optimally reflecting note onsets. The neural signatures derived by this procedure indeed resemble canonical onset-related ERPs, such as the N1-P2 complex. This EEG projection was then utilized to determine the Cortico-Acoustic Correlation (CACor), a measure of synchronization between EEG signal and stimulus. We demonstrate that a significant CACor (i) can be detected in an individual listener''s EEG of a single presentation of a full-length complex naturalistic music stimulus, and (ii) it co-varies with the stimuli’s average magnitudes of sharpness, spectral centroid, and rhythmic complexity. In particular, the subset of stimuli eliciting a strong CACor also produces strongly coordinated tension ratings obtained from an independent listener group in a separate behavioral experiment. Thus musical features that lead to a marked physiological reflection of tone onsets also contribute to perceived tension in music.     

Dynamics of EEG potentials at the beginning of a series of EEG-feedback sessions

We studied changes in the amplitudes of event-related EEG potentials (ERPs) and power spectra of background EEG in the course of a series of EEG-feedback sessions directed toward an increase in the ratio of powers of the α vs θ rhythms. The examined group included 70 volunteers divided into an experimental group (n = 37) and a control group (n = 33). The intensity of acoustic white noise overlapping the musical background served as a feedback signal; it became lower with increase in the above ratio, while in the control group it remained constant. The EEG potentials were recorded from C3 and C4 leads. The ERPs were recorded within a paradigm of measuring time intervals. Within a series of EEG-feedback sessions, the α/θ ratio decreased somewhat both in the control and experimental groups, but in subjects of the latter group this decrease was less significant, and the mean intragroup index became significantly greater than the respective value in the control group after the end of the third session. The EEG-feedback sessions also resulted in significant increases in the amplitudes of early components of the readiness potential in both hemispheres and in the amplitude of the contingent negative variation in the right hemisphere. We conclude that, in most healthy subjects, at least three sessions of α/θ training are necessary to form an effective series providing considerable changes in the pattern of EEG potentials. Neirofiziologiya/Neurophysiology, Vol. 39, No. 1, pp. 88–98, January–February, 2007.     

Sleep and EEG spectra in the pigeon (Columba livia) under baseline conditions and after sleep deprivation

Summary Sleep in adult domestic pigeons was studied by continuous 24-h recording of the EEG, EMG and EOG. Vigilance states were scored on the basis of behavioral observations, visual scoring of the polygraph records, and EEG power spectra.The animals showed a clear nocturnal preference for sleep. Throughout the dark period, EEG slow-wave activity was at a uniform level, whereas REM sleep (REMS) showed an increasing trend.EEG power density values differed significantly between the vigilance states. In general the values were highest in nonREM sleep (NREMS), intermediate in waking (W) and lowest in REMS.Twenty-four hour sleep deprivation reduced W and increased REMS, effects that are well documented in mammals. Unlike in mammals, EEG slow-wave activity remained unchanged, whereas EOG activity in W and NREMS was enhanced.Abbreviations EEG electroencephalogram - EMG electromyogram - EOG electrooculogram - SD sleep deprivation - L light - D dark - LD light dark - NREMS non rapid eye movement sleep - REMS REM sleep     

Spatial synchronization of the θ and α band cortical electrical oscillations in the formation of a set to an angry face expression in 5- to 11-year-old children

The θ and α band EEG coherence during forming and testing sets to an emotionally negative facial expression (angry) was studied in five- to six-year-old (n = 18) and 10- to 11-year-old (n = 25) children. Younger children displayed lower coherence values, especially in the right hemisphere, than older ones. Differences in θ- and α band coherence in the cases of rigid and flexible sets to an angry face expression were also revealed. With relatively rigid forms of cognitive activity, the EEG coherence values were lower. A correlation between the electrophysiological and behavioral data was established. It supports the hypothesis that two functional brain integration systems, the corticohippocampal and the frontothalamic ones, play a role in the processes of facial expression recognition and provide cognition flexibility.     

The Sleep Cycle Modelled as a Cortical Phase Transition

We present a mean-field model of the cortex that attempts to describe the gross changes in brain electrical activity for the cycles of natural sleep. We incorporate within the model two major sleep modulatory effects: slow changes in both synaptic efficiency and in neuron resting voltage caused by the ∼90-min cycling in acetylcholine, together with even slower changes in resting voltage caused by gradual elimination during sleep of somnogens (fatigue agents) such as adenosine. We argue that the change from slow-wave sleep (SWS) to rapid-eye-movement (REM) sleep can be understood as a first-order phase transition from a low-firing, coherent state to a high-firing, desychronized cortical state. We show that the model predictions for changes in EEG power, spectral distribution, and correlation time at the SWS-to-REM transition are consistent not only with those observed in clinical recordings of a sleeping human subject, but also with the on-cortex EEG patterns recently reported by Destexhe et al. [J. Neurosci. 19(11), (1999) 4595–4608] for the sleeping cat.     

Modulation of EEG Rhythms and Changes in Spike Activity of Noradrenergic Neurons of the <Emphasis Type="Italic">Locus Coeruleus</Emphasis> Related to Feedback Sessions by EEG Characteristics

In chronic experiments on awake cats, we studied the dynamics of the spectral power density (SPD) of the α rhythm vs SPD of the θ rhythm ratio and also of the characteristics of impulse activity generated by supposedly noradrenergic (NA) neurons of the locus coeruleus in the course of feedback (FB) sessions by EEG characteristics (EEG-FB). Trainings were performed using a technique analogous to that in EEG-FB sessions for humans. The level of a sound noise signal presented to the animal decreased with increase in the α/θ SPD ratio in the occipital lead. Changes in the level of the sound signal did not depend on EEG modulation in the control series. The animals were trained to correlate changes in the loudness of the sound signal with the power of EEG rhythms and, in such a way, to control the latter. The α/θ SPD ratio in EEG-FB sessions changed mostly due to a significant increase in the α rhythm power. The frequency of the impulse activity of NA neurons increased in a parallel manner with such EEG modulation. Possible mechanisms of the involvement of the cerebral NA system in the formation of the effects of EEG-FB sessions are discussed.     

The effects of morphine on the relationship between fetal EEG, breathing and blood pressure signals using fast wavelet transform

In this study, we introduce the fast wavelet transform (WT) as a method for investigating the effects of morphine on the electroencephalogram (EEG), respiratory activity and blood pressure in fetal lambs. Morphine was infused intravenously at 25 mg/h. The EEG, respiratory activity and blood pressure signals were analyzed using WT. We performed wavelet decomposition for five sets of parameters D _2j where -1 < j 5. The five series WTs represent the detail signal bandwidths: 1, 16–32 Hz; 2, 8–16 Hz; 3, 4–8 Hz; 4, 2–4 Hz; 5, 1–2 Hz. Before injection of the high-dose morphine, power in the EEG was high in all six frequency bandwidths. The respiratory and blood pressure signals showed common frequency components with respect to time and were coincident with the low-voltage fast activity (LVFA) EEG signal. Respiratory activity was observed during only some of the LVFA periods, and was completely absent during high-voltage slow activity (HVSA) EEG. The respiratory signal showed dominant power in the fourth wavelet band, and less power in the third and fifth bands. The blood pressure signal was also characterized by dominant power in the fourth wavelet band. This power was significantly increased during periods of respiratory activity. There was a strong relationship between fetal EEG, blood pressure and breathing movements. However, the injection of high-dose morphine resulted in a disruption of the normal cyclic pattern between the two EEG states and a significant increase in power in the first wavelet band. In addition, the high-dose drug resulted in a significant increase in the power of respiratory signal in the fourth and fifth wavelet bands, while power was reduced in the third wavelet band. Breathing activity was also continuous after the drug. The high-dose morphine also caused a temporary power shift from the third wavelet band to the fourth wavelet band for the 30-min period after injection of drug. Finally, high-dose morphine completely destroyed the correlation between EEG, breathing and blood pressure signals.     

Detection of EEG abnormalities with feedback stimulation

A feedback method for testing the reactivity of the occipital-parietal EEG in selected brain-lesioned patients revealed abnormalities of(a) insufficient reactivity,(b) bilateral differences in reactivity, and(3) asynchrony. These abnormalities were more evident during feedback stimulation than in the baseline conditions. The utility of feedback method for detecting EEG abnormalities rests on the increased stability or decreased noisy variation in the EEG during feedback. The EEG becomes more predictable even to the on-line human observer. This makes it easier to detect aberrations or deviations from normal effects. Some effects can only be seen with feedback such as the bilateral differences which occur when the left side controls the feedback compared to when the right side controls it. The results show that feedback EEG is a useful tool in clinical research and indicate that a clinical diagnostic test could be developed with more research. However, the feedback EEG method is not yet a proven diagnostic technique.The assistance of Eric Peper, Sylvia Runnals, Rosemary Billingslea, and Thomas McLaughlin in the EEG recording and data analysis was indispensible to this study.     

Induction of emotional states by reading aloud texts with various emotional valences and changes in the EEG power in the β and γ frequency bands

The EEG power in the β₁, β₂, and γ frequency bands during reading according to the method of “self-regulatory utterance” was compared in subjects reading aloud emotionally neutral business texts related to an unknown field of activity, fiction texts with clear positive or negative valences or personally important autobiographic texts with similar emotional valences. Two groups of subjects participated in the study: students training to be actors (N = 22) and students with other specializations (N = 23). We observed higher values of the EEG power in the γ (30–40 Hz) and β₂ (18.5–29.5 Hz) frequency bands when comparing the states during reading of emotionally positive and emotionally negative fiction texts and personally important texts. These data are similar to our previous studies with the use of techniques that apply internal induction of positive or negative emotions without speech, in different groups of subjects. Internal induction of positive emotions was associated with an increase in the EEG power in these bands compared to the performance of an emotionally neutral task, whereas induction of negative emotions resulted in a decrease in the EEG power.     

Brain’s alpha activity is highly reduced in euthymic bipolar disorder patients

Brain’s alpha activity and alpha responses belong to major electrical signals that are related to sensory/cognitive signal processing. The present study aims to analyze the spontaneous alpha activity and visual evoked alpha response in drug free euthymic bipolar patients. Eighteen DSM-IV euthymic bipolar patients (bipolar I n = 15, bipolar II n = 3) and 18 healthy controls were enrolled in the study. Patients needed to be euthymic at least for 4 weeks and psychotrop free for at least 2 weeks. Spontaneous EEG (4 min eyes closed, 4 min eyes open) and evoked alpha response upon application of simple visual stimuli were analyzed. EEG was recorded at 30 positions. The digital FFT-based power spectrum analysis was performed for spontaneous eyes closed and eyes open conditions and the response power spectrum was also analyzed for simple visual stimuli. In the analysis of spontaneous EEG, the ANOVA on alpha responses revealed significant results for groups (F(1,34) = 8.703; P < 0.007). Post-hoc comparisons showed that spontaneous EEG alpha power of healthy subjects was significantly higher than the spontaneous EEG alpha power of euthymic patients. Furthermore, visual evoked alpha power of healthy subjects was significantly higher than visual evoked alpha power of euthymic patients (F(1,34) = 4.981; P < 0.04). Decreased alpha activity in spontaneous EEG is an important pathological EEG finding in euthymic bipolar patients. Together with an evident decrease in evoked alpha responses, the findings may lead to a new pathway in search of biological correlates of cognitive impairment in bipolar disorder.     

Changes in EEG Rhythms and Spike Activity of Brainstem Dopaminergic Neurons Induced by Neurofeedback Sessions in Cats

We examined the dynamics of the ratios of spectral power densities (SPDs) of the alpha vs theta rhythms (α/θ ratio). of EEG and of the spiking frequency of supposedly dopaminergic (DA) neurons of the ventral tegmentum in the course of neurofeedback sessions directed toward changes in the EEG characteristics. Trainings were performed using techniques analogous to that used in neurofeedback sessions in humans. The level of the noise acoustic signal presented to the animal decreased with increase in the α/θ ratio in the occipital leads. In the control realizations, there were no dependences between the intensity of the acoustic signal and modulation of the current EEG. It was found that the animals learned, in a conditioned-reflex mode, to correlate changes in the intensity of the sound signal and power of the EEG rhythms and to control the latter; a high sound intensity was probably considered a factor of discomfort. The α/θ ratio in the course of neurofeedback sessions changed due to some increase in the SPD of the alpha EEG component and a noticeable drop in the SPD of theta oscillations. In a parallel manner with such modifications, augmentation of the spike activity of DA neurons was observed. Probable mechanisms of the involvement of the cerebral DA system in the formation of the effects of neurofeedback sessions are discussed.     

Electrophysiological Correlates of Attentional Disorders in 12- to 13-Year-Old Adolescents

The EEG was recorded in 12- to 13-year-old adolescents with attention deficit hyperactivity disorder (ADHD). During the EEG recording, the subjects performed an attention test. Pairs of tone stimuli in the combinations low–low and low–high were presented in random order with equal probabilities with intervals of 2.3 s. Subjects had to press a button as soon as possible in response to low–low tone pair presentation (a go trial) and not to press the button when a pair consisting of low and high tones was presented (a no-go trial). On the basis of the results of the test performance (the numbers of omissions and false alarms), the subjects were divided into three groups characterized by the degree of severity of the attentional disorder. The evoked potentials in go and no-go trials were detected individually for the subject groups and each derivation. The components N1, N2, P2, P3-go, and P3-no-go were identified in the evoked potentials. These components had different latencies and were differently localized over the head surface. Only the amplitudes of P3-go and P3-no-go in the subjects were statistically significantly correlated with the severity of the attentional disorder.