Maiseli et al.

Brain Informatics (2023) 10:20 Brain Informatics

https://doi.org/10.1186/s40708-023-00199-3

REVIEW Open Access

Brain–computer interface: trend, challenges,

and threats
Baraka Maiseli1*, Abdi T. Abdalla1, Libe V. Massawe1, Mercy Mbise2, Khadija Mkocha1, Nassor Ally Nassor1,
Moses Ismail1, James Michael1 and Samwel Kimambo1

Abstract
Brain–computer interface (BCI), an emerging technology that facilitates communication between brain and com-
puter, has attracted a great deal of research in recent years. Researchers provide experimental results demonstrating
that BCI can restore the capabilities of physically challenged people, hence improving the quality of their lives. BCI
has revolutionized and positively impacted several industries, including entertainment and gaming, automation
and control, education, neuromarketing, and neuroergonomics. Notwithstanding its broad range of applications,
the global trend of BCI remains lightly discussed in the literature. Understanding the trend may inform researchers
and practitioners on the direction of the field, and on where they should invest their efforts more. Noting this signifi-
cance, we have analyzed 25,336 metadata of BCI publications from Scopus to determine advancement of the field.
The analysis shows an exponential growth of BCI publications in China from 2019 onwards, exceeding those
from the United States that started to decline during the same period. Implications and reasons for this trend are
discussed. Furthermore, we have extensively discussed challenges and threats limiting exploitation of BCI capabilities.
A typical BCI architecture is hypothesized to address two prominent BCI threats, privacy and security, as an attempt
to make the technology commercially viable to the society.
Keywords Brain–computer interface, Brain activity, Machine learning, Neurological disease, Signal processing,
Augmented reality

1 Introduction healthy people that may not always be available. Chal-

Naturally, humans use their peripheral nerves and mus- lenged by the limitation, scientists and researchers have
cles to interact with the outside physical environments developed a brain–computer interface (BCI) technology
in carrying out the desired actions. This necessity and that can transform brain signals into human actions inde-
premise for survival comes with a cost for people with pendent of the peripheral nerves or muscles.
severe neurological diseases, including amyotrophic lat- BCI, also called brain–machine interface, provides
eral sclerosis and brainstem stroke. These people cannot direct communication between brain and external
control external devices, thus requiring assistance from devices, such as computers and robotic limbs [1–4].
Bypassing the conventional communication channels
for different tasks (e.g., vision, movement, and speech),
*Correspondence:
Baraka Maiseli BCI links the brain’s electrical activity and the external
[email protected] world to augment human capabilities in interacting with
1
Department of Electronics and Telecommunications Engineering, the physical environment [1]. BCI provides a non-mus-
College of Information and Communication Technologies, University
of Dar es Salaam, 14113 Dar es Salaam, Tanzania cular communication channel and facilitates acquisition,
2
Department of Computer Science and Engineering, College manipulation, analysis, and translation of brain signals to
of Information and Communication Technologies, University of Dar es control external devices or applications.
Salaam, 14113 Dar es Salaam, Tanzania

© The Author(s) 2023. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which
permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the
original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or
other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line
to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory
regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this
licence, visit http://creativecommons.org/licenses/by/4.0/
Maiseli et al. Brain Informatics (2023) 10:20 Page 2 of 16

Since its conception in 1973 by Vidal [5], BCI has Signal processing
Signal Feature Feature Feature
remained an active area of research with enormous acquisition extraction classification translation
promising opportunities [6–14]. Researchers have, for
instance, reported remarkable achievements demonstrat-
ing that BCI can efficiently restore capabilities of peo- Feedback BCI
ple with disabilities, such as those with schizophrenia Application

symptoms (psychosis, emotional disturbances, and cog-

nitive dysfunction) [15–21]. Generally, BCI applications Fig. 1 Main components of the brain–computer interface (BCI)
system
can be classified depending on the industry: gaming and
entertainment [22–24], security and authentication [25],
healthcare [21], education [26–28], advertisement and
neuromarketing (commercial marketing using principles
2.1 Signal acquisition
of neuroscience and cognitive science) [29–33], and neu-
This component comprises an electronic device with
roergonomics (application of neuroscience to ergonom-
electrodes for acquiring brain signals (oscillating elec-
ics) [34, 35]. Given its cross-cutting nature across many
trical voltages caused by biological activities of the
aspects of developments, BCI may remain an attractive
brain) that define its neurophysiological states. Signal
and a competitive research area over a longer period.
acquisition involves capturing of electrophysiologi-
Despite the promising applications of BCI, there has
cal signals that represent specific activities of the brain
been a paucity of studies on the future of this technol-
(e.g., movement, speech, hearing, and vision). Most
ogy and its possible threats when applied to humans.
BCI systems, including the commercial ones, deal with
The present study covers typical BCI threats, including
the following electrophysiological signals: electroen-
medical safety, privacy, ethics, and security. We stimulate
cephalography, brain’s electrical activity measured with
discussions within the scholarly community on the readi-
electrodes placed on the scalp [36, 37]; electrocorti-
ness to adopt the BCI technology and accommodate its
cography [38–40], electroencephalographic signals
challenges and potential threats. Furthermore, because
measured directly with electrodes placed on the surgi-
the natural working principles of the brain are not com-
cally exposed cerebral cortex; local field potential [41],
prehensively understood, recommendations have been
electric potential measured around the neuron’s extra-
provided for researchers to focus more on the short-
cellular space; and neuronal action potential [42, 43],
and long-term impacts of BCI on the general welfare of
rapid and temporary change in the neuron’s membrane
humans. In addition, our study surfaces several research
potential. Before being presented to the next BCI com-
opportunities in the field of brain–computer interface.
ponent, the captured brain signals undergo filtering,
Researchers and practitioners may capitalize on these
amplification, and digitization [21]. The overall perfor-
opportunities to develop safe BCI products that advance
mance of the BCI system depends heavily on the qual-
humanity and improve quality of our lives.
ity (signal-to-noise ratio) of the acquired brain signals.
Lastly, we extracted 25,336 metadata from Scopus to
Depending on the signal acquisition method, BCI
analyze patterns and trend of BCI research. Results show
can broadly be categorized into two types: invasive
an exponential growth of BCI publications, China being
(electrodes implanted under the scalp to record signals
the leading country between 2019 onwards followed by
directly from the brain) and non-invasive (electrodes
the United States within the same period. This observa-
implanted on the scalp). Invasive BCI provides a more
tion signals the significance of BCI to the community, but
accurate reading of brain signals, but requires surgery;
raises critical questions on the potential BCI threats to
non-invasive BCI does not require surgery, but suffers
humans.
from weak brain signals (poor signal-to-noise ratio)
that require expensive amplification hardware and
2 Fundamental components of BCI system sophisticated signal processing techniques.
The BCI system comprises three fundamental compo-
nents that serve specific roles: signal acquisition, signal
processing, and application (Fig. 1). These components 2.2 Signal processing
are interconnected and work together to allow the flow 2.2.1 Feature extraction
of brain signals to the target BCI application (e.g., robotic In this stage, the BCI system extracts critical elec-
arm). In particular situations, control signals from the trophysiological features from the acquired signals
BCI application may be sent back to the brain to stimu- to define brain activities, and hence encoding of the
late some common human functionalities, such as vision user’s intent [21]. Similar to the previous stage, feature
and hearing.
Maiseli et al. Brain Informatics (2023) 10:20 Page 3 of 16

extraction should be executed accurately, ensuring The BCI technology allows physically challenged people
that the features reflect high correlation with the user’s to control machines using their thoughts. This advantage
intent to enhance the effectiveness and performance gives such people a revealing experience to interact with
of the BCI system. Typical BCI systems employ time- the external environment and accomplish different activi-
domain or frequency-domain features [44–51] that ties without dependence from healthy people.
take different characteristics: amplitude or latency of The BCI field is moving fast with a number of promis-
event-evoked potentials (e.g., P300), frequency power ing outcomes that can significantly improve human lives.
spectra (e.g., sensorimotor rhythms), or neuronal firing Researchers require regular updates to address chal-
rates [21]. Therefore, before designing the BCI system, lenges hindering further advancement of the BCI tech-
the domain transform and characteristics of features nology. More importantly, given the multidisciplinary
should be established. Also, confounding artifacts nature of brain–computer interface, scientists and engi-
contained in the features that can negatively impact neers should work together to develop new and advanced
the subsequent stages of the BCI system should be BCI applications. Recently, the technology has found
eliminated. numerous industrial merits in a range of fields, including
mining and education. Combined with fourth industrial
2.2.2 Feature classification revolution, researchers have demonstrated that BCI may
The extracted features represent brain activities intended accelerate the evolution of robots and neurophysiological
for desired actions. The classification process helps to discoveries [98, 99, 150]. Other applications of the BCI
recognize patterns of the features corresponding to these technology include decoding of thoughts, extension of
actions. For example, we can recognize features repre- human memory, telepathy communication, automation
senting an instruction for moving a robotic arm. This and control, intelligence sharing, brain energy harvest-
component is usually implemented using machine learn- ing, and optimized (targeted) treatment of damaged body
ing and classification methods [52–54]. parts.

2.2.3 Feature translation 3.1 Decoding of thoughts

In this signal processing stage, the classified features are The brain, being a complex human organ, generates and
translated and transformed into actual commands to controls our thoughts and other physiological param-
operate an external device (BCI application). Examples eters: emotion, touch, breathing, hearing, motor skills,
of the outputs given after feature extraction include com- hunger, temperature, memory, and anger. Some param-
mands for cursor movement on the computer screen, eters, such as anger and changes of breathing rate, may
volume control on the audio device, or text writing. One be manifested outside through physical expressions or
important attribute of an algorithm for feature trans- actions. However, most parameters can only be mani-
lation is adaptability [55, 56]: ability of the translation fested internally (inside the brain) without the knowl-
algorithm to adaptively track changes of the features and edge of other people. The current technologies cannot,
generate an appropriate output. for example, predict with an acceptable accuracy the
thoughts of an individual. While this internalization
2.2.4 BCI application of human thoughts—represented as brain signals in
Feature translation generates commands that can con- a BCI system—may have advantages, some situations
trol external devices (BCI applications): cursor [57–60] may demand us to accurately decode such thoughts. In
for letter and text selection on the computer screen [44, criminology, for example, policemen would like to under-
45, 61], wheelchair [62, 63], and robotic arm [64, 65]. For stand whether a suspected criminal speaks the truth.
BCI restoration problems, the control signals from the Recently, researchers have been investigating how BCI
BCI application may be transmitted to the brain or other can improve the performance of polygraphs that measure
body organs. the degree of truth in the arguments from a person (e.g.,
criminal) [2, 66–68]. Perhaps the promising results in this
3 Applications and future of brain–computer direction may be achieved by combining BCI and artifi-
interface cial intelligence techniques.
In this contemporary society, scientists and engineers Can the BCI application facilitate translation of human
have been striving to apply advanced technologies in thoughts accurately into a readable text? How can the
improving quality of human life [144]. Of the available accuracy of the translated text be measured? Can our
technologies, BCI has gained considerable attention in imaginations be mapped into real objects, such as pic-
medicine for its ability to restore emotional and physical tures and texts printed on a piece of paper? Can events
strength of people with missing or damaged body parts. in the dreams be accurately decoded by the BCI system?
Maiseli et al. Brain Informatics (2023) 10:20 Page 4 of 16

Can we extend the applications of BCI to develop wear- conclusion, giving an appreciable impact to a person
able devices that monitor thoughts or sleeping pat- being counseled. Achieving this scientific endeavor
terns [69–71]? Can we extract a will directly from the requires an intensive multidisciplinary research.
thoughts of a dying person? Can we print physical docu-
ments by sending command signals and data from the
brain, through the BCI system, to the printer? These 3.3 Telepathy communication
interesting questions need further scientific inquiry. Rao et al. demonstrated that BCI, in conjunction with
This study envisages that future developments of the computer–brain interface (CBI) [78, 79], may
brain–computer interface will include sophisticated allow individuals to communicate without physical
products that can directly map human thoughts into interaction or sensory channels [80], a process called
physical objects. We believe that, with the growing trend telepathy communication. Integration of BCI and CBI
of BCI, people (especially those with physical disabilities) forms brain–brain interface that is still in early stages
will drive and control machines (e.g., drones, vehicles, of research and development [81–84]. In future, we
and airplanes) remotely using their thoughts [72]. The expect more work in this direction to expand the appli-
advanced developments of BCI may surface critical secu- cations of telepathy communications in various science
rity and privacy issues, and hence the technology needs and engineering fields. As an example, researchers may
to be well-regulated through universal standards [73, 74]. investigate how human brains can be interconnected
over the Internet of Things (IoT) network to enhance
3.2 Extension of human memory exchange of information and experiences among indi-
Stephen Hawking theorized the possibility of upload- viduals. While few studies demonstrate the possibility
ing the human mind into a computer [75]. This philo- of interfacing BCI and IoT [85–90], linking brains and
sophical argument, despite its focus on the human mind IoT over the network remains an open-ended chal-
(consciousness), raises a critical question on whether lenge that deserves attention of researchers. Further-
BCI may be a promising future technology to realize the more, integration of BCI-IoT and other communication
concept. Specifically, how do we extract memory signals modalities, such as mind–mind interface and mind–
from the brain and decode them for storage into a com- machine interface, need further investigation to explore
puter (memory extension)? If successfully implemented, additional capabilities and functionalities on human–
humans will be able to upload their memories into the machine–human communications. All these technolog-
computer for quicker processing, retrieval, and transmis- ical advancements should, however, be made in parallel
sion of information, or for control of external devices. with adherence to ethical principles of humanity.
In the recent developments of brain–computer inter-
face, scientists have generated outstanding results show-
ing that brain signals can be harvested and converted into 3.4 Automation and control
data reflecting human intended actions [76, 77]. Future The promising developments in BCI suggests that
studies on BCI may advance these results to investigate the technology may be useful in automation and con-
how BCI may be used to harvest behaviors and traits trol industries [91–96]. Currently, BCI has received a
from humans for research and scientific study purposes. significant deal of attention in home automation and
But this inquiry should be pursued under strict ethical control [97]. In this scenario, the technology assists
guidelines, a component that has not been well-captured physically challenged people to automate their daily
by the BCI researchers. home activities, making it possible for such people live
The sensitive information from the brain, if accu- independently. As the technology advances, we expect
rately harvested, may be stored into and retrieved from positive impacts of BCI in the industrial manufactur-
the external physical memory. Imagining the future ing processes. In essence, researchers may attempt to
of BCI, we envisage that scientists and practitioners investigate the role of BCI in the fourth industrial revo-
may develop portable flash drives (or other variations lution [98, 99]. For instance, the BCI application may be
of physical memories) that may be plugged into the connected over a secure wireless network to automate
BCI device to extract information from (or introduce processes in the manufacturing industry. Considering
information into) the brain. One may question a pos- sophistication and rapid development in the sensor tech-
sible area that may apply the proposed idea. Imagine nology, BCI may be applied in non-contact control and
a counselling psychologist armed with accurate infor- automation industrial systems. This research direction
mation (obtained through a BCI device) on the behav- requires intensive investigation to overcome inherent
iors and traits of a person. Evidently, this expert may limitations of the BCI technology and ensure seamless
be expected to provide a well-informed advice and interaction with intelligent sensors.
Maiseli et al. Brain Informatics (2023) 10:20 Page 5 of 16

3.5 Intelligence sharing Interface; Brain Computer Interface (BCI); and Brain-
Can the BCI, in conjunction with the CBI, help to repro- Computer Interface. Next, some publications incorrectly
gram the brain, hence allowing sharing of intelligence classified as related to BCI were omitted. In our extended
between individuals? Although it may be imagined as a dataset,2 all the extracted metadata were organized into
fiction, the fundamental principles of the technology continents, regions, and countries3 for analysis. The
suggest that brains may be reprogrammed artificially. VOSviewer4 served a purpose of organizing and analyz-
Achieving this milestone, however, requires solid under- ing the bibliographic networks of the investigated BCI
standing on the nature and functioning of our brains—a publications.
stage that has not been reached by the current state of Our analysis reveals that the BCI field has constantly
knowledge. been evolving over the years, with publications ranging
from theories and fundamental principals to practical
3.6 Brain energy harvesting applications. Studies demonstrate that BCI may signifi-
The human brain takes only 2% of the body’s mass and, cantly improve the quality of life for physically challenged
for an average adult in a normal state, consumes 20% people [77, 102]. Given its broad applications in many
of the whole body energy budget to execute its activi- fields, researchers have invested more time to address
ties [100]. This proportion of energy consumption makes practical challenges in BCI systems. Analyzing previous
it the third most energy-hungry body organ [101]. We BCI studies, we have observed an exponential growth
hypothesize that the BCI technology may be combined of the BCI field to date (Fig. 2a). Within a 5-year inter-
with other advanced technologies to harvest portion val (between 2016 and 2021), for instance, the number of
of this enormous amount of energy for powering low- BCI publications increased steadily by approximately 1.5
energy external devices. Studies are needed to realize the times. This trend suggests an increasing demand of BCI
idea, investigating how much energy can a typical BCI to the scientific and general community, an indicator call-
system harvest from the brain. ing for a need to conduct advanced BCI research.
Figure 2b, c shows that Asia, specifically the Eastern
3.7 Localized brain–computer interface region, has generated more BCI publications over the
In BCI, the process of brain signals acquisition is not dis- years. China demonstrates a steadily growing trend of
criminatory. Virtually, the electrodes acquire all the avail- the publications on brain–computer interface, topping
able signals within the vicinity of its location (under or on other countries from 2019 onwards (Fig. 2d). This inter-
the scalp). Consequently, a huge amount of signals and esting trend may be caused by an increased research
noise are collected for a single intended task (e.g., move- funding and support by the China government to under-
ment of the artificial leg), making the processing of such take advanced research [103, 104]. In the Made in China
signals rather difficult. We can, however, tap the specific 2025 [105] strategy, China has established ambitious
signals intended to control a targeted body part by local- plans to become a leading superpower by 2049. The
izing the BCI system. For example, considering a person strategy, coupled with a higher population size and an
with speech problems, the BCI system may be placed in increased number of academic and research institutions,
an area that directly receives speech control signals from could be a driving factor for China to achieve a remark-
the brain. This advancement may improve the perfor- able achievement in BCI research.
mance of the BCI system and reduce its size. The United States, however, remains a leading coun-
try in terms of the overall number of BCI publications
4 Trend of BCI research (Fig. 3). Given the higher technological and economi-
In analyzing the trend of BCI research, we, on 26 August cal muscle of the United States, this observation would
2022, extracted metadata of 25,336 publications from be expected. Perhaps an intriguing question for future
Scopus.1 The search string used was “brain computer inquiry would be on why the number of BCI publications
interface” that, as per the Scopus research rules, includes for this country started to decline from 2019 onwards.
other similar string variations: brain-machine interface; One way that the United States may improve the trend of
Brain Computer Interface; Brain-Computer Interfaces; BCI publications is to promote co-authorship with Chi-
Brain-computer Interface; Brain Machine Interface; nese universities and research institutions (Fig. 4).
Brain-computer Interface (BCI); Brain Computer Inter-
faces (BCIs); Brain-computer Interfaces; Brain-machine
2
https:// drive. google. com/ drive/u/ 0/ folde rs/ 1vcam Ddm4o NaPtm 5ktWk
TaiT6LMmOEE-h.
3
Link of continents/countries/regions:https://statisticstimes.com/geogr
aphy/countries-by-continents.php.
1 4
https://www.scopus.com. https://www.vosviewer.com/.
Maiseli et al. Brain Informatics (2023) 10:20 Page 6 of 16

NUMBER OF PUBLICATIONS 3000 Asia North America South America Oceania Europe Africa

2500 1400

NUMBER OF PUBLICATIONS
1200
2000
1000
1500 800

600
1000
400
500
200

0 0
1994 1999 2004 2009 2014 2019 2024 1994 1999 2004 2009 2014 2019 2024
PUBLICATION YEAR PUBLICATION YEAR

(a) Overall publication trend (b) Continental publication trend

Northern America Eastern Asia Southern Asia United States China Germany Japan
Northern Europe Western Europe Southern Europe United Kingdom India South Korea
900 600
NU MBER OF PUBLICATIONS

NUMBER OF PUBLICATIONS
800
500
700
600 400
500
300
400
300 200
200
100
100
0 0
1994 1999 2004 2009 2014 2019 2024 1994 1999 2004 2009 2014 2019 2024
PUBLICATION YEAR PUBLICATION YEAR

(c) Regional publication trend (d) Publication trend per top countries
Fig. 2 Evolution of brain–computer interface publications. (Data collected from Scopus on 26 August 2022.)

United States
China
Germany
Japan
United Kingdom
India
South Korea
Italy
Spain
COUNTRY

Canada
France
Austria
Australia
Singapore
Switzerland
Russian Federation
Taiwan
Netherlands
Iran
Poland
Brazil
Turkey
0 500 1000 1500 2000 2500 3000 3500 4000 4500 5000
NUMBER OF PUBLICATIONS
Fig. 3 Number of publications on brain–computer interface per country. (Data collected from Scopus on 26 August 2022.)
Maiseli et al. Brain Informatics (2023) 10:20 Page 7 of 16

viet nam
n viet nam

fiji romani
rom an a peru fiji malta
ma
romania peru
ka
azakhstan kazakhstan
indone
onessia
one indonesia
croa
oatia
a croatia
tia
south korea soutth kore
ea
sri lank
anka
an a i dia
in dia sri la
lanka indi
da
turrkey
ke cypr
prus
turrkey cyprus

taiwan taiwan
wa
ja
apa
pann united states japan un
nite
ed sta
ates
ca
ana
nada
da ca
anada
a
medi
dicine
c
ge
erm
rman
anyy
an medicine
g rm
ger an
ny
slovvakia
a slovakia
den
nma
mar
ak s aiin
sp denmar
markk spain
ieee
ee china ieee china
a
portug
tugal portugal
h g kong
hon kong cuba
uba
ba hong kong cu
cuba
mexic
mexic
ico mexico

bulg
garia
a bulgaria
russian fede
e ration
ed nort
o h ma
acedo
ce nia russian fede
ed ration north macedonia

thaila
land
la thailand

costa
a rica costa rica

(a) Network visualization (b) Density visualization

Fig. 4 Collaboration network among countries based on publications in brain–computer interface

division of biolog
log
ogy and biolog division of biolog
biology
y and biolog
department
me of neu
neur
e ology, unive departmentt of neu
urology, unive
depar
epartment off b
brain
in and cognit depar
pa tment of brain
n and cognit

schoo
ol o
of mecha
ec anical
ca engineeri school of mechanical en
sc engineeri

academy of me
edical
al engin
ngineeriing department of biomed
iome ical engin academy of medical engineeri
ring
ng department of biomed
dical engin
swartz cente
center fo
or c
or comp
putationa depar
artment
tm of biom
iomed
om ical engin swart
artz center for computati
tiona
depa tment off biomed
depar dical engin

university of chines
hines
se academy
dem
em depart
dep
epa ment of biomedical
biom engin universityy o
un of chinese academy
demy department off bio
biomedical engin
school of life
e sci
cience
cience and tec
enc sch
hool of life science and
nd tec
department
me of biom
iomed
edical engin depar
partm
tment o
of biomed
dic
ical engin
institu
ute of biomate
omaterials
rials
lss and depar
artment
ment off e
electr
l
lec ical engin institute of biomaterials
ials and
an dep
partment of electric
e cal engin
college
e of e
elec
ectri
ctrical
al and
an cont department of organis
orga
anis
ni
nismal biolo colleg
llege of elec
ectrical and con
ont departm
ment of organis
nismal biol
olo
key laboratory
y of a
adva
advanced con
n department of biomed
biomed
med
edical
ic engin
in key labora
boratory of advan
anced con
on department
nt of biomediccal en
ngin
ospedale
l san
n ca
amillo, irccs,
ccs
cs v ospedale san cam
millo, irccs, v
depar
at
artment off e
elec
ectr
ctrical
ica engin
eng dep
partm
ment of elect
ectrical en
engin
g.tec medicall engine
eer
ering
rin gmbh
gm department
m nt of neu
me neurosurgery, pa g.tec medica
al engine
ineering gm
gmbh
h departme
depart ment of neurosu
osurgery, pa

departmen
nt of brai
brain
n and
and
nd co
cognit departme
entt of brain and cog
gnit
kore univ
korea versit
ersity,
y, d
departm
tment
en o
ent korea univ
un ersity, de
epartment o
department
entt o
off phys
ph sics
ics, univers
nivers depart
rtment of bioen
oeng
enginee
ne ring, department of physics, univers departmen
depa nt of bioengine
neering,
department of psyc
depart psy hology, univ d
department
nt o
off psychology,
p univ
department of biol
bio ogy and biol department
ent of biology
gy and biol
riken brain sciience
enc institute, riken brain
n science
scien institute,

riken brain scienc

cie e institute,
cien institute of ne
eural engin
eng eerin riken brain science
e institute, institute of neural engi
gineerin

(a) Network visualization (b) Density visualization

Fig. 5 Collaboration network of organizations supporting research on brain–computer interface. (Data collected from Scopus on 26 August 2022.)

Figures 4 and 5 show five countries with higher vol- advanced BCI research with a focus on addressing the
ume of BCI publications: United States, China, Germany, third sustainable development goal, “good health and
Japan, and India. Authors from these countries collabo- well-being”.
rate to foster the development of BCI research. Given Africa lags behind in BCI research (Fig. 2b), generat-
the value of BCI technology in human socio-economic ing only 0.95% of all the BCI publications globally. This
development, we recommend the efforts to be adapted small proportion may be attributed to insufficient fund-
in other countries, specifically those in the global south. ing for supporting and advancing BCI research (Fig. 5).
Institutions from low-income economies, as defined Funding organizations may need to observe Africa as a
by the World Bank, should be empowered to conduct potential continent for BCI research. With an estimated
Maiseli et al. Brain Informatics (2023) 10:20 Page 8 of 16

population of 1.426 billion people by 20225—approxi- 5 Challenges and potential threats of brain–
mately three times that of Europe6—and with more computer interface
than 2,000 universities and institutions,7 Africa can sig- The BCI technology, despite its broad applications, poses
nificantly contribute in BCI research. The methods and threats to humans that need to be addressed. As we strive
results from studies on BCI can improve the quality of life to make the technology friendly and useful, research-
for millions of Africans. According to statistics from the ers should develop BCI applications that resonate with
United Nations, more than 80 million people in Africa the standard principles of humanity. In essence, a bet-
are disabled, including those with severe mental health ter technology should enhance our lives while consider-
conditions and physical impairments that may be benefi- ing human factors, including convenience, ease-of-use,
ciaries from BCI results. Therefore, supported by funding privacy, security, and safety [106–108]. Before adopting
organizations and governments, African researchers and the BCI technology for use by the community, research-
innovators should exploit the capabilities of BCI technol- ers and practitioners are obliged to engage users and
ogy to address the existing practical challenges in Africa. ensure that the technology has passed predefined quality
Another possible reason causing low number of BCI pub- standards.
lications in Africa could be the inadequate level of tech-
nology to undertake BCI research that requires advanced 5.1 Privacy
equipment and complex infrastructure. Collaboration In the article by Luigi Bianchi,11 the author informs
with the developed world, especially China and United lack of specific standards that govern development of
States, in undertaking BCI research may be an effective BCI applications. This challenge, as noted by Takabi
and a feasible strategy for Africa to achieve the desirable et al. [109], has resulted in BCI applications with unre-
output in BCI research. stricted access to brain signals. The authors’ results show
Generally, the BCI research opens up several interest- that these applications may, as a consequence, extract
ing problems that demand attention within the scholarly sensitive information from users without their knowl-
community. Our study discovered that countries address edge. As an attempt to address privacy concerns, stand-
the BCI problem differently depending upon their local ards should be established to define acquisition methods,
contexts. For example, while BCI studies from devel- access control protocols, and encryption techniques,
oped countries focus on the industrial applications of among other attributes. Klein and Ojemann suggest that
the technology, those from developing countries mostly the privacy concerns and other threats may be addressed
deal with how the technology contributes in improving through adherence to best practices when developing
life quality of humans (e.g., increasing life expectancy). BCI systems and incorporating such concerns into the
United States and China, which have shown significant informed consent protocols [110].
advances in BCI research, provide promising prospects In this work, we have hypothesized a functional model
of BCI in the fourth industrial revolution [98, 99] with, of the BCI system that accounts for privacy and secu-
however, a serious concern of the potential threats that rity issues (Fig. 6). This model, which extends the work
the technology may impose if misused. These countries of Mason and Birch [111], contains components that
have, in fact, practically applied BCI in the real-world may prevent unauthorized access of sensitive personal
to advance humanity. Critically analyzing metadata of information without the user’s awareness. Recalling
the 25,336 reviewed articles, we observed sophisticated Fig. 6, before acquisition of brain signals, the BCI sys-
BCI research laboratories8,9,10 that generates results with tem engages the user with predefined access rules to
positive practical impacts. Developing countries, such ensure high integrity and privacy of information. In the
as those in Africa, lack a support infrastructure for BCI signal processing block, a component “Feature selec-
research. Therefore, it may be relatively challenging in tion” retains quality features intended for classification
these countries to comprehensively explore competitive and translation. Next, for BCI applications linked with
advantages of the BCI technology. networked devices over the Internet, we propose encryp-
tion of the translated features (control commands) before
transmission. This process prevents attackers from alter-
ing the control commands, a consequence that may
5
https://worldpopulationreview.com/continents/africa-population. threaten the user’s safety. Other advanced technologies,
6
https://ec.europa.eu/eurostat/web/products-eurostat-news/-/ddn-20220 including blockchain [112], may also be used to prevent
711-1.
7
https://www.webometrics.info/en/Africa?page=20.
8
https://about.bci-lab.info/.
9
https://www.cmu.edu/bme/helab/Research/BCI/index.html. 11
https://lifesciences.ieee.org/lifesciences-newsletter/2019/april-2019/on-
10
https://www.etsu.edu/cas/psychology/bcilab/. brain-computer-interface-standards/.
Maiseli et al. Brain Informatics (2023) 10:20 Page 9 of 16

Signal processing
Feature Feature Feature Feature Feature
Signal Transmitter
extraction selection classification translation encryption
acquisition
Access rules
Internet
BCI Applications

Feedback Wheelchair

Fan
Feature
Receiver
decryption
Bulb

Television

Fig. 6 Brain–computer interface (BCI) system with encryption and decryption components for enhancing privacy

unauthorized access of the control commands by the a comprehensive knowledge on how the human body
attackers. Lastly, the model contains a feature decryption works and interacts with foreign matters. The knowl-
block that decodes the encrypted control commands for edge should be used by BCI scientists and engineers to
use by the BCI applications. develop safe and quality BCI applications. This knowl-
edge should, in addition, equip neurosurgeons with more
5.2 Security accurate information on specific brain regions to implant
The field of BCI has made a significant progress in the BCI electrodes.
development of medical applications and products to
improve the patients’ quality of life (e.g., restoration 5.4 Ethical, legal, and social concerns
of damaged sight or hearing) [113]. However, given the The BCI research raises a number of ethical, legal, and
increasing demand for BCI-internet communications, social concerns on privacy, security, safety, accountabil-
security concerns have emerged [114–116]. The advance- ity, and accessibility [118]. The society would prefer the
ment of brain–computer interface creates opportunities BCI technology that addresses their questions. For exam-
for cyber attackers to intervene in the normal operations ple, should people be concerned by privacy and security
of the BCI application [117]. The attackers may alter of the BCI applications? Does the technology guarantee
commands derived from the feature translation compo- safety? Does the society get equal access to the technol-
nent (Fig. 1) and cause adverse effects to the target sub- ogy? In a situation of negative technological or technical
ject. Therefore, researchers should investigate security impacts, who will be accountable and what are the legal
threats and vulnerable BCI components that can be easily implications? These questions require careful considera-
attacked, then find robust solutions. tions and further research before administering this tech-
nology to the society.
5.3 Safety
Safety concerns can generally be observed in invasive 5.5 Convenience and flexibility
BCI types. Because of being implanted into the brain tis- Most BCI applications require calibration data to reverse
sue, invasive BCI can damage nerve cells and blood ves- undesirable changes caused by neural plasticity or micro-
sels, hence increasing the risk of infection.12 Additionally, movements of the electrode arrays [77]. This necessity
the natural defence system of the body may reject the calls for frequent decoder retraining, an inconvenient
implant, treating it as a foreign entity (biocompatibility and time-consuming process that unnecessarily burdens
concern). Another safety concern of invasive BCI is the the user. Willett et al. [77] highlight the challenge in their
possible formation of scar tissue after surgery, a conse- seminal work on brain-to-text communication through
quence that may gradually degrade the quality of the handwriting. Despite the promising performance
acquired brain signals. Addressing this challenge requires achieved by the authors’ model, daily decoder retrain-
ing was unavoidable. Future studies may investigate more
effective techniques for decoder training without physi-
12
UK Parliament POST, Brain-Computer Interface; POSTNOTE: Number cally engaging the user. In essence, the BCI application
614 January 2020.
Maiseli et al. Brain Informatics (2023) 10:20 Page 10 of 16

Computer Science
Engineering
Medicine

DISCIPLINE
Neuroscience
Mathematics
Biochemistry, Genetics and Molecular Biology
Physics and Astronomy
Materials Science
Chemical Engineering
Decision Sciences
Others
0 2000 4000 6000 8000 10000 12000
NUMBER OF PUBLICATIONS
Fig. 7 Number of brain–computer interface publications per discipline

should operate adaptively with respect to the stochastic invasive BCI requires electrodes optimally positioned
changes in the neural activities of the brain. Automatic under the scalp.
self-calibration approaches may be employed to update
operation of the BCI application accordingly, hence pro- 5.8 Availability of participants for clinical trials
moting convenience and flexibility. BCI, being an emerging and a relatively new technology,
offers promising opportunities to several disadvantaged
5.6 Multidisciplinarity groups. Most people, especially those from developing
The BCI field involves multiple disciplines that should countries, are unaware of the merits and demerits of the
be linked to establish advanced principles and more technology as evidenced from a smaller number of BCI
effective BCI applications. In our analysis from Scopus, publications collected from such countries (Fig. 2b).
we observed that some important disciplines have not Therefore, engaging an acceptable number of people
been adequately engaged in the BCI research (Fig. 7). in testing the BCI medical products may be relatively
For example, only 1% of the BCI-related publications challenging.
originate from psychology, a discipline dealing with study Following ethical guidelines, people should express
of human mind and behavior. Psychology, when com- their consent to accept, adopt and use the BCI technol-
bined with other disciplines, may provide a milestone to ogy. In this work, we noted limited attempts to start
develop even better and practical BCI systems that can clinical trials of BCI devices. On 28 July 2021, Synchron
revolutionize humanity positively. Establishing research became the first BCI company to receive approval from
teams from varied disciplines may require strategic plans the United States Food and Drug Administration for con-
and funding, but such multidisciplinary teams are impor- ducting (investigational device exemption) clinical trial
tant to fully harness the BCI promising capabilities. of a permanently implanted device, Stentrode13 [120].
Other initiatives for clinical trials of BCI products can
5.7 Big data be observed at the University of Pittsburgh14 (sensori-
The brain stores an enormous amount of information motor microelectrode brain–machine interface) and
serving different human tasks. In addition, this central the United States National Library of Medicine15 (e.g.,
body organ generates a vast amount of electrical signals BrainGate216 [121] and BCI device from the University of
that control, monitor, and regulate human activities. Evi- Grenoble [122, 123]). Morinière et al. introduced a dual-
dently, BCI raises a big data problem that needs sophis- arm exoskeleton for evaluating BCI products in clini-
ticated techniques to address. Unfortunately, because of cal trials [124]. Despite these initiatives, including those
insufficient knowledge on the brain working principles, from startups and companies, the number of participants
BCI researchers may not have collected and utilized all involved in the clinical trials seems insufficient for gen-
the brain data and signals. Researchers need to under- eralization across the global community. We recommend
stand key neurological features, including neuroplasticity
that flexibly allows re-organization of neurons in learning
13
or injury recovery [119]. In non-invasive BCI, researchers https://synchron.com/.
14
should determine resolution of the electrode network on https://www.rnel.pitt.edu/.
15
the scalp for optimal collection of brain signals. Similarly, https://clinicaltrials.gov/.
16
https://www.braingate.org/.
Maiseli et al. Brain Informatics (2023) 10:20 Page 11 of 16

BCI Device information on clinical trials for the commercially viable

BCI devices remains unclear. We could locate from pub-
Approval by Standards and lic medical databases only a few clinical trials with lim-
regulatory authority best practices ited number of participants. Considering the delicacy
Rejection and possible long-term impact of BCI technology to
Revision
Success humans, approval procedures from respective regulatory
Certificate
authorities seem necessary before commercialization of
Clinical trials BCI devices (Fig. 8). This necessity, however, introduces
Negative feedback another challenge that some developing countries may
be inadequately equipped with advanced facilities and
Positive feedback

expertise to test and approve BCI devices.

5.10 Battery lifetime

Implantable BCIs require materials that can sustainably
Adoption of
BCI Device
operate over longer periods of time, preferably decades,
without deterioration [119, 128–130]. The warm aque-
Metrics
&
ous nature of our brains, however, affects the power-
Revision Assessment
human factors retention capability of the implants. Water (cerebrospinal
Fig. 8 Proposed procedures for practical administration of brain– fluid), being a powerful solvent, gradually corrodes the
computer interface devices insulating materials of the electrodes. Over time, short
circuits may be created, increasing crosstalks between
electrodes. This challenge reduces battery lifetime and
diversification and increased number of participants for limits the amount of signals collected by electrodes.
clinical trials from different countries, considering cul- Researchers need to study different insulating materials
tural and traditional values. Furthermore, studies may be to understand how they interact with the brain relative
needed to understand acceptance of the BCI technology to the BCIs battery lifetime. In addition, computation-
to the society. In this work, we located a few studies that ally efficient algorithms should be developed to ensure
attempt to determine human behavioral factors towards optimum utilization of battery power. Even more impor-
acceptance of BCI devices [125, 126]. Our recommenda- tantly, alternative energy sources (e.g., micromovements
tion is that, despite the advantages that this technology inside the brain) for powering implantable BCIs should
provides, the development of such devices should con- be investigated.
sider the factors.
5.11 Affordability and portability
5.9 Standardization and approval by regulatory Commercially available BCI devices can hardly be
authorities afforded by the general public because of their prohibi-
We have witnessed an increasing number of initiatives tively high costs [131–134], perhaps due to their sophisti-
to develop BCI devices with advanced features17,18 [119, cation and construction materials. Also, the current BCI
127]. Startups and companies have been developing com- systems are complex and bulkier, making them suitable
mercial BCI devices for use by the society. Our study only in laboratory and industrial settings. Researchers
found ongoing efforts for developing universal standards should develop cost-effective and portable BCI systems
governing neurotechnologies for BCI devices.19 These for ordinary people, potential users of the technology.
efforts should be accelerated to match with the increased This solution will be more useful for people in developing
commercial demands of the BCI devices. Currently, countries.
people may raise concerns on the practical suitability of
the BCI technology with respect to general quality and 6 Conclusion
ethical guidelines. In addition, guided by the best prac- In this study, insights have been given on the perspective
tices for developing and administering medical devices, of the brain–computer interface. Inspired by its benefits,
the society needs to seize the available opportunities that
the technology advocates. To maximize the benefits and
17
https://www.ai-startups.org/top/brain_computer_interface/. increase usability of the BCI technology across the soci-
18
https:// tracxn. com/ explo re/ Brain- Compu ter- Inter face-Start ups- in- ety, researchers and scientists should address the poten-
United-States. tial threats of the technology highlighted in our work.
19
https://standards.ieee.org/industry-conne ctions/neurotechnologies-for- We can fully exploit the benefits and capabilities of the
brain-machine-interfacing/.
Maiseli et al. Brain Informatics (2023) 10:20 Page 12 of 16

technology through multidisciplinary efforts to address dataset21,22 [145–149], we will develop computationally
limitations of the current BCI systems. inexpensive algorithms for encrypting, extracting, clas-
In view of the BCI components, five possible research sifying, and translating features from the brain. Measures
directions can be taken: cognitive psychology, medicine, of accuracy will be established to ensure that the devel-
biomedical electronics, signal processing, and engi- oped algorithms give computer commands that accu-
neering. These directions necessitate multidisciplinary rately emulate users’ actions. Note that there has been no
research where researchers work closely to address the universally acceptable standards for measuring the accu-
BCI sub-challenges. Psychologists and medical doc- racy of BCI applications, and we will attempt to narrow
tors should provide the fundamental working principle this research gap.
of the brain; scientists should develop effective signal
acquisition devices along with algorithms for processing
Author contributions
brain signals (extraction, classification, and translation BM conceived the idea and wrote the initial draft of the paper; ATA reviewed
of features); and engineers should develop physical BCI the technical correctness of the paper; LVM, MM, KM, NAN, MI, JM, and SK
applications and evaluate their performance based on the proofread the manuscript, added missing information, and assisted in data
collection and analysis.
predefined standards.
We assert that the BCI field has many research oppor- Funding
tunities that have not been explored. From all the This research is not supported by any organization.
reviewed literature, an observation was made that the Availability of data and materials
existing challenges in brain–computer interface have The authors declare that the data supporting the findings of this study are
received little attention. The research community is rec- available within the article and its supplementary information files.

ommended to address the challenges and extend the

capabilities that BCI offers, including development of Declarations
BCI-Internet and BCI-CBI communication devices. In Ethics approval and consent to participate
addition, researchers may explore how mind–body inter- Not applicable.
vention methods, such as hypnotherapy, can improve
Competing interests
BCI systems [135–137]. In whatever situation of devel- The authors declare that they have no competing interests.
opment, however, the primary goal of BCI should be to
advance humanity by improving the quality of people’s
Received: 5 January 2023 Accepted: 1 July 2023
lives.
Notwithstanding the promising capabilities and mer-
its of BCI, a significant number of challenges and threats
have not been adequately addressed. In addition, the References
current number of participants in the clinical trials 1. Pfurtscheller G, Neuper C (2009) Brain-computer interface
seems low and undiversified, making generalization of 2. Zander TO, Kothe C (2011) Towards passive brain-computer interfaces:
applying brain-computer interface technology to human-machine
the results questionable. Furthermore, global standards systems in general. J Neural Eng 8:025005
should be established to develop safe and quality BCI 3. Wolpaw JR, Birbaumer N, Heetderks WJ, McFarland DJ, Peckham PH,
products with threats significantly minimized. In this Schalk G, Donchin E, Quatrano LA, Robinson CJ, Vaughan TM et al
(2000) Brain-computer interface technology: a review of the first inter-
regard, although BCI unlocks our future for well-being, national meeting. IEEE Trans Rehabil Eng 8:164–173
this emerging technology requires intensive research, 4. Mudgal SK, Sharma SK, Chaturvedi J, Sharma A (2020) Brain computer
including many clinical trials, for practical applications. interface advancement in neurosciences: applications and issues.
Interdiscip Neurosurg 20:100694
With the existing challenges and threats unsatisfactorily 5. Vidal JJ (1973) Toward direct brain-computer communication. Annu
addressed, the technology may not be ready for con- Rev Biophys Bioeng 2:157–180
sumption by the society. This conclusion is partly sup- 6. Wang Y, Wang R, Gao X, Hong B, Gao S (2006) A practical vep-
based brain-computer interface. IEEE Trans Neural Syst Rehabil Eng
ported by a few other studies [138–144] and scholarly 14:234–240
communications.20 7. Wolpaw JR, McFarland DJ, Neat GW, Forneris CA (1991) An eeg-based
Our future work will be focused on addressing brain-computer interface for cursor control. Electroencephalogr Clin
Neurophysiol 78:252–259
some threats originating from the middle BCI com- 8. Abiri R, Borhani S, Sellers EW, Jiang Y, Zhao X (2019) A comprehensive
ponent, signal processing. Using the publicly available review of eeg-based brain-computer interface paradigms. J Neural Eng
16:011001

21
20
https://www.technologynetworks.com/neuroscience/blog/exploring-the- http://www.bbci.de/competition/iii/.
22
ethical-challenges-of-brain-computer-interface-technology-363367. http://bnci-horizon-2020.eu/database/data-sets.
Maiseli et al. Brain Informatics (2023) 10:20 Page 13 of 16

9. Rashid M, Sulaiman N, Abdul Majeed A, Musa RM, Bari BS, Khatun S et al 31. Ali A, Soomro TA, Memon F, Khan MYA, Kumar P, Keerio MU, Chowdhry
(2020) Current status, challenges, and possible solutions of eeg-based BS (2022) Eeg signals based choice classification for neuromarketing
brain-computer interface: a comprehensive review. Front Neurorobot- applications. A Fusion of Artificial Intelligence and Internet of Things for
ics 14:25 Emerging Cyber Systems 371–394
10. Silversmith DB, Abiri R, Hardy NF, Natraj N, Tu-Chan A, Chang EF, 32. Aldayel M, Ykhlef M, Al-Nafjan A (2021) Consumers’ preference rec-
Ganguly K (2021) Plug-and-play control of a brain-computer interface ognition based on brain-computer interfaces: advances, trends, and
through neural map stabilization. Nat Biotechnol 39:326–335 applications. Arab J Sci Eng 46:8983–8997
11. Aggarwal S, Chugh N (2022) Review of machine learning techniques for 33. Abdulkader SN, Atia A, Mostafa M-SM (2015) Brain computer interfac-
eeg based brain computer interface. Arch Comput Methods Eng 1–20 ing: applications and challenges. Egypt Inf J 16:213–230
12. Pino A, Tovar N, Barria P, Baleta K, Múnera M, Cifuentes CA (2022) 34. Nam CS, Traylor Z, Chen M, Jiang X, Feng W, Chhatbar PY (2021) Direct
Brain–computer interface for controlling lower-limb exoskeletons, in: communication between brains: a systematic Prisma review of brain-
Interfacing Humans and Robots for Gait Assistance and Rehabilitation, to-brain interface. Front Neurorobot 15:656943
Springer, pp. 237–258 35. Asgher U, Khan MJ, Asif Nizami MH, Khalil K, Ahmad R, Ayaz Y, Naseer
13. Saha S, Mamun KA, Ahmed K, Mostafa R, Naik GR, Darvishi S, Khandoker N (2021) Motor training using mental workload (mwl) with an assistive
AH, Baumert M (2021) Progress in brain computer interface: challenges soft exoskeleton system: a functional near-infrared spectroscopy (fnirs)
and opportunities. Front Syst Neurosci 15:578875 study for brain-machine interface (bmi). Front Neurorobotics 15:605751
14. Kinney-Lang E, Kelly D, Floreani ED, Jadavji Z, Rowley D, Zewdie ET, 36. Antonenko P, Paas F, Grabner R, Van Gog T (2010) Using electroenceph-
Anaraki JR, Bahari H, Beckers K, Castelane K et al (2020) Advancing alography to measure cognitive load. Educ Psychol Rev 22:425–438
brain-computer interface applications for severely disabled children 37. Knoll A, Wang Y, Chen F, Xu J, Ruiz N, Epps J, Zarjam P (2011) Measur-
through a multidisciplinary national network: summary of the inaugural ing cognitive workload with low-cost electroencephalograph, in: Ifip
pediatric bci canada meeting. Front Hum Neurosci 14:593883 conference on human-computer interaction, Springer, pp. 568–571
15. Ruiz S, Birbaumer N, Sitaram R (2013) Abnormal neural connectivity in 38. Miller KJ, Shenoy P, Miller JW, Rao RP, Ojemann JG et al (2007) Real-time
schizophrenia and fmri-brain-computer interface as a potential thera- functional brain mapping using electrocorticography. Neuroimage
peutic approach. Front Psych 4:17 37:504–507
16. Hoffmann U, Vesin J-M, Ebrahimi T, Diserens K (2008) An efficient 39. Leuthardt EC, Miller KJ, Schalk G, Rao RP, Ojemann JG (2006) Electrocor-
p300-based brain-computer interface for disabled subjects. J Neurosci ticography-based brain computer interface-the seattle experience. IEEE
Methods 167:115–125 Trans Neural Syst Rehabil Eng 14:194–198
17. Anitha T, Shanthi N, Sathiyasheelan R, Emayavaramban G, Rajendran T 40. Keene D, Whiting S, Ventureyra E (2000) Electrocorticography. Epileptic
(2019) Brain-computer interface for persons with motor disabilities-a Disord 2:57–64
review. Open Biomed Eng J 13 41. Kajikawa Y, Schroeder CE (2011) How local is the local field potential?
18. Moghimi S, Kushki A, Marie Guerguerian A, Chau T (2013) A review of Neuron 72:847–858
eeg-based brain-computer interfaces as access pathways for individu- 42. Smetters D, Majewska A, Yuste R (1999) Detecting action potentials in
als with severe disabilities. Assistive Technol 25:99–110 neuronal populations with calcium imaging. Methods 18:215–221
19. Manyakov NV, Chumerin N, Combaz A, Van Hulle MM (2011) Compari- 43. Khodagholy D, Gelinas JN, Thesen T, Doyle W, Devinsky O, Malliaras
son of classification methods for p300 brain-computer interface on GG, Buzsáki G (2015) Neurogrid: recording action potentials from the
disabled subjects. Comput Intell Neurosci 2011 surface of the brain. Nat Neurosci 18:310–315
20. Soman S, Murthy B (2015) Using brain computer interface for synthe- 44. Farwell LA, Donchin E (1988) Talking off the top of your head: toward
sized speech communication for the physically disabled. Proc Comput a mental prosthesis utilizing event-related brain potentials. Electroen-
Sci 46:292–298 cephalogr Clin Neurophysiol 70:510–523
21. Mak JN, Wolpaw JR (2009) Clinical applications of brain-computer 45. Donchin E, Spencer KM, Wijesinghe R (2000) The mental prosthesis:
interfaces: current state and future prospects. IEEE Rev Biomed Eng assessing the speed of a p300-based brain-computer interface. IEEE
2:187–199 Trans Rehabil Eng 8:174–179
22. Lécuyer A, Lotte F, Reilly RB, Leeb R, Hirose M, Slater M (2008) Brain- 46. Kennedy PR, Bakay RA, Moore MM, Adams K, Goldwaithe J (2000) Direct
computer interfaces, virtual reality, and videogames. Computer control of a computer from the human central nervous system. IEEE
41:66–72 Trans Rehabil Eng 8:198–202
23. Nijholt A, Tan D, Allison B, del R. Milan J, Graimann B (2008) Brain- 47. Krusienski DJ, Sellers EW, McFarland DJ, Vaughan TM, Wolpaw JR (2008)
computer interfaces for hci and games, in: CHI’08 extended abstracts Toward enhanced p300 speller performance. J Neurosci Methods
on Human factors in computing systems, pp. 3925–3928 167:15–21
24. Van Erp J, Lotte F, Tangermann M (2012) Brain-computer interfaces: 48. McFarland DJ, Krusienski DJ, Wolpaw JR (2006) Brain-computer inter-
beyond medical applications. Computer 45:26–34 face signal processing at the wadsworth center: mu and sensorimotor
25. Orenda MP, Garg L, Garg G (2017) Exploring the feasibility to authenti- beta rhythms. Prog Brain Res 159:411–419
cate users of web and cloud services using a brain-computer interface 49. McFarland DJ, Wolpaw JR (2008) Sensorimotor rhythm-based brain-
(bci), in: International conference on image analysis and processing, computer interface (bci): model order selection for autoregressive
Springer, pp. 353–363 spectral analysis. J Neural Eng 5:155
26. Spüler M, Krumpe T, Walter C, Scharinger C, Rosenstiel W, Gerjets P 50. Pardey J, Roberts S, Tarassenko L (1996) A review of parametric model-
(2017) Brain-computer interfaces for educational applications, in: Infor- ling techniques for eeg analysis. Med Eng Phys 18:2–11
mational Environments, Springer, pp. 177–201 51. Schalk G, Wolpaw JR, McFarland DJ, Pfurtscheller G (2000) Eeg-based
27. Katona J, Kovari A (2016) A brain-computer interface project applied in communication: presence of an error potential. Clin Neurophysiol
computer engineering. IEEE Trans Educ 59:319–326 111:2138–2144
28. Verkijika SF, De Wet L (2015) Using a brain-computer interface (bci) in 52. Blankertz B, Dornhege G, Lemm S, Krauledat M, Curio G, Müller K-R
reducing math anxiety: evidence from South Africa. Compute Educ (2006) The berlin brain-computer interface: Machine learning based
81:113–122 detection of user specific brain states. J Univ Comput Sci 12:581–607
29. Mashrur FR, Rahman KM, Miya MTI, Vaidyanathan R, Anwar SF, Sarker F, 53. Lv Z, Qiao L, Wang Q, Piccialli F (2020) Advanced machine-learning
Mamun KA (2022) An intelligent neuromarketing system for predicting methods for brain-computer interfacing, IEEE/ACM Transactions on
consumers’ choice from electroencephalography signals. Physiol Behav Computational Biology and Bioinformatics
113847 54. Elsayed NE, Tolba AS, Rashad MZ, Belal T, Sarhan S (2021) A deep learn-
30. Mashrur FR, Rahman KM, Miya MTI, Vaidyanathan R, Anwar SF, Sarker ing approach for brain computer interaction-motor execution eeg
F, Mamun KA (2022) Bci-based consumers’ choice prediction from eeg signal classification. IEEE Access 9:101513–101529
signals: an intelligent neuromarketing framework. Front Human Neuro- 55. Wolpaw JR, Birbaumer N, McFarland DJ, Pfurtscheller G, Vaughan TM
sci 16 (2002) Brain-computer interfaces for communication and control. Clin
Neurophysiol 113:767–791
Maiseli et al. Brain Informatics (2023) 10:20 Page 14 of 16

56. Wolpaw JR, McFarland DJ (2004) Control of a two-dimensional move- 79. Xie S, Gao C, Yang Z, Wang R (2005) Computer-brain interface, in: Pro-
ment signal by a noninvasive brain-computer interface in humans. Proc ceedings. 2005 First International Conference on Neural Interface and
Natl Acad Sci 101:17849–17854 Control, IEEE, 2005, pp. 32–36
57. Pfurtscheller G, Neuper C, Guger C, Harkam W, Ramoser H, Schlogl A, 80. Rao RP, Stocco A, Bryan M, Sarma D, Youngquist TM, Wu J, Prat CS (2014)
Obermaier B, Pregenzer M (2000) Current trends in graz brain-computer A direct brain-to-brain interface in humans. PLoS ONE 9:e111332
interface (bci) research. IEEE Trans Rehabil Eng 8:216–219 81. Hongladarom S (2015) Brain-brain integration in 2035: metaphysical
58. Kayagil TA, Bai O, Henriquez CS, Lin P, Furlani SJ, Vorbach S, Hallett M and ethical implications. J Inf Commun Ethics Soc
(2009) A binary method for simple and accurate two-dimensional cur- 82. Jiang L, Stocco A, Losey DM, Abernethy JA, Prat CS, Rao RP (2019)
sor control from eeg with minimal subject training. J Neuroeng Rehabil Brainnet: a multi-person brain-to-brain interface for direct collaboration
6:1–16 between brains. Sci Rep 9:1–11
59. McFarland DJ, Krusienski DJ, Sarnacki WA, Wolpaw JR (2008) Emula- 83. Goodman G, Poznanski R, Cacha L, Bercovich D (2015) The two-brains
tion of computer mouse control with a noninvasive brain-computer hypothesis: towards a guide for brain-brain and brain-machine inter-
interface. J Neural Eng 5:101 faces. J Integr Neurosci 14:281–293
60. Mohammadi L, Einalou Z, Hosseinzadeh H, Dadgostar M (2021) Cursor 84. Hildt E (2019) Multi-person brain-to-brain interfaces: ethical issues.
movement detection in brain-computer-interface systems using the Front Neurosci 13:1177
k-means clustering method and lsvm. J Big Data 8:1–15 85. Laport F, Vazquez-Araujo FJ, Castro PM, Dapena A (2018) Brain-com-
61. Rezeika A, Benda M, Stawicki P, Gembler F, Saboor A, Volosyak I (2018) puter interfaces for internet of things. Multidiscip Digital Publ Inst Proc
Brain-computer interface spellers: a review. Brain Sci 8:57 2:1179
62. Pires G, Castelo-Branco M, Nunes U (2008) Visual p300-based bci to 86. Chu NN (2017) Surprising prevalence of electroencephalogram
steer a wheelchair: a bayesian approach, in: 2008 30th Annual Inter- brain-computer interface to internet of things [future directions]. IEEE
national Conference of the IEEE Engineering in Medicine and Biology Consumer Electron Magazine 6:31–39
Society, IEEE, pp. 658–661 87. Teles A, Cagy M, Silva F, Endler M, Bastos V, Teixeira S (2017) Using
63. Galán F, Nuttin M, Lew E, Ferrez PW, Vanacker G, Philips J, Millán JdR brain-computer interface and internet of things to improve healthcare
(2008) A brain-actuated wheelchair: asynchronous and non-invasive for wheelchair users, in: UBICOMM 2017: The Eleventh International
brain-computer interfaces for continuous control of robots. Clin Neuro- Conference on Mobile Ubiquitous Computing, Systems, Services and
physiol 119:2159–2169 Technologies, volume 1, pp. 92–94
64. McFarland DJ, Wolpaw JR (2008) Brain-computer interface operation of 88. Mathe E, Spyrou E (2016) Connecting a consumer brain-computer
robotic and prosthetic devices. Computer 41:52–56 interface to an internet-of-things ecosystem, in: Proceedings of the 9th
65. Flesher SN, Downey JE, Weiss JM, Hughes CL, Herrera AJ, Tyler-Kabara ACM International Conference on PErvasive Technologies Related to
EC, Boninger ML, Collinger JL, Gaunt RA (2021) A brain-computer Assistive Environments, pp. 1–2
interface that evokes tactile sensations improves robotic arm control. 89. Zhang X, Yao L, Zhang S, Kanhere S, Sheng M, Liu Y (2018) Internet of
Science 372:831–836 things meets brain-computer interface: a unified deep learning frame-
66. Haider A, Fazel-Rezai R (2017) Application of p300 event-related poten- work for enabling human-thing cognitive interactivity. IEEE Internet
tial in brain-computer interface, Event-Related Potentials and Evoked. Things J 6:2084–2092
Potentials 1:19–36 90. Coogan CG, He B (2018) Brain-computer interface control in a virtual
67. Wang H, Chang W, Zhang C (2016) Functional brain network and multi- reality environment and applications for the internet of things. IEEE
channel analysis for the p300-based brain computer interface system of Access 6:10840–10849
lying detection. Expert Syst Appl 53:117–128 91. Aricò P, Borghini G, Di Flumeri G, Colosimo A, Bonelli S, Golfetti A, Pozzi
68. Świec J (2021) Brain-computer interface in lie detection, in: Interna- S, Imbert J-P, Granger G, Benhacene R et al (2016) Adaptive automation
tional Scientific Conference on Brain-Computer Interfaces BCI Opole, triggered by eeg-based mental workload index: a passive brain-com-
Springer, pp. 166–175 puter interface application in realistic air traffic control environment.
69. Sathyanarayana A, Srivastava J, Fernandez-Luque L (2017) The science Front Hum Neurosci 10:539
of sweet dreams: predicting sleep efficiency from wearable device data. 92. Yang D, Nguyen T-H, Chung W-Y (2020) A bipolar-channel hybrid
Computer 50:30–38 brain-computer interface system for home automation control utilizing
70. Shelgikar AV, Anderson PF, Stephens MR (2016) Sleep tracking, wearable steady-state visually evoked potential and eye-blink signals. Sensors
technology, and opportunities for research and clinical care. Chest 20:5474
150:732–743 93. Shivappa VKK, Luu B, Solis M, George K (2018) Home automation
71. Martin S, Mikutta C, Knight RT, Pasley BN (2016) Understanding and system using brain computer interface paradigm based on auditory
decoding thoughts in the human brain. Neuroscience selection attention, in: 2018 IEEE international instrumentation and
72. Ascari L, Marchenkova A, Bellotti A, Lai S, Moro L, Koshmak K, Mantoan measurement technology conference (I2MTC), IEEE, pp. 1–6
A, Barsotti M, Brondi R, Avveduto G et al (2021) Validation of a novel 94. Di Flumeri G, De Crescenzio F, Berberian B, Ohneiser O, Kramer J, Aricò
wearable multistream data acquisition and analysis system for ergo- P, Borghini G, Babiloni F, Bagassi S, Piastra S (2019) Brain-computer
nomic studies. Sensors 21:8167 interface-based adaptive automation to prevent out-of-the-loop
73. Sujatha Ravindran A, Aleksi T, Ramos-Murguialday A, Biasiucci A, Fors- phenomenon in air traffic controllers dealing with highly automated
land A, Paek A, et al (2020) Standards Roadmap: Neurotechnologies for systems. Front Hum Neurosci 13:296
Brain-Machine Interfacing, typeTechnical Report, Technical report. IEEE. 95. Aloise F, Schettini F, Aricò P, Leotta F, Salinari S, Mattia D, Babiloni F, Cin-
[Google Scholar] cotti F (2011) P300-based brain-computer interface for environmental
74. Easttom C, Bianchi L, Valeriani D, Nam CS, Hossaini A, Zapała D, Roman- control: an asynchronous approach. J Neural Eng 8:025025
Gonzalez A, Singh AK, Antonietti A, Sahonero-Alvarez G et al (2021) A 96. Corralejo R, Nicolás-Alonso LF, Álvarez D, Hornero R (2014) A p300-
functional model for unifying brain computer interface terminology. based brain-computer interface aimed at operating electronic
IEEE Open J Eng Med Biol 2:65–70 devices at home for severely disabled people. Med Biol Eng Comput
75. Collins N (2013) Hawking in the future brains could be separated from 52:861–872
the body. Telegraph 20:2013 97. Srijony TH, Rashid MKHU, Chakraborty U, Badsha I, Morol MK (2021)
76. Faisal SN, Amjadipour M, Izzo K, Singer JA, Bendavid A, Lin C-T, Iacopi F A proposed home automation system for disable people using bci
(2021) Non-invasive on-skin sensors for brain machine interfaces with system, in: Proceedings of International Joint Conference on Advances
epitaxial graphene. J Neural Eng 18:066035 in Computational Intelligence, Springer, pp. 257–270
77. Willett FR, Avansino DT, Hochberg LR, Henderson JM, Shenoy KV (2021) 98. Xu M, David JM, Kim SH et al (2018) The fourth industrial revolution:
High-performance brain-to-text communication via handwriting. opportunities and challenges. Int J Financial Res 9:90–95
Nature 593:249–254 99. Douibi K, Le Bars S, Lemontey A, Nag L, Balp R, Breda G (2021) Toward
78. Allison BZ, Wolpaw EW, Wolpaw JR (2007) Brain-computer interface eeg-based bci applications for industry 4.0: challenges and possible
systems: progress and prospects. Expert Rev Med Devices 4:463–474 applications, Front Human Neurosci 456
Maiseli et al. Brain Informatics (2023) 10:20 Page 15 of 16

100. Engl E, Attwell D (2015) Non-signalling energy use in the brain. J Physiol interface in clinical trials, in: 2015 IEEE/RSJ International Conference on
593:3417–3429 Intelligent Robots and Systems (IROS), IEEE, pp. 5333–5338
101. Herculano-Houzel S (2011) Scaling of brain metabolism with a fixed 125. Nijboer F (2015) Technology transfer of brain-computer interfaces as
energy budget per neuron: implications for neuronal activity, plasticity assistive technology: barriers and opportunities. Ann Phys Rehabil Med
and evolution. PLoS ONE 6:e17514 58:35–38
102. Capogrosso M, Milekovic T, Borton D, Wagner F, Moraud EM, Mignardot 126. Wang Y-M, Wei C-L, Wang M-W (2022) Factors influencing students’
J-B, Buse N, Gandar J, Barraud Q, Xing D et al (2016) A brain-spine inter- adoption intention of brain–computer interfaces in a game-learning
face alleviating gait deficits after spinal cord injury in primates. Nature context, Library Hi Tech
539:284–288 127. Paszkiel S (2020) Using bci and vr technology in neurogaming, in:
103. Zhou P, Leydesdorff L (2006) The emergence of china as a leading Analysis and Classification of EEG Signals for Brain–Computer Inter-
nation in science. Res Policy 35:83–104 faces, Springer, pp. 93–99
104. Qiu J et al (2014) China goes back to basics on research funding. Nature 128. Sarpeshkar R, Wattanapanitch W, Arfin SK, Rapoport BI, Mandal S, Baker
507:148–149 MW, Fee MS, Musallam S, Andersen RA (2008) Low-power circuits for
105. Zenglein MJ, Holzmann A (2019) Evolving made in china 2025. MERICS brain-machine interfaces. IEEE Trans Biomed Circuits Syst 2:173–183
papers on China 8:78 129. Herron JA, Thompson MC, Brown T, Chizeck HJ, Ojemann JG, Ko AL
106. Marangunić N, Granić A, Technology acceptance model: a literature (2017) Cortical brain-computer interface for closed-loop deep brain
review from (1986) to 2013. Universal access in the information society stimulation. IEEE Trans Neural Syst Rehabil Eng 25:2180–2187
14(2015):81–95 130. Bjorninen T, Muller R, Ledochowitsch P, Sydanheimo L, Ukkonen L,
107. Lee Y, Kozar KA, Larsen KR (2003) The technology acceptance model: Maharbiz MM, Rabaey JM (2012) Design of wireless links to implanted
past, present, and future. Commun Assoc Inf Syst 12:50 brain-machine interface microelectronic systems. IEEE Antennas Wirel
108. Matemba ED, Li G, Gogan ICW, Maiseli BJ (2020) Technology accept- Propag Lett 11:1663–1666
ance model: recent developments, future directions, and proposal for 131. McCrimmon CM, Fu JL, Wang M, Lopes LS, Wang PT, Karimi-Bidhendi A,
hypothetical extensions. Int J Technol Intell Planning 12:315–348 Liu CY, Heydari P, Nenadic Z, Do AH (2017) Performance assessment of a
109. Takabi H, Bhalotiya A, Alohaly M (2016) Brain computer interface (bci) custom, portable, and low-cost brain-computer interface platform. IEEE
applications: Privacy threats and countermeasures, in: 2016 IEEE 2nd Trans Biomed Eng 64:2313–2320
International Conference on Collaboration and Internet Computing 132. Yohanandan SA, Kiral-Kornek I, Tang J, Mshford BS, Asif U, Harrer S (2018)
(CIC), IEEE, pp. 102–111 A robust low-cost eeg motor imagery-based brain-computer interface,
110. Klein E, Ojemann J (2016) Informed consent in implantable bci in: 2018 40th Annual International Conference of the IEEE Engineering
research: identification of research risks and recommendations for in Medicine and Biology Society (EMBC), IEEE, pp. 5089–5092
development of best practices. J Neural Eng 13:043001 133. Rakhmatulin I, Parfenov A, Traylor Z, Nam CS, Lebedev M (2021)
111. Mason SG, Birch GE (2003) A general framework for brain-computer Low-cost brain computer interface for everyday use. Exp Brain Res
interface design. IEEE Trans Neural Syst Rehabil Eng 11:70–85 239:3573–3583
112. Khan AA, Laghari AA, Shaikh AA, Dootio MA, Estrela VV, Lopes RT (2021) 134. Zhang L, Guo X-j, Wu X-p, Zhou B-y (2013) Low-cost circuit design of
A blockchain security module for brain-computer interface (bci) with eeg signal acquisition for the brain-computer interface system, in: 2013
multimedia life cycle framework (mlcf ). Neurosci Inf 100030 6th International Conference on Biomedical Engineering and Informat-
113. Bernal SL, Celdrán AH, Pérez GM, Barros MT, Balasubramaniam S (2021) ics, IEEE, pp. 245–250
Security in brain-computer interfaces: state-of-the-art, opportunities, 135. Alimardani M, Hiraki K (2017) Development of a real-time brain-com-
and future challenges. ACM Comput Surv (CSUR) 54:1–35 puter interface for interactive robot therapy: an exploration of eeg and
114. Denning T, Matsuoka Y, Kohno T (2009) Neurosecurity: security and emg features during hypnosis. Int J Comput Electric Autom Control Inf
privacy for neural devices. Neurosurg Focus 27:E7 Eng 11:187–195
115. Ienca M (2015) Neuroprivacy, neurosecurity and brain-hacking: 136. Rimbert S, Avilov O, Adam P, Bougrain L (2019) Can suggestive hypnosis
Emerging issues in neural engineering, in: Bioethica Forum, volume 8, be used to improve brain-computer interface performance?, in: 8th
Schwabe, pp. 51–53 Graz Brain-Computer Interface Conference 2019
116. Ienca M, Haselager P (2016) Hacking the brain: brain-computer interfac- 137. Deivanayagi S, Manivannan M, Fernandez P (2007) Spectral analysis of
ing technology and the ethics of neurosecurity. Ethics Inf Technol eeg signals during hypnosis. Int J Syst Cybern Inf 4:75–80
18:117–129 138. Cattan G (2021) The use of brain-computer interfaces in games is not
117. Ajrawi S, Rao R, Sarkar M (2021) Cybersecurity in brain-computer ready for the general public. Front Comput Sci 3:628773
interfaces: Rfid-based design-theoretical framework. Inf Med Unlocked 139. Belkacem AN. Real-time human-like robot control based on brain-com-
22:100489 puter interface, in: 2020 2nd International Workshop on Human-Centric
118. Yuste R, Goering S, Bi G, Carmena JM, Carter A, Fins JJ, Friesen P, Gallant Smart Environments for Health and Well-being (IHSH), IEEE, 2021, pp.
J, Huggins JE, Illes J et al (2017) Four ethical priorities for neurotech- xi–xi
nologies and ai. Nature 551:159–163 140. LaGrandeur K (2021) Are we ready for direct brain links to machines
119. Smalley E (2019) The business of brain-computer interfaces. Nat Bio- and each other? A real-world application of posthuman bioethics. J
technol 37:978 Posthumanism 1:87–91
120. Han JJ (2021) Synchron receives fda approval to begin early feasibility 141. Davis KR (2022) Brain-computer interfaces: the technology of our
study of their endovascular, brain-computer interface device future. UC Merced Undergraduate Res J 14
121. Simeral JD, Hosman T, Saab J, Flesher SN, Vilela M, Franco B, Kelemen 142. Arico P, Borghini G, Di Flumeri G, Sciaraffa N, Colosimo A, Babiloni
JN, Brandman DM, Ciancibello JG, Rezaii PG et al (2021) Home use F (2017) Passive bci in operational environments: insights, recent
of a percutaneous wireless intracortical brain-computer interface by advances, and future trends. IEEE Trans Biomed Eng 64:1431–1436
individuals with tetraplegia. IEEE Trans Biomed Eng 68:2313–2325 143. Aricò P, Sciaraffa N, Babiloni F (2020) Brain–computer interfaces: toward
122. Benabid AL, Costecalde T, Eliseyev A, Charvet G, Verney A, Karakas S, a daily life employment
Foerster M, Lambert A, Morinière B, Abroug N et al (2019) An exoskel- 144. Fry A, Chan HW, Harel NY, Spielman LA, Escalon MX, Putrino DF (2022)
eton controlled by an epidural wireless brain-machine interface in a Evaluating the clinical benefit of brain-computer interfaces for control
tetraplegic patient: a proof-of-concept demonstration. Lancet Neurol of a personal computer. J Neural Eng 19:021001
18:1112–1122 145. Daly I, Matran-Fernandez A, Valeriani D, Lebedev M, Kübler A (2021)
123. Larzabal C, Bonnet S, Costecalde T, Auboiroux V, Charvet G, Chabardes Datasets for brain-computer interface applications. Front Media SA
S, Aksenova T, Sauter-Starace F (2021) Long-term stability of the chronic 146. Kaya M, Binli MK, Ozbay E, Yanar H, Mishchenko Y (2018) A large electro-
epidural wireless recorder wimagine in tetraplegic patients. J Neural encephalographic motor imagery dataset for electroencephalographic
Eng 18:056026 brain computer interfaces. Sci Data 5:1–16
124. Moriniere B, Verney A, Abroug N, Garrec P, Perrot Y (2015) Emy: a 147. Cho H, Ahn M, Ahn S, Kwon M, Jun SC (2017) Eeg datasets for motor
dual arm exoskeleton dedicated to the evaluation of brain machine imagery brain–computer interface, GigaScience 6 gix034
Maiseli et al. Brain Informatics (2023) 10:20 Page 16 of 16

148. Wang Y, Chen X, Gao X, Gao S (2016) A benchmark dataset for ssvep-
based brain-computer interfaces. IEEE Trans Neural Syst Rehabil Eng
25:1746–1752
149. Zhu F, Jiang L, Dong G, Gao X, Wang Y (2021) An open dataset for wear-
able ssvep-based brain-computer interfaces. Sensors 21:1256
150. Zhang X, Ma Z, Zheng H, Li T, Chen K, Wang X, Liu C, Xu L, Wu X, Lin
D, Lin H (2020) The combination of brain-computer interfaces and
artificial intelligence: applications and challenges. Ann Transl Med 8:712

Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in pub-
lished maps and institutional affiliations.