Study of Hardware Direction of EEG Acquisition Device in BCI

Chen Liu

School of Electronic Information and Electrical Engineering, Yangtze University, Jingzhou, Hubei, 434023, China

Keywords: Brain-Computer Interface (BCI), Electroencephalogram (EEG), Acquisition Device, Hardware Technology.

Abstract: Brain-computer interface (BCI) has been the focus of research all over the world and is an important way for

the human brain to interact with external devices. Electroencephalogram (EEG) acquisition device, as the

main way to collect EEG signals, is an indispensable part of this field. The research results of various countries

and companies in related fields are endless, and there is a wide range of application space in various fields in

the future. At present, the main composition of EEG acquisition devices has tended to be stable, but there are

still shortcomings in portability, accuracy, and stability in function and performance. In order to clarify the

general situation of the current EEG acquisition device and the main research direction in the future, this paper

will divide its overall structure into electrical signal acquisition sensor, which includes seven parts: pre-

processing circuit including filter and signal amplification circuit, analog-to-digital converter, signal

processing unit, data storage module, data transmission interface and power management module for analysis

and description. At the same time, the current, more advanced products, the shortcomings of current research,

and the direction of optimization are described and analyzed, and finally, the development trend is prospected.

1 INTRODUCTION

Brain-computer interface technology has great

application potential in many fields. At present, it has

been used in the treatment of patients with paralysis,

neurological, and other related diseases. It can be

predicted that in the future, many diseases that are

currently incurable are expected to be improved and

even solved. Brain-computer interface refers to the

use of modern computer science, biological science,

brain science, electronic information, and other fields

of technology to create a human brain or other

biological brain and electronic devices between the

information interaction channel so as to achieve the

exchange and control of information and data. The

first step is to capture brain activity signals, including

Electroencephalogram (EEG) acquisition, functional

magnetic resonance imaging, near-infrared imaging,

and so on. Among them, EEG signal acquisition is a

way with the highest precision and fastest response,

which is conducive to more real-time and accurate

interaction and control with computers, meeting

people's needs for brain-computer interaction in some

scenes. Therefore, the acquisition of EEG signals has

https://orcid.org/0009-0008-4358-8162

received more attention and has become the focus of

brain signal acquisition.

In 1924, the German psychiatrist Hans Berger was

the first to collect rhythmic point changes from the

human scalp and named it electroencephalogram.

This officially opened the embryonic stage of brain-

computer interface theory. The accuracy and safety of

early EEG acquisition devices are low. With the

continuous development of brain science,

neuroscience, and computer science, in the 1990s,

Europe and the United States began the brain program

one after another, which rapidly promoted the

development of brain-computer-related technologies

and reached unity in the concept of brain-computer

interface. At the beginning of the 21st century, brain-

computer interface technology entered a period of

explosion, and various non-implanting minimally

invasive and non-invasive EEG signal acquisition

systems and flexible implanting acquisition systems

that pursue high precision and long-term acquisition

have also been developed. In 2015, Mniev et al.

proposed a study related to the long-term

performance improvement of neural prosthetics by

flexible neural implants, which further promoted the

development of invasive brain-computer interfaces

Liu, C.

Study of Hardware Direction of EEG Acquisition Device in BCI.

DOI: 10.5220/0013679300004670

Paper published under CC license (CC BY-NC-ND 4.0)

In Proceedings of the 2nd International Conference on Data Science and Engineering (ICDSE 2025), pages 98-103

ISBN: 978-989-758-765-8

(Minev et al., 2015). In 2024, Yanshan University

proposed a method of mixed EEG and myoelectric

examination to improve the recognition rate of

features. The equipment systems and methods of

EEG acquisition are constantly updated and

optimized, which will also promote the application of

brain computers in more scenarios.

This paper mainly reviews the research status and

future development trend of EEG acquisition

technology and equipment, summarizes the structure

of the current EEG acquisition system, as well as the

optimization direction of different methods, and

compares the currently developed acquisition

systems. In the end, the shortcomings of current EEG

acquisition technology and the key problems that

need to be solved are summarized, and the future

development of related directions is prospected.

2 PRINCIPLE

An electroencephalogram (EEG) is an image drawn

by the electrical signals collected by various EEG

acquisition devices in the head of the organism, which

shows the spontaneous and rhythmic electrical

activities of the brain cells of the organism. EEG

acquisition device is a kind of precision measuring

instrument that needs to protect the test target and

ensure the accuracy of the measurement signal. In

addition, it is necessary to consider the power

consumption and volume of the entire measurement

system as much as possible so as to reduce costs. The

current brain-computer interface technology is

mainly classified into three categories: invasive,

semi-invasive, and non-invasive, and the acquisition

of electrical signals at different locations. They

generally have the following main components:

electrical signal acquisition sensor, including

filtering, signal amplification circuit pre-processing

circuit, analog-to-digital converter, signal processing

unit, data storage module, data transmission interface,

and power management module. The approximate

structure is shown in Figure 1.

Figure 1: Structure of EEG acquisition device

The electrical signal acquisition sensor is

responsible for collecting the original signal of the

corresponding position of the organism into the

brain's electrical equipment. The organism can be

regarded as a conductor with high impedance, so the

brain signals that can be detected are rather weak, and

at the same time, they will be interfered with by

various external factors and other strong models of

the organism itself, and there will be large DC offset,

which has a great impact on direct acquisition (Zhou,

Du, and Duan, 2022). At present, there are three main

lead modes in scientific research, namely unipolar

lead, bipolar lead, and average lead. The unipolar lead

connects the head measurement sensor and the

reference electrode located in the earlobe or mastoid,

respectively, to the amplifier. It has the advantage of

higher amplitude and more stable measurement

locations when recording the absolute value of

changes in brain potentials. However, the reference

potential at the earlobe or mastoid cannot maintain

zero potential, so it is susceptible to interference from

other bioelectrical signals so that abnormal signals

other than the EEG signal at the target location can be

measured. The bipolar leads do not use reference

electrodes, but two measurement sensors at different

positions are connected to both ends of the amplifier

to record the difference in potential change between

the two places. The advantage of this is that it can

effectively suppress the interference of common

mode signal to the acquisition target position signal

and improve the signal-to-noise ratio of the

acquisition system, but it can not measure the

absolute value of point change. The average lead is to

average the signals collected by multiple EEG in the

head and connect them to one end of the amplifier as

the average reference electrode, which is used to

Study of Hardware Direction of EEG Acquisition Device in BCI

replace the reference electrode in the earlobe or

mastoid in the single lead so as to reduce the influence

of the reference electrode on the collected EEG signal.

After the original EEG signal is obtained by the

sensor, it needs to be processed by the pre-processing

circuit. For example, the above EEG signal itself is

quite weak and susceptible to common-mode

interference. Common mode interference is caused by

many reasons; the most common is the human body's

own physiological activities generated by other signal

interference. Therefore, a good performance filter is

needed to eliminate the effect of interference as much

as possible. In this circuit, it is also necessary to

amplify the originally weak EEG signal. This part of

the circuit determines the quality of the signal

collected by the whole EEG acquisition device, so it

is also the key research part of the hardware circuit in

the current EEG acquisition.

EEG-related research requires high accuracy of

data. EEG signal acquisition device to this part of the

circuit requirements are also quite high. In addition to

the high-resolution requirements (usually not less

than 16 bits), but also a high demand for the sampling

rate of the device, the current research field of EEG

acquisition device sampling rate is basically above

1kHz, but also to choose less noise devices.

At present, most processor modules are

implemented on platforms such as Digital Signal

Processing (DSP) or Field Programmable Gate Array

(FPGA), and microcomputers are also used. This part

of the module mainly sends the processed signal to

the host computer or other peripherals through the

communication interface for subsequent calculation

processing or direct control of other peripherals.

3 TECHNOLOGY STATUS

In the aspect of non-invasive EEG acquisition devices,

due to their advantages of relative convenience and

high security, they have great potential in various

civil industries. Traditional EEG acquisition devices

usually have 8-128 acquisition channels with a

collection frequency of 100-1000Hz and use earlobes

and papillae as reference electrodes (Gao et al., 2015).

However, with the continuous improvement of

demand and technology, the relevant performance of

the instrument is also improving. In May 2023,

Cumulus Neuroscience published a certified dry-

sensor EEG headset with 16 electrodes, a sampling

frequency of up to 500Hz, and a sampling bandwidth

of 0.5Hz to 250Hz. Such conditions enable the

recording and analysis of more complex EEG signals

with low latency (Julie, 2023). It has great advantages

in the number of channels, and the use of electrodes,

and the dry sensor maintains high accuracy while

avoiding the cleaning problems that are prone to the

use of traditional electrodes. In the same year, non-

invasive EEG acquisition devices were also used in

medical applications, such as Cognixion combined

with AR technology to develop Cognixion ONE

Axon (COA) to help improve the quality of life of

patients with muscular dystrophy, traumatic brain

injury, and other diseases. The device has six infinite

EEG sensors and one infinite eye movement sensor to

record brain signals and eye movement signals, with

a sampling frequency of 250Hz and a sampling

bandwidth of 0.5Hz to 125Hz. Compared with

traditional electrode caps, it has great advantages in

terms of comfort, portability, and flexibility (Sagar,

2023). China's Changzhou Boricon Technology Co.,

Ltd. has also developed a non-invasive EEG

acquisition device, NeuSen W, with 128-1024

channels, 0.5Hz to 70Hz sampling bandwidth, and

16kHz sampling rate, and equipped with a nine-axis

motion sensor, which can eliminate motion noise. It

has the advantages of good portability, strong signal

stability, and strong shielding ability (Neuracle,

2020). At the same time, it can be connected through

WIFI, supporting multiple devices for simultaneous

acquisition and multi-end synchronous EEG imaging

and data interaction analysis. The above equipment

parameters are shown in Table 1.

Table 1: Hardware information of non-invasive EEG acquisition device

Name Number of

channels

Sampling

fre

uenc

/Hz

Sampling

andwidth/Hz

Reference electrode

Traditional EEG acquisition device 8-128 100-1000 0.1-1000 Earlobe, mastoid, average

reference electrode

-sensor EEG headset 16 500 0.5-250 avera

e reference electrode

COA 6+1 250 0.5-125 avera

e reference electrode

NeuSen W 128-1024 16

0.5-70

In addition to non-invasive technologies, there are

also companies that focus on invasive technologies

and related equipment, such as Neuralink, BrainGate,

and others. Neuralink used monkey brains to control

ICDSE 2025 - The International Conference on Data Science and Engineering

100

computers in 2019, implanted devices in the brains of

live pigs in 2020, and completed the first device

implantation of human brains in 2024, where a brain-

computer interface system the size of a coin is

implanted in the brain region to record EEG signals,

but this technology is not yet mature. A few weeks

after surgery, part of the wiring connected to the brain

falls off, resulting in a loss of signal acquisition, and

the system cannot work properly (Xu, Xue, and Xu,

2024). In February 2022, Xuanwu Hospital Capital

Medical University completed China's first closed-

loop neurostimulator CNS 061 and used it to achieve

the implantation of the first patient with Parkinson's

disease. Its acquisition system has 16 channels, which

can connect various types of probe electrodes or chip

electrodes, and has the function of 16 channels for

stimulation and eight channels for collection. Its

communication function can also realize Bluetooth

communication and near-field communication,

through which the collected information can be

transmitted to the relevant monitoring equipment and

synchronized to the corresponding server through

WIFI. EEG acquisition devices with relatively less

trauma have also made great progress. In April 2023,

brain technology company Precision Neuroscience

completed the implantation of an EEG acquisition

device with a wound of just 1mm. The device is a

flexible membrane of just 1cm square, containing

1024 channels. Higher accuracy of EEG

measurements was achieved with less overall

biological damage. The technique has been implanted

in three patients with tumors in the language region

of the brain (Ho et al., 2022). In the less invasive

technology, in 2022, the personnel of relevant

research institutions in China implanted the EEG

acquisition device through functional magnetic

resonance to precisely locate the target location and

realized the use of the device to complete the

character output of the brain-computer interface, and

the speed reached 12 characters per minute. The

device used two Utah electrodes made of miniature

silicon, with a total of 192 channels. And the

equivalent information transmission rate of each

channel reaches 2 bits per minute.

4 TECHNICAL CHALLENGES

AND IMPROVEMENTS

With the continuous research and exploration of

scientists in this field, the relevant technologies of

EEG acquisition devices are also maturing step by

step. At present, there are still many difficulties to be

overcome, from the research field to the application

of various fields. The current development mainly

focuses on the following:

1. Portability. In the future, the brain-computer

interface will be applied in daily life, and the daily

brain-computer interface in many application

scenarios needs to be portable and comfortable

enough while being wearable, so the volume and

weight of the device and even the material will have

relatively demanding requirements. At the hardware

level of the EEG acquisition device, power

consumption is an important point that needs to be

optimized. At present, the weight of the power supply

in wearable EEG acquisition devices accounts for a

large part, so reducing power consumption can

effectively reduce the weight of the device itself and

thus reduce the pressure borne by the wearer (Le et

al., 2022). In terms of volume, the current

optimization direction is to integrate the entire system

on one chip as much as possible, thereby reducing the

volume of the overall device (Guo et al., 2023).

2. Information processing and transmission rate.

A major shortcoming of the current wearable EEG

acquisition device is that the transmission rate is

relatively low, which cannot meet the needs of the

current application environment. At present, the rapid

development of communication technology has led to

the commercial use of 5G communication and the

research and development of 6G. In the future, the

problems in communication will be further solved

with the development of related fields.

3. Hardware performance. The bandwidth,

acquisition rate, accuracy, and speed of analog-to-

digital conversion mentioned above are the keys to

determining the quality of the signal finally collected

by the equipment, and the sufficiently accurate and

reliable signal is the basis of the subsequent

processing and decoding process. At present, from the

product comparison of various EEG acquisition

device manufacturers, it can be seen that the

equipment at the front end of the industry has obvious

advantages in its hardware configuration. With the

further improvement of scientific research in this field,

the understanding of EEG signals is becoming more

and more profound, and the hardware will be more

targeted and optimized for the acquisition of EEG

signals, such as more accurate positioning of specific

signals, widening of the spectrum range of EEG

signals, and the development of electronics will also

derive hardware with ideal sampling rate and

bandwidth. At present, the acquisition signal is still

subject to a lot of interference in the acquisition

process, and the power frequency interference of

hardware equipment is one of them. In future research,

Study of Hardware Direction of EEG Acquisition Device in BCI

101

we should try to remove the corresponding

interference signal without affecting the signal of the

same frequency in the original EEG signal (He et al.,

2020).

3. Expansion of sensors and peripherals At present,

many brain-computer devices have assembled a

variety of peripherals, such as the nine-axis motion

sensor mentioned above, and currently, in the field of

application, there are already peripherals such as the

use of EEG acquisition device to control robotic arms,

character input, and drones. In order to optimize the

collected information or adjust the wearing problems,

flexible sensors for ear acquisition, combined with

myoelectric acquisition devices, etc., are also

constantly being developed. In the future, the

development of brain peripheral devices will also

extend to various industries, such as the game

industry, transportation, industry, furniture, and even

other strategic fields, which will further change

people's lifestyles (Ruan et al., 2024).

5 CONCLUSION

The current EEG acquisition device technology is still

improving, the overall structure of the EEG

acquisition device will not change much in the short

term; the key area is mainly in the pre-processing

circuit and analog-to-digital conversion circuit, the

acquisition rate, the size of the bandwidth range, the

accuracy and speed performance of the analog-to-

digital converter, to a large extent, determine the

quality of the final EEG signal. The current

mainstream equipment has a high measurement

accuracy, as small as possible input noise, and a fast

sampling rate, but there are still some shortcomings,

in the medical aspect, invasive EEG acquisition

devices also have long-term connection problems.

For future applications in various fields, the volume

and weight, information processing and transmission

rate, hardware performance, sensor, and peripheral

expansion are all areas that need to be further

developed and optimized for EEG acquisition devices

and even brain-computer interfaces to meet the needs

of wireless, portable, comfortable and high-density in

future daily or medical fields.

The current EEG acquisition device still has

unsolved problems in some aspects because of

hardware and other problems. There are still

problems with the stability of EEG sensors such as

electrodes and probes, and it is difficult to maintain

the acquisition accuracy or other effects during long-

term wearing or implantation. In the case of

interference by other signals of the organism itself

and power frequency interference, the signal quality

will also be affected in the process of acquisition.

After continuous optimization and improvement to

solve these problems in the future, EEG acquisition

devices and even brain-computer interfaces will have

a wider range of applications.

It can be predicted that EEG acquisition devices

will still be one of the key research directions in the

field of brain-computer interface. When the

technology becomes more mature in the future, EEG

signals can be more accurately observed to better

interpret the intention of the human brain and better

establish the information interaction channel between

the human brain and electronic devices so as to treat

patients with disabilities or neurological diseases in

the future medical industry. In terms of entertainment,

it is combined with movies and games to create new

entertainment ways, and in terms of life, it is

combined with smart homes to further optimize life

experience and so on. Brain-computer interfaces may

become a new type of terminal to replace mobile

phones in the future.

REFERENCES

Cumulus Neuroscience, 2023. Cumulus Neuroscience

receives FDA 510(k) clearance for an award-winning,

first-in-class neurophysiology platform for at-home

use. Available at:

https://cumulusneuro.com/articles/2023_05_04/

[Accessed 16 Feb. 2025].

Gao, X., Wang, Y., Chen, X., 2021. Interface, interaction,

and intelligence in generalized brain-computer

interfaces. Trends in Cognitive Sciences, 25(8), pp.

671–684.

Guo, J., Zhang, S., Xue, K. and Ji, B., 2023. Key hardware

technologies and future applications of brain-computer

interface. Unmanned System Technology, 5, pp. 1–16.

He, Q., Hao, S., Si, J., Wu, Y. and Cheng, J., 2020. A review

of hardware systems of EEG acquisition equipment for

BCI. Chinese Journal of Biomedical Engineering, 6, pp.

747–758.

Ho, E., Hettick, M., Papageorgiou, D., 2022. The layer

seven cortical interfaces: a scalable and minimally

invasive brain-computer interface platform. BioRxiv,

2022.01.02.474656.

Minev, I.R., Musienko, P., Hirsch, A., 2015. Electronic

dura mater for long-term multimodal neural interfaces.

Science, 347, pp. 159–163.

Neuracle, 2020. NeuSen W series wireless EEG acquisition

system. Available at:

http://www.neuracle.cn/productinfo/148706.html

[Accessed 7 Jun. 2020].

Ruan, M., Zhang, L., Ling, J., Yuan, T., Zhang, X., Zhu, C.,

et al., 2024. Developments in the field of brain-

ICDSE 2025 - The International Conference on Data Science and Engineering

102

computer interfaces in 2023. Life Science, 1, pp. 39–

47.

Sagar, V., 2023. Cognixion gets FDA breakthrough device

designation for Cognixion ONE Axon. Available at:

https://www.nsmedicaldevices.com/news/cognixion-

gets-fda-breakthrough-designation-for-cognixion-one/

[Accessed 16 Feb. 2025].

Xiao, L., Zhu, Z., Yuan, S., Liu, Z., Gao, L., Ye, J. and

Zhang, X., 2022. Portable multi-channel EEG signal

acquisition system. Chinese Journal of Medical

Devices, 4, pp. 404–407.

Xu, Y., Xue, L. and Xu, Y., 2024. Global BCI industry

development status. Shanghai Renmin University

Monthly, 10, pp. 52–53.

Zhou, L., Du, Y. and Duan, D., 2022. Study on the

extraction method of weak EEG signal. Automation

Technology and Application, 10, pp. 97–100.

Study of Hardware Direction of EEG Acquisition Device in BCI

103