One of the key aspects to consider when we plan to start an EEG-neurotechnology project is the hardware. There are several technical specifications and parameters that should be considered when choosing a system that best adapts to our project requirements. In this post, we will help you understand the main features of an electroencephalogram (EEG) system.
Overview of EEG technical features
When looking at EEG technology, an EEG headset or an EEG amplifier to record the electrical activity of the brain (electroencephalography), one should have a clear idea of the main requirements in order to choose the hardware that best adapts to the project specifications, giving the maximum performance within an available budget. The top performance EEG system may not be the best choice if we are not going to use their advanced features. Alternately, a low-cost device may not provide adequate performance, putting the research or application at risk.
To select the best cost-for-value, we need to understand all the EEG features that are appropriate for our project requirements. This post will help you select the best system for your specific application. This is primarily oriented to EEG, although many descriptions provided herein could be extrapolated to any biosignal amplifier.
An EEG technical features sheet typically looks like this:
Figure 1: Hardware technical specifications of the Versatile EEG from Bitbrain.
We can see that the technical features of an EEG system can be divided into three main areas:
2) The EEG amplifier, which can be subdivided into analog and digital subareas;
EEG headset sensor layer
The sensor layer is probably the most visible when deciding which EEG headset to go for. Here, we will consider aspects like the number of electrodes and if they have fixed or interchangeable positions, the type of electrode-skin contact (i.e., dry-EEG vs semi-dry EEG vs gel-EEG), and the shielding of the electrodes and wires.
1. The number of EEG electrodes
The number of EEG electrodes will determine the amount of information that we can measure from the brain. Commonly, the number of electrodes in most EEG applications ranges between 8 and 128. These numbers refer to the “recording” electrodes, and we usually need to add one reference (an electrode that is used to subtract the common mode noise from the “recording” electrodes) and one ground electrode. A higher number of electrodes will allow more detailed measurements from different brain areas. However, the increase in the number of electrodes comes with an increase in the cost and the complexity of both the experimental set-up and the data analysis.
2. The placement of EEG sensors
The placement of EEG sensors generally follows the International 10-20 System, which labels the positions according to the scalp location. Commercial EEG headsets can have either fixed or interchangeable electrodes. The systems with fixed positions do not allow moving the electrodes from one location to another, while the systems with interchangeable electrodes can be configured to accommodate different experimental setups. Systems with interchangeable electrodes are preferred for research, where it can be important to have a dense and flexible head sensor coverage. However, systems with fixed locations are optimized in terms of ergonomics and cost, and they can be preferred in some neurotechnology applications where the location of the electrodes is always the same and the priorities are usability and comfort.
3. EEG Electrode contact
Regarding the electrode contact with the skin, we can divide them into two main groups:
- Dry EEG electrodes do not require the use of any electrolytic substance, making the contact directly with the scalp.
- Wet EEG electrodes require the application of an electrolytic substance to improve the contact impedance. Different types of wet electrodes can work with electrolytic gels, saline solutions, or just tap water.
As a general rule, electrodes with higher contact impedance (e.g., dry electrodes) require amplifiers with higher technical specs to achieve similar signal quality but allow for better usability and easier set-up.
4. EEG shielding
Finally, electrode and cable shielding are features that reduce artifacts and noise, leading to higher signal quality. Electrodes can be active, which are those which have an embedded pre-amplification layer attached to the sensor, or passive, which do not have this electronic stage. Active electrodes reduce movement artifacts but can lead to more common noise than passive ones if not perfectly calibrated. Cable shielding minimizes the influence of external noises during the transmission of the signals from the electrode to the amplifier. Given the low magnitude of EEG activity, proper cable shielding is highly recommended in order to achieve a quality recording, especially when using passive electrodes.
See here for more information about the EEG sensor layer and typical values for research.
The EEG amplifier
The EEG amplifier is the part of the data acquisition system responsible for accommodating, amplifying and converting the analog electrical signals captured by the sensors into digital signals that can be processed by the computer. There are several features that characterize the operation of the amplifier.
1. EEG Sampling Rate
The sampling rate specifies the number of times that the signal is measured per unit of time, usually given in Hertz (Hz).
In origin, the EEG is an analog signal (continuous in time), but to process it with a computer we need to convert it into a digital signal (discrete in time). Since the EEG activity carries information about brain signals within a bandwidth between <0.5 and 80 Hz (usual range of brain waves and potentials), the sampling rate needs to allow measuring these frequencies.
According to the Nyquist theorem, we know that the minimum sampling rate to measure activity at 80 Hz will be 160 Hz. A common sampling rate value like 256 Hz (which is a common value in EEG amplifiers) allows full spectral resolution within the bandwidth. Higher sampling rates will provide a higher resolution on that frequency range, but no additional information for the large majority of brain processes.
The resolution of an amplifier is the number of bits used to encode the analog EEG signal voltage values into discrete numbers. This process is carried out by a component called the analog to digital converter (ADC). The more bits we use to encode these values, the more resolution the recorded digital signal will have.
3. Input Range
Related to this, we also have the input range of the amplifier, which corresponds to the maximum amplitude that can be recorded before saturation. EEG amplifiers need an input range that comprises the minimum and maximum values of EEG signals (tens of microvolts), but also those values from other physiological/mechanical processes that interfere with EEG and offset voltages, which can range in an order of magnitude of hundreds of millivolts (EMG, EOG,...).
Given that the EEG signals are within the range of few microvolts, one critical aspect in measurement is how the amplifiers minimize the effect of the electrical noise of the circuit. The input-referred noise is the noise generated by the circuitry of the amplifier even in the absence of input signal and should be as low as possible to avoid contaminating the signal. The common-mode rejection ratio (CMRR) is the capacity of the amplifier to eliminate the common voltage of its inputs. The higher this value is, the better the amplifier performance. For example, in EEG, the CMRR should attenuate the 50/60 Hz electrical noise while amplifying the pure brain activity.
Finally, the last factor which affects signal quality is the impedance contact between an electrode and the skin. The higher this impedance is, the more noise that is generated. To address this, one important parameter is the input impedance of the amplifier (the impedance of its first stage), which should be as high as possible (i.e., around 100 times the electrode-skin impedance) to avoid attenuating the signal amplitude and reducing CMRR. Also, a helpful tool to have is an impedance check and monitor (electrode-skin impedances) which will allow you to keep impedance below a certain limit. EEG amplifiers sometimes include this feature to measure the electrode impedances before starting a measurement (which assumes the impedance will not vary during the experimentation), or, better, during recording so that changes can be monitored.
See here for more information about the EEG amplifier layer and typical values for research.
Other features: Connectivity, data streaming, power
Although the sensors and the amplification layers receive the most attention, there are several other features that are important and that can determine the type of experiment that we can perform and the EEG results that we get.
1. Power supply
First, the system power supply can be cable-powered or battery-powered. Connecting the EEG amplifier directly to the power line allows unlimited recording duration, but it constraints mobility. EEG devices powered by batteries enable freedom of movement. How the batteries are connected to the device can be done in two ways, internal integrated batteries which optimize the size and weight of the device but impede its use when it needs to be recharged, and swappable batteries that can be replaced for charging, allowing more flexibility in recording but at the cost of larger size.
The connection between the amplifier and the computer can be done by means of wired or wireless communication (wireless EEG headsets). Wired communication is commonly done through USB, and although it is very robust and allows high-density transmission, the wires sometimes complicate the set-up and constrain mobility. Wireless communication for EEG data collection allows complete mobility at the cost of a slightly lower robustness in the communication, especially in cases where the electromagnetic space is saturated. In this case, reconnection protocols and data backup, either internal or removable storage, are especially useful to avoid data losses. Different technologies for wireless data transmission include WIFI, Bluetooth, or proprietary technologies.
EEG amplifiers usually feature auxiliary inputs and outputs, which can be necessary for many experiments. Analog inputs are used to record additional electrical ExG signals (e.g., EMG, ECG, EOG) or other physiological sensors (e.g., respiration, galvanic skin response), as well as other sensors that measure optical or auditory information (e.g., photodiodes, microphones). Digital input/output ports can be used to synchronize the amplifier with other external devices, like stimulus systems, response button boxes, foot pedals, robots, etc. In addition, modern EEG headsets often include kinematic or inertial sensors to facilitate the detection or filtering of head movements during EEG monitoring.
Last but not least, the output format of the EEG data files has to be taken into account. Generated files can follow either proprietary or a standardized format. Standardized formats are gaining popularity as they enhance interoperability. Systems with proprietary formats should, at least, provide software plugins to facilitate the integration of these datasets within the common software analysis tools.
See here for more information about the Connectivity Layer and Other Features.
In this post we provided an overview of the main technical features of an EEG system (including the headset and the amplifier), to understand some of the basic concepts that should support the decision of both EEG technicians or professionals without previous experience on EEG when purchasing a new hardware for their research or application.
In our series of posts, we go deeper into each of these technical aspects of the EEG systems: EEG technical features can be divided into three main areas:
The EEG amplifier (acquisition layer), which can be subdivided into analog and digital areas, addressed in this post.
The amplifier connectivity area (connectivity layer) and other complementary features like dimensions, power or weight.
We encourage you to read also other content that could be helpful for the selection of your EEG system: The differences between wet and dry EEG systems, the differences between fixed and variable electrode placement, or the artifacts and filtering tools.
We hope that all this knowledge, in conjunction with the requirements that your particular project has, will make your choice much easier. If you would like to know more about our products and how they fit within your needs, you can also contact us. Please, feel free to contact us if you have any comments or questions.
You might also be interested in
- EEG Synchronization With Other Biosensors and Software
- Bringing BCIs to the user’s home for neurorehabilitation and assistive applications
- What is BCI? An introduction to brain-computer interface using EEG signals
- Modern BCI-based Neurofeedback for Cognitive Enhancement
- EEG-based Neurotechnology for Human Enhancement and Rehabilitation
- EEG Electrode Placement: Fixed vs. Variable
- What is QEEG brain mapping and normative databases?
- Epilepsy and EEG seizure-detection
- The Use of EEG for ADHD Diagnosis and Treatment
- The Procedure and Uses of the EEG Test
- Grozea, C., Voinescu, C. D., & Fazli, S. (2011). Bristle-sensors—low-cost flexible passive dry EEG electrodes for neurofeedback and BCI applications. Journal of Neural Engineering, 8(2), 025008. DOI: 10.1088/1741-2560/8/2/025008
- Searle, A., & Kirkup, L. (2000). A direct comparison of wet, dry, and insulating bioelectric recording electrodes. Physiological Measurement, 21(2), 271–283. DOI: 10.1088/0967-3334/21/2/30
- (2003) The American Board of Registration of Electroencephalographic and Evoked Potential Technologists, Inc., American Journal of Electroneurodiagnostic Technology, 43:4, 270-273, DOI: 10.1080/1086508X.2003.11079452
- Mettingvanrijn, A. C., Peper, A., & Grimbergen, C. A. (1994). Amplifiers for bioelectric events: A design with a minimal number of parts. Medical & Biological Engineering & Computing, 32(3), 305–310. DOI: 10.1007/bf02512527
- Garipelli, G., Chavarriaga, R., & Millán, J. D. R. (2013). Single-trial analysis of slow cortical potentials: a study on anticipation related potentials. Journal of Neural Engineering, 10(3), 036014. DOI: 10.1088/1741-2560/10/3/036014
- Kemp, B., & Olivan, J. (2003). European data format ‘plus’(EDF+), an EDF alike standard format for the exchange of physiological data. Clinical Neurophysiology, 114(9), 1755-1761.