CC BY
32
Section BRAIN-COMPUTER INTERFACES, COGNITIVE NAVIGATION WORKSHOP AND NEUROENGINEERING
4. Kreibig S. D. Autonomic nervous system activity in emotion: A review //Biological psychology. - 2010. - T. 84. - №. 3. - C. 394-421.
5. Ravaja N., Somervuori O., Salminen M. Predicting purchase decision: The role of hemispheric asymmetry over the frontal cortex //Journal of Neuroscience, Psychology, and Economics. - 2013. - T. 6. - №. 1. - C. 1.
OM&P
Steady State Visual Evoked Potential Based BCI Study in Overt and Covert Attention
Zafer Iscan1 *, Elena Sokolova2'3, Vadim V. Nikulin1,4
1 Centre for Cognition and Decision Making, National Research University Higher School of Economics, Moscow, Russian Federation;
2 Faculty of Materials Science, Moscow State University, Moscow, Russian Federation ;
3 Energy Science and Technology, Skolkovo Institute of Science and Technology, Moscow, Russian Federation;
4 Neurophysics Group, Department of Neurology, Charité - University Medicine Berlin, Berlin, Germany. * Presenting e-mail: zaferiscan@yahoo.com
Introduction
Brain-computer interfaces (BCIs) have potential to help severely disabled people by translation of the intentions of subjects into a number of different commands. Due to its safety and high time resolution, electroencephalogram (EEG) based BCIs have become popular and various designs using different signals (e.g. P300, oscillations) have been proposed. Among them, steady state visual evoked potentials (SSVEPs) are particularly attractive due to their high signal to noise ratio (SNR). In this study, we proposed a four-class BCI design based on SSVEPs to study the differences between the overt and covert attention.
Methods
Four circles with individual flickering frequencies (5.45, 8.57, 12 and 15 Hz) were presented to healthy participants on an LCD monitor. EEG was recorded from 60 channels with three electrooculogram (EOG) channels in 30 trials. In each trial, subjects focused either on the fixation cross or one of the four circles and paid attention to the circle indicated by a red oval frame for three seconds. Decision tree, Naive Bayes and K-Nearest Neighbor classifiers were used to evaluate the classification performance using features generated by canonical correlation analysis.
Results
The offline classification accuracy for the overt attention was positively correlated with the duration of stimuli and was more than 90% when it was longer than two seconds. The accuracy dropped drastically in the covert attention case. Discussion: Classification performance for the overt attention condition validates the robustness of the SSVEP-based BCIs. However, different classification approaches should be developed in order to classify the covert attention responses.
Experimental Measurements of Human Brain Noise Intensity in Perception of Ambiguous Images
Alexander E. Hramov1,2, Vadim V. Grubov1,2, Alexey A. Koronovskii1,2 *, Maria K. Kurovskaya1,2, Anastasiya E. Runnova1,2,Maxim O. Zhuravlev1,2, Alexander N. Pisarchik3,4
1 Saratov State University, Astrakhanskaya, 83, Saratov, 410012, Russia
2 Saratov State Technical University, Politehnicheskaya, 77, Saratov, 410054, Russia;
3 Centro de Investigaciones en Optica, Loma del Bosque 115, Lomas del Campestre, 37150 Leon, Guanajuato, Mexico;
4 Center for Biomedical Technology, Technical University of Madrid, Campus Montegancedo, 28223 Pozuelo de Alarcon, Madrid, Spain.
* Presenting e-mail: alexey.koronovskii@gmail.com
Section BRAIN-COMPUTER INTERFACES, COGNITIVE NAVIGATION WORKSHOP AND NEUROENGINEERING
The human brain is likely to be the most convoluted and enigmatic object for the comprehensive studies. Due to their outstanding importance and the extraordinary complexity, the investigations of the brain require the active interdisciplinary cooperation of the scientists belonging to the different branches of science. In the present work we study cognitive brain activity in visual perception of ambiguous images being just one, but a very exciting task among an enormous number of open problems of brain researches. We propose the theoretical approach associated with the experimental technique to quantitatively characterize cognitive brain activity in perception of ambiguous images. The internal noise seems to plays an important role in the visual perception of such images. Based on the developed theoretical background and the obtained experimental data, we introduce the concept of effective noise intensity characterizing cognitive brain activity and propose the experimental technique for its measurement. The developed theory, using the methods of statistical physics, provides the solid experimentally approved basis for further understanding of brain functionality. Our theoretical and experimental findings are in excellent agreement with each other. The rather simple way to quantitatively characterize brain activity connected with the perception of ambiguous images may be a powerful tool to be used, e.g., in neurotechnology, as a braincomputer interface task, and in medicine for diagnostic and prognostic purposes. We expect that our work will be interesting and useful for scientists carrying out interdisciplinary research at the cutting edge of physics, neurophysiology and medicine.
OM&P
Modern Trends in Brain-Machine Interfaces
Mikhail A Lebedev*
Department of Neurobiology, Duke University, Durham, North Carolina 27710, USA. * Presenting e-mail: mikhail.a.lebedev@gmail.com
Considerable advances in brain-machine interface (BMI) technologies bring medicine closer to solving such challenges as treatment of paralysis and sensory disabilities caused by neural trauma and diseases. Toward this goal, there is an ongoing research on prosthetic limbs controlled by brain signals, and neural stimulation systems that restore sensations by stimulating sensory brain areas. Not so long ago depictions of such BMIs could be found only in science fiction. Nowadays, even the most futuristic ideas are becoming real. Several recent studies have demonstrated direct functional connections between the brain and robotic arms. Significant achievements have been made in the systems that restore hearing, vision, vestibular function and tactile sensations to people who suffer from sensory loss.
In addition to medical applications, BMIs are being developed for augmentation of brain function in normal humans. Examples include BMIs for computer gaming, and neurofeedback systems that detect drowsiness in long-distance drivers. In the future, BMI-based technologies will lead to new means of communication and hybrid systems that merge the nervous systems with artificial neural nets.
BMIs are an interdisciplinary field that involves neurophysiologists, neurosurgeons, neurologists, robotic and electrical engineers, mathematicians and programmers. Facilitated by these collaborative efforts, BMI filed is developing very rapidly, with the number of publications growing exponentially. Current BMIs can be classified into:
1. Motor BMIs. These systems record neural signals in the brain motor areas and transform them into control commands to external devices. Aided with motor BMIs, paralyzed patients can control prostheses of the arms and legs, and motorized wheelchairs.
2. Sensory BMIs. These are systems for restoration of vision, hearing, tactile sensations, proprioception and vestibular functions. In a typical design, sensory information is collected by an artificial sensor and transmitted to the brain using electrical stimulation of the brain sensory areas.
3. Cognitive BMIs. These devices decode higher-order brain signals, such as neural representation of decision making, emotions, and even thoughts.
A large number of methods have been developed for sampling neural signals and utilizing them in BMIs. These methods can be subdivided into two major classes: invasive and noninvasive recordings.
Noninvasive BMIs are safe to use and easy to implement. These include systems based on electroencephalography (EEG), magnetoencephalography (MEG), near-infrared spectrometry (NIRS), and functional magnetic resonance imaging (fMRI). Notwithstanding a number of advantages of these methods, their generally have low information transfer rate and are susceptible to artifacts. Additionally, noninvasive BMIs often require a considerable degree of concentration from the user, leading to fatigue.
Invasive BMIs utilize brain implants placed by a neurosurgeon on the brain surface or inserted into the brain tissue. With these methods, activity of single neurons and their populations can be recorded, which enables high-quality decoding. Modern invasive BMIs incorporate multiple recording channels. Invasive BMIs are also used to stimulate nervous tissue. Several clinical trials of invasive BMIs in humans have been already conducted.
