Electroenzephalography/BCI

The new technology is based on the capacitive measurement of brain signals, which was combined with special measures to suppress environmental electrical disturbances. Capacitive measurement means that there is no direct electrical contact to the head, so conductive gel is not required.
The project is supported by the BMBF (Federal Ministry of Education and Research) and is done in collaboration with Prof. Dr. Gabriel Curio of the Department of Neurology of the Charité Berlin and Prof. Dr. Klaus-Robert Müller of the Fraunhofer Institute FIRST in Berlin. The possibilities of a quick and simple recording of the brain signals for medical diagnostics as well as for new applications, e.g. control of computer games and further devices, are studied in this productive and interdisciplinary cooperation between neurology, computer sciences and electrical engineering.
A basic task of the device is the complex evaluation of signals, by means of which the relevant parts are filtered from the innumerable different brain signals using modern methods of signal analysis. The control information is extracted from these data and transmitted to the vehicle.

Electroencephalography is a routine method of medical diagnostics for assessing human brain activity. The electrical signals, induced by activity of the cranial nerves, are measured on the scalp. Thinking consists of a complex interaction of electrical signal processing and chemical signal storage in the brain. Certain parts of the brain are clearly related to activities of the body, e.g. the processing of eye signals used here which takes place in the visual cortex in the back of the human brain.

Such electrical processes occur in all parts of the brain and they interfere with each other. They can be measured by accompanying voltage on the skin. With the electroencephalography, caps are usually used for applying electrodes which can measure the voltage of a few millionth volts.

The signals of normal vision are too complicated and too weak to be used in the EEG. But if a large part of the field of view is occupied by a blinking pattern with a frequency between 8 and 15 Hz, stronger electrical signals of the visual centre are induced in the brain. They can be recorded by the EEG, filtered by the computer and processed. Such signals induced by external, defined stimuli are called visually evoked potentials (VEP).

The signals are represented as a map (see figure). In medical applications, the doctor makes a diagnosis by comparing the brain activity of the ill person with that of healthy ones.

Electrodes are usually used for EEG recording which minimise the contact resistance between electrode and skin surface and guarantee a stable, reliable contact. So-called silver / silver chloride electrodes meet these demands best. By means of a special conductive gel, which is applied in the hair between electrodes and scalp, the electrical resistance can be minimised and the derivation of the EEG can be optimised. Since an EEG usually requires some dozens of electrodes (figure on the right), all electrodes have to be applied to the represented cap carefully one by one and tested. For particularly good contacts, the surface of the scalp has to be roughened mechanically or abraded. The time expenditure and the required know how are considerable, so the EEG derivation is generally left for specialists and the derivation of brain waves for controlling machines so far appeared to be too complex.

The effect that there are also charge displacements on the body surface due to brain activity is used for the capacitive measurement of the EEG. This change of the charge can in turn affect the charge on a metal plate close to the body. Since this electrical plate does not require direct electrical contact to the body, it can be isolated from the body. The measurement of the capacitive EEG (cEEG) is therefore also possible through the hair. A supersensitive signal amplifier is connected to this plate which amplifies the brain signal and processes it, so it can be displayed on the screen later on. Plate, amplifier and further signal processing electronic systems are integrated in the compact electrodes (30 mm in diameter, approx. the size of a 2 euro coin). The electrode is a bit larger than a standard EEG electrode.

28 electrodes are integrated in one helmet for covering different brain areas. The electrodes can be adjusted mechanically to allow the adaptation to various head shapes.

A brain computer interface is meant to allow the direct information flow between brain and computer. For the technical implementation, signals from the brain have to be recorded. There are several methods which can be considered for this purpose. The method we use is the relatively cost-effective and simple recording of electrical brain signals (EEG) by means of capacitive electrodes.

In our system, the signals of the visual region of the brain are induced by looking at two blinking chequerboard patterns on the computer screen. If the controlling person concentrates on the blinking pattern on the left side, the vehicle is supposed to drive to the left and accordingly to the right if looking at the pattern on the right side. If none of the chequerboard patterns is looked at, the vehicle drives straight ahead. The signals for left and right differ in terms of blinking frequency (Martin Oehler, Peter Neumann, Matthias Becker, Gabriel Curio, Meinhard Schilling, “Extraction of SSVEP Signals of a Capacitive EEG Helmet for Human Machine Interface”, Proceedings of the 30th Annual International Conference IEEE EMBS, Vancouver, Canada, 2008).

The recorded signals are amplified and disturbances are filtered. The signals are then transmitted to the computer (wireless or via cable) and evaluated. The intention of the controlling person can only be determined from the various signals and converted into control signals for a machine by means of a computer programme. The control commands are then transmitted to a vehicle model via radio.

Advanced BCI systems, e.g. studied by our colleagues in Berlin, are only based on electrical brain signals influenced by will in the motor centre which are, however, much weaker and which are recorded with conventional EEG electrodes.

At the TU Braunschweig a light weight electrodes-helmet was build and used to simplify by mobile and wireless transmission the analysis of the electrical brain activities significantly.
With the EEG-helmet brain signals can be measured directly without electrical contact between electrodes and head. This is possible because of the new kind of capacitive electrodes.

It is already the third generation of electrode-helmets developed by the Institut für Elektrische Messtechnik und Grundlagen der Elektrotechnik (EMG). The new helmet weighs only 500 gram, uses 24 electrodes and can be adjusted individually to different users. Using 3D printers it is no problem to produce fast high numbers of helmets. “The helmet is not pre-commercial yet, but it can be used for studies at clinics and surgeries.” says Prof. Meinhard Schilling of the EMG.

The new technology is distinguished by a very good signal-to-noise-ratio. It combines the capacitive measurement of the brain signals and special elimination methods of electrical interference from the environment. Using the capacitive electrodes no direct contact to the head is needed preventing time-consuming preparations with contact gel. This helmet is directly mounted and the EEG is measured. [1]

For a few years the Institut für Elektrische Messtechnik und Grundlagen der Elektrotechnik has investigated biomedical sensor technology and is cooperating with the Charité Berlin and the information scientists of the TU Berlin. The aim is to achieve a fast, simple and wireless instrument to record brain signals that can be used in the medical diagnosis like for example to diagnose epilepsy quickly or in the research of sleeping laboratories. Furthermore, new applications as human machine interfaces, like steering computer games or controlling other devices are thinkable just as potential usage for neuro-ergonomics. In future the concentration state of persons in critical situations can be measured, for example in the cockpit of an airplane or at the control panel of a large factory building.

[1] Extraction of SSVEP Signals of a Capacitive EEG Helmet for Human Machine Interface Martin Oehler et. al., Proceedings of the 30th IEEE EMBS, Vancouver, 4495–4498, 2008 .

Electroencephalography is a routine method of the medical diagnostics to appraise the human brain activities. The electrical signals caused by activities of the neurons in the brain are measured on the scalp. Thinking is a complex cooperation of electrical signal processing and chemical signal storage of the brain. Certain parts of the brain are closely related to certain activities of the body for example the signals caused by the eyes are treated in the visual cortex at the back of the head.

Those electrical actions running all over the brain and their signals are superposed outwards. They are measureable due to the electrical potentials on the skin. Usually there is an elastic cap that takes care for a good contact of the electrodes for a record of an electroencephalogram, though the electrodes can measure the potentials that are a few millionth Volts high.

The signals of normal sight are too complicated and too weak to be measured with an EEG. But if the eyes are focused on a pattern flashing with a frequency between 8 and 15Hz, higher electrical potentials are generated in the visual centre of the brain. This procedure is called pattern-reversal stimulation. Those signals can be recorded with EEG, filtered by the computer and be processed. Those signals provoked by an external specified impulse are called visual evoked potentials (VEP), or when they are steady-state during a long time stimulation steady-state visual evoked potential (SSVEP).

For some people usual communication is not possible anymore because of serious illnesses of the central nervous system (locked-in syndrome).  For such cases a communication due to the analysis of brain signals (brain computer interfaces by EEG) was developed. Such systems are based on the principle that a computer gives a selection of letters and the patient chooses a letter by generating a measureable signal in the EEG. Such an application was implemented and tested with the capacitive EEG-helmet, too.

The selection of the letters was realized with flashing chessboard patterns. The influence of different color combinations on the successful detection by different test persons was analyzed, too. There are significant differences between the individual results [2]. Moreover, the successful detection is dependent on the ability for concentration and the personal subjective most comfortable color combination.

The color patterns are assigned to the letters and by focusing on one certain pattern the flashing frequency can be detected in the visual centre of the brain. In this manner the letter can be recognized. By doing so letters can be written one by one and can built words to communicate.

[2] Subject response variability in terms of colour and frequency of capacitive SSVEP measurements. Marianne Gerloff, Meinhard Schilling, Proceedings BMT 2012, Jahrestagung, Jena, 46. DGBMT, 95-98, 2012.