Electroencephalography (EEG)

Recording electrical signals on the scalp surface:
The investigation of the human brain activity

The first technology with which human brain activity could be measured was developed by Hans Berger in the first half of the 20th century at the University of Jena in Germany.
Electrical signals produced by nerve cell activity in the human brain are measured by electroencephalography (EEG) on the surface of the scalp. Measuring the summed electrical activity of the nerve cells in the brain on the scalp has the advantage that it is non-invasive to the research subject.

Averaging: isolating event-related potentials (ERP)

In order to extract the activity generated in a specific brain region and assumed to be associated with a particular cognitive function, the summed EEG signal is processed with special mathematical methods.

Since the 1960's, the averaging method has been employed to the EEG signal to isolate so-called event-related potentials (ERP). ERPs are signal changes in the measured electrical potential that are temporally bound to a sensory, motor, cognitive or emotional event. By averaging the EEG-epochs temporally centered around specific events which occur repetiously in an experiment, it is possible to isolate potentials characteristic for certain cognitive functions. For example, the error-related negativity occurs immediately following erroneous responses in reaction time experiments.  The ERN is a negative potential deflection which reaches its maximum approximately 50-100 milliseconds following an erroneous response, has a frontocentral scalp distribution and can have a magnitude of up to 15 µV.      

Schematic depiction of the measurement and averaging of event-related potentials. Singular "waves" temporally associated with a specific experimental event are usually labeled according to their polarity and latency from the triggering event. (e.g. N100 = negatively poled ERP with peak latency at roughly 100 milliseconds; P300 = positively poled ERP with peak latency at roughly 300 milliseconds). Negative potential deflections are often presented facing upwards (© MPI for for Human Cognitive and Brain Sciences, Leipzig)

To top

Requirements, advantages and limitations of ERP measurement

ERPs can be measured on the scalp surface only under certain conditions. First of all, a sufficient number of synchronized neurons must exhibit identical activity patterns.  In order for the aggregated electrical signal to be clearly transmitted to the scalp surface, the geometrical configuration of the neurons must be more or less parallel.  Signals from structures located deep within the brain are particularly difficult to measure.  Furthermore, it is not possible to determine the neuroanatomical generator of the ERP from the measured potential change alone (so-called inverse problem).  Thus, the real advantage of EEG and ERP is the virtually unlimited temporal resolution.  In contrast, the spatial resolution is relatively low.        

Important information such as amplitude oscillations or signal fluctuations over the course of the single events is lost by averaging the event-locked EEG signal to isolate ERPs. 
However, due to technological developments it is now possible to better analyze the dynamics of electrical brain activity over the course of an experiment and in each experimental trial. Such new methods of source separation include, for example, independent component analysis (ICA) and time frequency analysis.     

To top

EEG-Experimental Procedure

Electrodes which record the summed electrical brain potentials are placed on subjects' scalps in EEG experiments.  Usually, the electrodes are mounted in a type of hat, which resembles a bathing cap.  In order to accurately measure the minute voltage fluctuations of the EEG signal, the skin-electrode impedance must be held as low as possible. This is insured by applying a special electrode gel.  Ideally, subjects' scalps are degreased by means of a hair wash - without conditioners because these can have a non-conductive effect.  For EEG experiments, subjects are seated in an electrically and acoustically shielded chamber. While in the chamber, stimuli such as images or tones are presented via a computer monitor or loudspeakers. In most experiments, subjects are instructed to respond to the presented stimuli according to certain rules by means of a button press. EEG is recorded for analysis at a later date while the subject performs the task. EEG investigations are completely harmless for subjects. 


Technical Equipment

The EEG laboratory is equipped with two 32-channel EEG-amplifiers (BrainAmp MR+) which are compatible with magnetic resonance scanners (MRI).  The amplifiers enable simultaneous EEG and fMRI data collection.  64 Ag/AgCl electrodes are embedded in the EEG caps so that the scalp surface is nearly entirely covered. 


Above: Typical temporal course of the error-related negativity (ERN) measured at a frontocentral EEG electrode.

Below: Spatial distribution of the ERN across the scalp. Blue colors indicate the relative amplitude strength of the negativity.

Regarding the images see also:
Paragraph Averaging
and Junior Research Group Cognitive Neurology