Anastasiya E. Runnova

Classifying the Perceptual Interpretations of a Bistable Image Using EEG and Artificial Neural Networks

In order to classify different human brain states related to visual perception of ambiguous images, we use an artificial neural network (ANN) to analyze multichannel EEG. The classifier built on the basis of a multilayer perceptron achieves up to 95% accuracy in classifying EEG patterns corresponding to two different interpretations of the Necker cube. The important feature of our classifier is that trained on one subject it can be used for the classification of EEG traces of other subjects. This result suggests the existence of common features in the EEG structure associated with distinct interpretations of bistable objects. We firmly believe that the significance of our results is not limited to visual perception of the Necker cube images; the proposed experimental approach and developed computational technique based on ANN can also be applied to study and classify different brain states using neurophysiological data recordings. This may give new directions for future research in the field of cognitive and pathological brain activity, and for the development of brain-computer interfaces.

Detection of EEG-patterns associated with real and imaginary movements using detrended fluctuation analysis

Authentic recognition of specific patterns of electroencephalograms (EEGs) associated with real and imagi- nary movements is an important stage for the development of brain-computer interfaces. In experiments with untrained participants, the ability to detect the motor-related brain activity based on the multichannel EEG processing is demonstrated. Using the detrended fluctuation analysis, changes in the EEG patterns during the imagination of hand movements are reported. It is discussed how the ability to recognize brain activity related to motor executions depends on the electrode position.

Adaptive filtering of electroencephalogram signals using the empirical-modes method

A new method for the removal of physiological artifacts in the experimental signals of human electroencephalograms (EEGs) has been developed. The method is based on decomposition of the signal in terms of empirical modes. The algorithm involves EEG signal decomposition in terms of empirical modes, searching for modes with artifacts, removing these modes, and restoration of the EEG signal. The method was tested on experimental data and showed high efficiency in the removal of various physiological artifacts in EEGs.

Multifractal spectrum of physiological signals: a mechanism-related approach

In this paper we discuss an approach for mechanism-related analysis of physiological signals performed with the wavelet-based multifractal formalism. This approach assumes estimation of the singularity spectrum for the band-pass filtered processes at different physiological conditions in order to provide explanation of the occurred changes in the Hölder exponents and the multi-fractality degree. We illustrate the considered approach using two examples, namely, the dynamics of the cerebral blood flow (CBF) and the electrical activity of the brain.

Multifractal analysis of real and imaginary movements: EEG study

We study abilities of the wavelet-based multifractal analysis in recognition specific dynamics of electrical brain activity associated with real and imaginary movements. Based on the singularity spectra we analyze electroencephalograms (EEGs) acquired in untrained humans (operators) during imagination of hands movements, and show a possibility to distinguish between the related EEG patterns and the recordings performed during real movements or the background electrical brain activity. We discuss how such recognition depends on the selected brain region.

Power-law statistics of neurophysiological processes analyzed using short signals

We discuss the problem of quantifying power-law statistics of complex processes from short signals. Based on the analysis of electroencephalograms (EEG) we compare three interrelated approaches which enable characterization of the power spectral density (PSD) and show that an application of the detrended fluctuation analysis (DFA) or the wavelet-transform modulus maxima (WTMM) method represents a useful way of indirect characterization of the PSD features from short data sets. We conclude that despite DFA- and WTMM-based measures can be obtained from the estimated PSD, these tools outperform the standard spectral analysis when characterization of the analyzed regime should be provided based on a very limited amount of data.

Analysis of psycho-physiological features of a subject in simple tests with the registration of electroencephalograms

In this paper we found a correlation between the characteristics of a person revealed in classical psychological testing on the basis of Schulte tables, and its neurophysiological features of the functioning of the brain obtained from the time-frequency analysis of EEG. The results obtained are interesting from the point of view of the choice of training strategies for a particular individual. We believe that the obtained results are of interest for fundamental science and applied works of psychological testing and diagnostics. The study of such forming strategies on EEG data can be automated and do not require the work of highly skilled psychologists.

Brain states recognition during visual perception by means of artificial neural network in the different EEG frequency ranges

In the present paper, the possibility of classification by artificial neural networks of a certain architecture of ambiguous images is investigated using the example of the Necker cube from the experimentally obtained EEG recording data of several operators. The possibilities of artificial neural network classification of ambiguous images are investigated in the different frequency ranges of EEG recording signals.

Identification of the patterns of brain activity during the imagination of movements using an artificial neural network

In this paper, we investigate the problem of identification of patterns on magnetoencephalography signals of a brain associated with human movements. The design of registration of experimental data during magnetoencephalography (MEG) is developed and described. Consecutive imaginary movements of the hands and legs of the person are chosen as the basic movements. We solve the problem of recognition and classification of patterns using artificial neural networks. For a multilayer perceptron, good results of recognition of patterns of brain activity associated with different types of motion have been obtained.

Nonlinear dynamics and coherent resonance in a network of coupled neural-like oscillators

In this paper we study the spiking behaviour of a neuronal network consisting of 100 Rulkov elements coupled to each other with randomly chosen coupling strength. We find periodical grouping forming in the signal from all neurons in the network. We discovered the phenomenon of coherent resonance when signal-to-noise ration takes the maximum value at certain values of such parameters as number of neurons in the system, number of stimulated neurons, amplitude of external stimulus and amplitude of internal noise.