Jan Cepek - Bionics semetral work - 2004

Brain computer interface

www.koerperwelten.com INTRODUCTION








Brain Computer interface (BCI) is a communication system that recognized users's command only from his or her brainwaves and reacts according to them. For this purpose PC and subject is trained. Simple task can consist of desired motion of an arrow displayed on the screen only through subject's imaginary of something (e.g. motion of his or her left or right hand). As the consequence of imaging process, certain characteristics of the brainwaves are raised and can be used for user's command recognition, e.g. motor mu waves (brain waves of alpha range frequency associated with physical movements or intention to move).


Brain patterns form wave shapes that are commmonly sinusoidal. Usually, they are measured from peak to peak and normally range from 0.5 to 100 µV in amplitude, which is about 100 times lower than ECG signals. By means of Fourier transform power spectrum from the raw EEG signal is derived. In power spectrum contribution of sine waves with different frequencies are visible. Although the spectrum is continuous, ranging from 0 Hz to one half of sampling frequency, the brain state of the individual may make certain frequencies more dominant.[1] Brain waves have been categorized into four basic groups.

The best - known and most extensively studied rhythm of the human brain is the normal alpha rhytm.It can be usually observed better in the posterior and occipital regions with typical apmlitude about 50 µV (P-P). Alpha activity is induced by closing the eyes and by relaxation, and abolished by eye opening or alerting by any mechanism (thinking, calculating).


Encephalographic measurements is consisted of

Electrodes read signal from the head surface, amplifiers bring the microvolt signals into range where they can be digitalized accurately, converter changes signals from analog to digital form, and personal computer proces this data.


The EEG electrodes and their proper function are critical for acquiring appropriately high quality data for interpretation. Many types of electrodes exist, often with different characteristics. Basically there are following types of electrodes:

For multichannel montages, electrode caps are preferred, with number of electrodes installed on this surface. Commonly used scalp electrodes consist of Ag-AgCl discs, 1 to 3 mm in diameter, with long felxible leads that be pluged into an amplifier. AgCl electrodes can accurately record also very slow changes in potential. Needle eectrodes are used for long time recordings and are invasively inserted under scalp. In 1958, International Federation on Electroencephalography and Clinical Neurphysiology adopted standartisation for electrode placement called 10-20 electrode placement system. This system standardized physical placement and designations of electrodes an the scalp. The head is didvided into proportional distances from prominent skull landmarks (nasion, preauricural points, inion) to provide adequate coverage of all regions of the brain. Label 10-20 designates proportional distance on percents between ears and nose where points for electrodes are chosen.Best results are in with invasive measurement techniques, where are electrodes direct on the brain and are scaning only the small location.

Picture above describes usuall electrodes placement

Amplifiers and filters

The signals need to be amplified to make them compatibile with A/D converters. Amplifiers adequate to measure these sinals have to satisfy very specific requirements. They have to provide amplification selective to the physiological signal, reject superimposed noise and interference signals, and guarantee protection from demages through voltage and current surges for both patients and electronic equipment. The basic requirements that a biopotential amplifier has no satisfy are:


Among basic evaluation of the EEG traces belongs scanning for signal distorsions called arteffacts. Usually it is a sequence with higher amplitude and differnt shape on comparison to signal sequences that doesn't suffer by anu large contamination. The artefact in the recorded EEG may be either patient- related or technical. Patinet - related artefacts are unwanted physiological signals that may significintly disturb the EEG. Technical artefacts, such as AC power line noise, can be decreased by decreasing electrode impedance and by shorter electrode wires. The most common EEG artefact sources can be classified in following way:

Patient related:


Excluding the artefact segments from the EEG traces can be managed by the trained experts or automaticially.For better discrimination of different physiological artefacts, additional electrodes for monitoring eye movement,ECG, and muscle activity may be important.

In Brain science Institute RIKEN[2] was developed the ICELAB for signal procesing which is describing the picture below.

The preprocessing tools include: Principal Component Analysis (PCA), prewhitening, filtering: High Pass Filtering (HPF), Low Pass Filtering (LPF), Subband filters (Butterworth, Chebyshev, Elliptic) with adjustable order of filters, frequency subbands and the number of subbands.

The postprocessing tools includes: Deflation and Reconstruction ("cleaning") of original raw data by removing undesirable components, noise or artifacts.

Moreover, the ICALAB Toolboxes have flexible and extendable structure with ;the possibility to extend the toolbox by the users by adding their own algorithms. The algorithms can perform not only ICA ;but also Second Order Statistics Blind Source Separation (BSS) Sparse Component Analysis (SCA), Nonnegative Matrix Factorization (NMF), Smooth Component Analysis (SmoCA), Factor Analysis (FA) and any other possible matrix factorization of the form X=HS+N or Y=WX where H=W+ is a mixing matrix or a matrix of basis vectors.

The ICA/BSS algorithms are pure mathematical formulas, powerful, but rather mechanical procedures: There is not very much left for the user to do after the machinery has been optimally implemented. The successful and efficient use of the ICALAB strongly depends on a priori knowledge, common sense and appropriate use of the preprocessing and postprocessing tools.


In the context of BCI, EEG signals are mainly analyzed in time, frequency, and time-frquency domains. Most of the research groups work in the frequency domain and extract the information characterizing mental activities from the nonparametric and parametric spectral representations of EEG. Also joint spectral preoperities of the EEG components are analyzed dor detecting particular emotional states. The relationship between the time courses of the signals coming from different electrodes serves as an indication of motor activities. Useful information can alos be extracted from particular brain configurations that can be interpreted in terms of brain states.

Time-frequency and time-scale representations of EEG signals were exploited for finding those neuronal groups that synchronize their activity as a response to a particular stimulus. From that above considerations it can be stated that mental activities, when mapped onto the time-frequency representation of EEG signals, display a picture that ilustrates the cooperative activity consists in analyzing the joint time-frequency-space correlations between the components of an EEG signal.[3]

The Brain Communicator is well-suited for patients who are severely paralyzed or locked-in, and who therefore have very limited options in their communications with others, such as ALS (Amyotrophic Lateral Sclerosis) patients on a ventilator. Patients must be cognitively intact with no history of epilepsy.

There are many apllications that are still in developement but also some of them are quite usefull. At Graz university[4] of technology was developed BCI that uses oscillatory EEG signals, recorded during specific mental activity, as input and provides a control option by its output. The obtained output signals are presently evaluated for different purposes, such as cursor control, selection of letters or words, or control of a prosthesis or orthosis. They are already working on Direct Brain Interface that recognizes voluntary activity within the brain and can be used to control assistive technologies without requiring any physical movement. This technology uses the electrocorticogram (ECoG), recorded from implanted electrodes which are placed directly on the cortex.

There are two choices. One is the patented Neurotrophic Electrode, whereby the electrode tip is implanted 5mm under the surface of the brain and the outer end is attached to amplifiers and FM transmitters located on the skull, under the scalp.[5] No wires or batteries are used. Power is provided by a power induction system similar to your toothbrush holder that charges the toothbrush overnight. This implantation requires major surgery lasting about 10 hours. The neural signals are transmitted to and processed by a computer to activate a switch or drive a cursor and hence provide communication. The other option is to implant a patented conductive skull screw that does not enter the brain. It records from local field potentials over the surface of the cortex, rather like a very pricise EEG (electroencephalogram). These signals can be used to activate a switch and hence provide communication.

Before implantation, the subject undergoes a functional MRI. This determines if there is brain activity even when there is no movement. The implant target is thus chosen. The system is also used at surgery to guide the surgeon onto target for accurate implantation.

The Army is also interested in using BCI to make faster responses possible for figter pilots. The combination of EEG signals and artifacts combine to create a signal that can be used to fly a virtual plane. One can imagine that the military would have multiple uses for a system that speeds up response times in areas as tactical maneuvering and even targeting and firing weapons. Currently,the main focus of Air Force research is for Alternative Control Technology(ACT). The goal of the ACT program is to enable communication with computers while the computer users's hands are busy with other tasks.


There are meny ways to future work exist in BCIs. Signal algorithms need to be improved.Espaecially in recognition abilities without a change in the amount of time taken to recognize the signal.

The system need to be optimized to the BCI system fo users.The question is how much the individual schould adapt to the system vs. how much the computer schould adapt to the user. User applications must improve in order to make individuals want to learn to use BCI.Current systems assume that the human schould do most of the learning and current BCI users have been trained in lengths of time up to a year. Future systems should allow more variability in training with a hevier weight on computer adaptality.

The known problems of the BCI schould be solved and the program relased as an open source program [6].In this way,researchers will be able to use the program in their experiments. IN order to achieve this goal,better software testing abilities need to be added as well the ability to display EEG data on-line.


As we can see there are many usefull applications of brain computer interface.It can be very helpfull for people with moving disabilities as human - machine interface. But it can be also used for control of human body muscles.There are also many possibilitess in military domain.Last are are the applications for making our lives easier. So one day maybe all people are wearing bci-caps and using hands only for eating.Or even without caps but with implants right in CNS.To bring this in reality it has to been developed more adaptible bci system and foremost avoid all risks.



[2] RIKEN Brain Science Institute

[3] Signal Processing Institute at Swiss Federal Institute of Technology Lausanne

[4] On-line Service Expert Labourer

[5] Institute for Human-Computer Interfaces at Graz University of technology

[6] Jessica D.Bayliss: A Flexible Brain - Computer Interface, University of Rochester,New York,2001