Brain-Computer Interface systems with EEG signals

The control signals used in the BCI systems based on electroencephalographic signals, with particular regard to the phenomena of synchronization and DE synchronization of the brain rhythms related to a mental state.

Introduction on Brain-Computer Interface systems

For more than ten years the research has been aimed at the implementation of Brain-Computer Interface systems which, according to the definition:

"It is a communication system in which the messages and commands that the individual sends to the external environment do not pass through the normal output channels of the brain, represented by peripheral nerves and muscles."

In an EEG-based BCI, for example, messages are encoded through activities electroencephalography. A BCI system, therefore, aims to decode and classify the signals deriving from brain activity to provide data commands for controlling various applications.

Elements of Brain Anatomy

The brain, together with the spinal cord, creates the central nervous system (CNS), which has the function of interpreting the signals coming from both outside and inside the body and processing the responses. Structurally, the brain is divided into three parts: a middle part called the diencephalon and two symmetrical parts (cerebral hemispheres) that make up the telencephalon. The two hemispheres are separated, up to the nucleus of white matter called the corpus callosum, by the inter-hemispherical fissure which, with the fissures of Rolando and Silvio, divides each hemisphere into four lobes: frontal, parietal, temporal and occipital.

Each lobe is, in turn, divided by furrows in convolutions in which the areas of projection, with specific motor and sensory functions. In 1909 Brodmann evaluated the distribution of neuronal layers throughout the cerebral cortex and thus characterized 52 different cortical areas. Many years later, the cortical site of many functions as described, and it was confirmed that the boundaries of these areas often coincide with those of the areas described by Brodmann. Schematically, the main brain areas can be recognized, primary sensory and primary motor areas secondary sensory areas and secondary motor areas; membership areas.

However, the concept of localization cannot be rigidly defined, since each brain area integrates with the others in a framework of general coordination. The sensorial cortex is of particular interest and motor movement, which is connected to each other by a bridge of nerve fibers. In the primary sensitive areas, the conscious perception of elementary stimuli occurs. The largest primary sensory area is that for general somatic sensitivity (somatosensory cortex), located in the post-central convolution of the parietal lobe. It consists of a typical six-layer bark in which the granular one is very developed. So, it is possible to recognize this area.

Method of Acquiring Brain Signals

There are various techniques based on different physical principles to measure brain activity. Below is a short description of these.

fMRI

The functional Magnetic Resonance Imaging is a non-invasive imaging technique that allows detecting information on brain metabolism using the BOLD signal (Blood Oxygen Level Dependent). Due to the good spatial resolution through fMRI analysis, studies have been conducted on the possible control signals for a BCI system. The low temporal resolution, the size, and the cost of the equipment do not allow the implementation of portable and accessible to all BCI systems.

MEG

Magneto Encephalography measures the magnetic field produced by internal currents. It has a good spatial and temporal resolution that has allowed the careful study of the characteristics of the brain signals. But also this technique, due to the small magnetic fields to be measured and the consistency of the measurement system is not used for the implementation of BCI systems aimed at mass consumption but rather for neurophysiological investigations.

PET

Positron Emission Tomography measures metabolic activity by detecting the activity of radioisotopes placed in the patient. It is not possible to use this technique for a 6 BCI system due to the low spatial resolution and above all, due to the invasiveness and the production cost of the radiopharmaceutical.

NIRS

Near-infrared Spectroscopy is a non-invasive and real-time diagnostic technique capable of measuring tissue oxygenation using portable instruments, relatively low cost. This technique can be used in the future for BCI applications.

ECoG

Electrocorticography, technique has an excellent spatial and temporal resolution, low vulnerability to muscle and environmental artifacts. But, given its invasiveness is not conceivable to apply it for portable BCI systems, although in the last ten years, systems that use ECoG have also been developed on human subjects.

EEG

Electroencephalography, due to good temporal resolution, ease of use and non-invasiveness is the most common choice for data acquisition in BCI systems even if it brings with it the disadvantage of a low spatial resolution due to the dispersion of the signal in the conducting

Introduction to Electroencephalography

EEG records the resulting brain activity on the surface of the brain. The electroencephalographic (EEG) signals do not derive from the sum of the action potentials of the neuronal axons but from the crossed dendritic potentials of the pyramidal cells, which are oriented vertically in the cortex, with their dendrites arranged parallel to each other. The variation in potential of one part of the cell with respect to another creates a field that gives an extracellular current, therefore a difference in potential results measurable on the surface.

Standard 10-20

To date, the use of the 10-20 standard for the registration of the EEG has been consolidated, which consists in finding the center of the head in the middle of the length n Asian-inion and place the electrodes along five transverse lines at distances of 10% or 20% of that length

Types of s-based BCI systems EEG

We can classify BCI systems according to two main characteristics:

- exogenous / endogenous

- depending on the need to have or not an external stimulus to elicit the signal of interest;

- dependent / independent

- referred to the dependence of the stimulation of the norms to elicit the desired signal

Generally exogenous BCIs, compared to endogenous ones, have the advantage of being more robust to inter-individual variability and having a high rate to transfer information at the expense of the need to have equipment for stimulation and, therefore, less comfortable use. In an independent BCI system, normal exit routes do not play an essential role; therefore, even people with serious disabilities are able to use it. These reasons mean that the focus of research is exogenous and independent BCI systems. Furthermore, we can distinguish synchronous BCI systems from systems asynchronous (self-paced) depending on whether the user is bound by a precise moment scanned externally in which to provide the command or is free to provide it when he wants.

Author: Vicki Lezama