REVIEW

published: 25 February 2021

doi: 10.3389/fnsys.2021.578875

Progress in Brain Computer

Interface: Challenges and
Opportunities
Simanto Saha 1,2*, Khondaker A. Mamun 3 , Khawza Ahmed 2 , Raqibul Mostafa 2 ,
Ganesh R. Naik 4 , Sam Darvishi 1 , Ahsan H. Khandoker 5 and Mathias Baumert 1*
1
School of Electrical and Electronic Engineering, The University of Adelaide, Adelaide, SA, Australia, 2 Department of
Electrical and Electronic Engineering, United International University, Dhaka, Bangladesh, 3 Advanced Intelligent
Multidisciplinary Systems (AIMS) Lab, Department of Computer Science and Engineering, United International University,
Dhaka, Bangladesh, 4 Adelaide Institute for Sleep Health, College of Medicine and Public Health, Flinders University, Adelaide,
SA, Australia, 5 Healthcare Engineering Innovation Center, Department of Biomedical Engineering, Khalifa University of
Science and Technology, Abu Dhabi, United Arab Emirates

Brain computer interfaces (BCI) provide a direct communication link between the brain
and a computer or other external devices. They offer an extended degree of freedom
either by strengthening or by substituting human peripheral working capacity and
have potential applications in various fields such as rehabilitation, affective computing,
robotics, gaming, and neuroscience. Significant research efforts on a global scale
have delivered common platforms for technology standardization and help tackle
highly complex and non-linear brain dynamics and related feature extraction and
classification challenges. Time-variant psycho-neurophysiological fluctuations and their
impact on brain signals impose another challenge for BCI researchers to transform
the technology from laboratory experiments to plug-and-play daily life. This review
Edited by:
Alessandro E. P. Villa,
summarizes state-of-the-art progress in the BCI field over the last decades and highlights
Neuro-Heuristic Research Group critical challenges.
(NHRG), Switzerland
Keywords: brain computer interface, hybrid/multimodal BCI, neuroimaging techniques, neurosensors,
Reviewed by:
electrical/hemodynamic brain signals, cognitive rehabilitation
Dongrui Wu,
Huazhong University of Science and
Technology, China
David R. Painter, 1. INTRODUCTION
The University of Queensland,
Australia The brain computer interface (BCI) is a direct and sometimes bidirectional communication tie-up
*Correspondence:
between the brain and a computer or an external device, which involves no muscular stimulation.
Simanto Saha It has shown promise for rehabilitating subjects with motor impairments as well as for augmenting
[email protected] human working capacity either physically or cognitively (Lebedev and Nicolelis, 2017; Saha and
Mathias Baumert Baumert, 2020). BCI was historically envisioned as a potential technology for augmenting/replacing
[email protected] existing neural rehabilitations or serving assistive devices controlled directly by the brain (Vidal,
1973; Birbaumer et al., 1999; Alcaide-Aguirre et al., 2017; Shahriari et al., 2019). The first systematic
Received: 01 July 2020 attempt to implement an electroencephalogram (EEG)-based BCI was made by J. J. Vidal in 1973,
Accepted: 06 January 2021 who recorded the evoked electrical activity of the cerebral cortex from the intact skull using EEG
Published: 25 February 2021
(Vidal, 1973), a non-invasive technique first studied in humans invented by Berger (1929). Another
Citation: early endeavor to establish direct communication between a computer and the brain of people with
Saha S, Mamun KA, Ahmed K,
severe motor impairments had utilized P300, an event related brain potential (Farwell and Donchin,
Mostafa R, Naik GR, Darvishi S,
Khandoker AH and Baumert M (2021)
1988). As an alternative to conventional therapeutic rehabilitation for motor impairments, BCI
Progress in Brain Computer Interface: technology helps to artificially augment or re-excite synaptic plasticity in affected neural circuits.
Challenges and Opportunities. By exploiting undamaged cognitive and emotional functions, BCI aims at re-establishing the
Front. Syst. Neurosci. 15:578875. link between the brain and an impaired peripheral site (Vansteensel et al., 2016). However, the
doi: 10.3389/fnsys.2021.578875 research applications of BCI technology evolved significantly over the years, including brain

Frontiers in Systems Neuroscience | www.frontiersin.org 1 February 2021 | Volume 15 | Article 578875

Saha et al. Progress in Brain Computer Interface

fingerprinting for lie detection (Farwell et al., 2014), detecting summarizes signal acquisition modalities and their suitability for
drowsiness for improving human working performances (Aricò BCI applications.
et al., 2016; Wei et al., 2018), estimating reaction time (Wu et al., Recent technological advancements allow both the decoding
2017b), controlling virtual reality (Vourvopoulos et al., 2019), of neural activities and the delivery of external signals into
quadcopters (LaFleur et al., 2013) and video games (Singh et al., targeted brain areas to induce plasticity, i.e., remodeling of
2020), and driving humanoid robots (Choi and Jo, 2013; Spataro neurosynaptic organization (Lajoie et al., 2017). Plasticity
et al., 2017). Figure 1 demonstrates the progression of BCI in is an inherent characteristic of the brain and peripheral
various application fields since its conception. nervous system underpinning BCI-based rehabilitation and
According to the Brain/Neural Computer Interaction Horizon other neuroscientific applications. While most of the BCI systems
2020 project, an initiative by the European Commission for translate brain signals to computer commands, some systems
coordinating BCI research, six major application themes, i.e., utilize external stimulation modalities such as transcranial
restore (e.g., unlocking the completely locked-in), replace (e.g., magnetic stimulation (Grau et al., 2014; Rao et al., 2014;
BCI-controlled neuroprosthesis), enhance (e.g., enhanced user Schaworonkow et al., 2019) and transcranial direct current
experience in computer games), supplement (e.g., augmented stimulation (Baxter et al., 2017) to stimulate specific brain
reality glasses), improve (e.g., upper limb rehabilitation after areas. The bidirectional framework of BCI comprises either one
stroke), and research tool (e.g., decoding brain activity with real- brain with feedback modality or two brains. Transcranial direct
time feedback) have been outlined as feasible and promising fields current stimulation directed by MI-related EEG signals alters
(Brunner et al., 2015). This overview encompasses a wide range the connectivity in sensorimotor networks of healthy individuals
of challenges and trends in BCI field. For specialized reviews on (Baxter et al., 2017). Another possible application of bidirectional
particular BCI topics we refer to the recent literature (McFarland BCI framework is direct brain-to-brain communication (Grau
et al., 2006; Schwartz et al., 2006; Bashashati et al., 2007; Lotte et al., 2014; Rao et al., 2014). Moreover, some BCI applications
et al., 2007, 2018; Matthews et al., 2007; Sitaram et al., 2007; Mak require auxiliary modalities, e.g., proprioceptive feedback and
and Wolpaw, 2009; Menon et al., 2009; Nicolelis and Lebedev, functional electrical stimulation driven by brain signals as
2009; Summerer et al., 2009; Vaadia and Birbaumer, 2009; Milan feedback for augmenting or regaining peripheral motor actions
and Carmena, 2010; Min et al., 2010; Clausen, 2011; Krusienski (Darvishi et al., 2017; Bhattacharyya et al., 2019; Bockbrader et al.,
et al., 2011; Liao et al., 2012; Nicolas-Alonso and Gomez-Gil, 2019; Murovec et al., 2020).
2012; Shih et al., 2012; Jebari, 2013; McCullagh et al., 2014; Ahn
and Jun, 2015; Jayaram et al., 2016; Lebedev and Nicolelis, 2017;
Mudgal et al., 2020; Rashid et al., 2020; Saha and Baumert, 2020). 1.2. Factors Influencing BCI Performance
For medical applications of BCI, three criteria are essential: (1) a
comfortable and convenient signal acquisition device, (2) system
1.1. Characterization of BCI Systems validation and dissemination, and (3) reliability and potentiality
BCI systems can be categorized by the way they use the brain: of BCI (Shih et al., 2012). For the restoration of mobility
Passive BCI decode unintentional affective/cognitive states of the in patients with motor impairments, invasive intracortical
brain (Zander et al., 2009), while active BCI directly involve recordings show better BCI performance (Hochberg, 2013)
the user’s voluntary intention-induced brain activity. Reactive than non-invasive methods such as EEG (Milan and Carmena,
BCI use brain waves generated as response to external stimuli. 2010). The performance determines how efficiently a patient
Detecting driver’s drowsiness to prevent road accidents is an can perform an impaired motor task or communicate with an
example of passive BCI (Lin et al., 2008; Gao et al., 2019). external device. Invasive modalities are also suitable for locked-in
BCI systems driven by users’ intentional motor imagery (MI) patients, because the benefits (significantly improved quality-of-
(Marchesotti et al., 2016; Saha et al., 2019a; Saha and Baumert, life) outweigh the risks associated with implantation (Gilja et al.,
2020) and visually evoked P300 produced by external stimulation 2011). A pilot study found no adverse effects pertaining to surgery
(Farwell et al., 2014) can be considered active BCI and reactive or tissue reaction at 1 year follow-up (Friehs et al., 2006). Invasive
BCI, respectively. BCI should generally not be considered for neurologically intact
The modality of signal acquisition has been used to people due to risks associated with surgery. However, invasive
divide systems into invasive and non-invasive BCI (Min recordings may enable the utilization of localized inner cortex
et al., 2010; Rosenfeld and Wong, 2017). Non-invasive BCI activities and a better interpretation of surface recordings from
exploiting EEG are most common, although more recently, a non-invasive modality (Schalk, 2010; Lina et al., 2012).
functional near infrared spectroscopy (fNIRS) (Matthews et al., Many factors influence BCI performance; taking the
2007), magnetoencephalography (MEG) (Fukuma et al., 2016), underlying cortical-subcortical networks into consideration is
functional magnetic resonance imaging (fMRI) (Kaas et al., 2019) of crucial importance. For example, MI-induced signals are best
and functional transcranial Doppler ultrasonography (Faress recorded from premotor and motor areas, because premotor
and Chau, 2013; Lu et al., 2015; Khalaf et al., 2019) have cortex, primary motor cortex and supplementary motor area
been exploited. In contrast, invasive intracortical electrodes along with basal ganglia and thalamus of the subcortical areas are
(Pandarinath et al., 2017) and electrocorticography (ECoG) the mostly activated areas during MI (Marchesotti et al., 2017).
(Kaiju et al., 2017) have been used, providing a superior signal- While EEG can capture premotor and motor area activation
to-noise ratio and better localization of brain activity. Table 1 (Edelman et al., 2015; Saha et al., 2019a), intracortical electrodes

Frontiers in Systems Neuroscience | www.frontiersin.org 2 February 2021 | Volume 15 | Article 578875

Saha et al. Progress in Brain Computer Interface

FIGURE 1 | The number of publications over the years: The statistics was based on a search on PubMed in which “brain computer interface” was the search
keyword. The publications those were listed until 4th December 2020 have been accounted only. A significant increase in the number of publications in this decade as
compared to the last decade implicates the engagement of a greater community in this field and, thus the importance of BCI technology.

can record signals from basal ganglia and thalamus (Sand et al., associativity, demonstrated in previous works in case of
2017). natural vision (Hasson et al., 2004) and natural music listening
Several issues can significantly impede BCI performance. (Abrams et al., 2013), could be exploited toward inter-subject
Maintaining an acceptable signal-to-noise ratio in non-invasive operable BCI. Recent studies suggest that inter-subject operable
long-term recordings is critical. Event-induced brain waves sensorimotor rhythm-based BCI might become feasible for
or oscillations are dynamic and affected by unstable resting subjects who share common brain dynamics (Saha et al.,
state networks (RSNs) (Mantini et al., 2007). Time-variant 2017, 2018, 2019a; Saha and Baumert, 2020). Inter-subject
psychophysiological (Gonçalves et al., 2006; Zhang et al., BCI holds promise predominantly for healthy people and in
2015; Acqualagna et al., 2016; Saha and Baumert, 2020), applications such as gaming, drowsiness and lie detection,
neuroanatomical (Kasahara et al., 2015) factors and users’ because rehabilitative BCI must consider the characteristics and
fundamental traits (Ahn and Jun, 2015) cause unreliable severity of individual impairment (Park et al., 2016). Transfer
estimates of RSNs, causing short and long-term signal variation learning can also reduce the effects of session-to-session and
within and across individuals (Saha and Baumert, 2020). subject-to-subject variabilities, by using systems that were
Due to these intrinsic signal variations, BCI systems require trained on data from different people exploiting commonalties
subject-specific training, during which subjects attend a and reducing training requirements (Jayaram et al., 2016;
calibration session that is tedious and often frustrating. To Saha et al., 2017, 2018, 2019a; He and Wu, 2019; Wu et al.,
eliminate subject-specific training, the concept of inter-subject 2020).

Frontiers in Systems Neuroscience | www.frontiersin.org 3 February 2021 | Volume 15 | Article 578875

Saha et al. Progress in Brain Computer Interface

TABLE 1 | A list of neuroimaging techniques and their suitability in brain computer interface (BCI) applications.

Feature EEG MEG ECoG Intracortical fMRI fNIRS PET

Recording

Activity type Electrical Magnetic Electrical Electrical Metabolic Metabolic Metabolic

Measurement type Direct Direct Direct Direct Indirect Indirect Indirect
Invasiveness Non-invasive Non-invasive Invasive Invasive Non-invasive Non-invasive Invasive
Portability Yes No Yes Yes No Yes No
Temporal resolution ∼ 0.05 s ∼ 0.05 s ∼ 0.003 s ∼ 0.003 s ∼1s ∼1s 1–2 min
Spatial resolution ∼ 10 mm ∼ 5 mm ∼ 1 mm ∼ 0.5 mm (LFP) ∼ 1 mm ∼ 5 mm ∼ 4 mm
∼ 0.1 mm (MUA)
∼ 0.05 mm (SUA)
BCI applicability Acceptable Mobility Unfavorable Unfavorable Slow and Slow, but Limited
spatio-temporal constraint for healthy for healthy mobility mobile and potentiality
resolution with BCI users BCI users constraint a potential
high-density alternative
electrodes to fMRI

EEG, electroencephalography; MEG, magnetoencephalography; ECoG, electrocorticography; fMRI, functional magnetic resonance imaging; fNIRS, functional near infrared
spectroscopy; PET, positron emission tomography.

2. CHALLENGES volume in sensorimotor cortical areas is associated with BCI

success (Kasahara et al., 2015). Sensorimotor rhythm-based BCI
2.1. Psychophysiological and Neurological has implicated that physiological predictors such as spectral
Challenges entropy and power spectral density, derived from resting state
Emotional and mental processes, neurophysiology associated EEG recordings are correlated with BCI performance (Zhang
with cognition and neurological factors, i.e., functions, anatomy, et al., 2015; Acqualagna et al., 2016). Psychological predictors
play crucial roles in BCI performance and give rise to significant such as attention and motivation, are also associated with
intra- and inter-individual variability (Saha and Baumert, 2020). sensorimotor rhythm-based BCI performance (Hammer et al.,
Psychological factors such as attention, memory load, fatigue and 2012). Corticospinal excitability could be used as another reliable
competing cognitive processes (Gonçalves et al., 2006; Käthner marker for BCI performance (Vasilyev et al., 2017). Taking
et al., 2014; Calhoun and Adali, 2016) as well as users’ basic head anatomy into consideration augments BCI performance
characteristics such as lifestyle, gender, and age, (Kasahara et al., (Wronkiewicz et al., 2015; Saha et al., 2019a).
2015) influence instantaneous brain dynamics. For example, Around 15–30% of individuals are inherently not able
individuals with lower empathy participate less emotionally in to produce brain signals robust enough to operate a BCI
a P300-BCI paradigm and can produce higher amplitudes of (Blankertz et al., 2009; Halder et al., 2019; Cecotti, 2020).
P300 waves than subjects with greater empathetic involvement Considering neurophysiological phenomena may reduce BCI
(Kleih and Kübler, 2013). Motivation is also related to P300-BCI illiteracy. An adaptive machine learning approach incorporating
performance (Nijboer et al., 2010). neurophysiological and psychological traits has been proposed
Besides psychological traits, resting state physiological to reduce BCI illiteracy (Vidaurre and Blankertz, 2010). The
parameters, for example, frequency domain features of resting causes of BCI illiteracy do not exclusively rely on users’ ability to
state heart rate variability are associated with BCI performance produce signals. Sometimes technological limitations may hinder
(Kaufmann et al., 2012). In addition, the baselines of RSNs are essential features extraction for a successful BCI operation for
dynamic and modify any cortical signature instantaneously an individual. For example, measurements of scalp EEG/MEG
(Mantini et al., 2007). Age alters RSNs and associated cognitive may not show good task-specific signals due to the folding of the
responses (Wang et al., 2016b). Adapting to such time-variant cortex or scalp-to-cortex distance for that individual (Andersen
RSNs is more demanding when the effects of RSNs mask et al., 2020).
event-related cortical responses (Jensen et al., 2011). Moreover, Other case-specific investigations on neuro-psycho-
the inherent complexity and diversity in the formation of physiological parameters contributing to BCI performance
human brains (Sporns, 2013) that influence the functional are essential. For the rehabilitation of stroke survivors, affected
neural networks (Honey et al., 2010), construct highly volatile neural circuits, i.e., lesions are to be identified carefully, because
neuronal connectivity over time and across subjects (Honey brain responses fluctuate according to the spatial location of the
et al., 2009). An efficient BCI system must be robust to such stroke lesion (Park et al., 2016). Although current neuroimaging
inherent physiological fluctuations over time to enable more methods are effective in capturing stroke lesion sites, a case-
generalized systems (Saha and Baumert, 2020). specific BCI design that incorporates residual brain function is
Experiments correlating BCI performance with required for rehabilitative interventions. Highly individualized
neuroanatomical, neurophysiological and psychological design impedes wide dissemination of BCI-driven rehabilitation
parameters have provided fascinating results: gray matter of neurological conditions.

Frontiers in Systems Neuroscience | www.frontiersin.org 4 February 2021 | Volume 15 | Article 578875

Saha et al. Progress in Brain Computer Interface

2.2. Technological Challenges resolution (Lystad and Pollard, 2009; Min et al., 2010; He
Event related potential (ERP) (McCane et al., 2015), steady- et al., 2011; Nicolas-Alonso and Gomez-Gil, 2012). Employing
state visual evoked potential (SSVEP) (Chen et al., 2015; Abu- high density EEG mapping increases spatial resolution but
Alqumsan and Peer, 2016), auditory evoked potential (AEP) results in high computational cost and efforts to maintain
(Schreuder et al., 2010), steady-state somatosensory evoked a reasonable signal-to-noise ratio across all channels (Chen
potential (SSSEP) (Muller-Putz et al., 2006; Oxley et al., 2017), et al., 2015). Since EEG captures only the electrical field
and motor imagery (MI) (Marchesotti et al., 2016; Saha et al., associated cognitive processes, concomitant assessment of blood-
2019a; Saha and Baumert, 2020), have been proposed to detect oxygen level-dependent (BOLD) activity may improve BCI
cognitive signatures although none of the approaches performs performance. BOLD activity is typically captured with fMRI
well for all BCI applications. For example, ERPs and SSVEPs (Sitaram et al., 2007), which is not feasible for most BCI
are target-specific and elicited by external stimuli; however, applications, due to unmanageable size and cost of the device.
if ERPs depend on visual stimuli, they cannot be used for fNIRS provides a safe, non-invasive, relatively inexpensive
communication by locked-in patients with impaired visual and portable neuroimaging alternative for recording BOLD
processing. In that case, auditory-based ERP (e.g., AEP) could be activity (Matthews et al., 2007). Integrating fNIRS with EEG
used if auditory processing remains intact. The SSVEP method can significantly enhance classification performances regardless
provides the highest information transfer rate of a non-invasive of low information transfer rate caused by inherent delays
EEG-based BCI (Chen et al., 2015; Abu-Alqumsan and Peer, in hemodynamics (Fazli et al., 2012). A recent study has
2016). Limitations of the SSVEP technique include visual fatigue suggested that fNIRS is unable to adequately offer acceptable
caused by looking at a flickering display for a long time. performances on its own, but can be combined with EEG to
When using this technique, the control signal could be arbitrary boost the performances (Ge et al., 2017). However, continuous
and counter-intuitive, although it might depend mostly on the technological advances could promote fNIRS as an exclusive tool
experimental context. For example, when using a BCI speller for neuroscience research, including the development of BCI
based on SSVEP, an individual looks at the letter “A”, which (Scholkmann et al., 2014; Naseer and Hong, 2015).
flickers at 10 Hz. It is generally not given importance to any Probing sources in cortico-subcortical networks is another
inherent relationship between “A” and 10 Hz, instead the control important limitation of scalp-based sensors such as EEG.
signal is arbitrary mapped to interface with a computer. An Reconstructing task-induced networks while resolving the so-
advantage of an MI-based BCI is the use of explicit mapping called inverse problem imposes a significant challenge. A two-
of task-related brain signals to operate (Saha and Baumert, equivalent-dipole model was applied on EEG data to discern
2020). However, MI seems too slow for action control, thus the anatomical nature of the MI induced sources and to aid
they are not suitable for controlling virtual reality environments the classification performances (Kamousi et al., 2005). Saha
or videogames (Lécuyer et al., 2008). Recently proposed hybrid et al. proposed a wavelet-based source localization approach
BCIs which utilize more than one signature, i.e., SSVEP/ERP to investigate MI-related sources and their impact on BCI
(Combaz and Van Hulle, 2015; Yin et al., 2015) and SSVEP/MI performance (Saha et al., 2019a). The neuronal potentials
(Pfurtscheller et al., 2010; Horki et al., 2011), seem to offer more attenuate through several tissue layers of complex geometry and
robust features. Considering asynchronous BCI where the user diverse electrical properties; however, the magnetic permeability
decides to activate a command when necessary, the performance in the cerebrospinal fluid, skull, and skin, is consistent (da Silva,
is still unsatisfactory (Han et al., 2020). 2013). Thus, MEG can capture signal with less distortion than
The intrinsic neurophysiological instability of brain dynamics EEG. Although MEG provides better spatiotemporal resolution
poses critical challenges for making BCI systems efficient. The as compared to EEG, the magnetic field created by the brain
major components of a BCI system are signal acquisition, signal is very small, requiring costly, stationary recording equipment
processing and effector device (Schwartz et al., 2006). Various (Mellinger et al., 2007; Corsi et al., 2019).
neuroimaging techniques have been used to explore cortical The BCI classifier design has to address two issues (Bashashati
activities through either electrical or hemodynamic signatures et al., 2007; Lotte et al., 2007, 2018). First, the dimensionality
(Min et al., 2010), but none of the methods shows any advantage of the features set used for estimating the model parameters
for a lucrative BCI design meeting the four important criteria: should be chosen for optimal performance based on the
cost efficiency, portability, easy maintenance, and little or no nature of the classifier. Second, the trade-off between bias and
involvement of surgery. EEG-based BCI are relatively more variance has to be considered and may involve regularizing the
compliant with the abovementioned criteria as compared to parameter estimation.
other signal acquisition modalities. Tables 2, 3 list a diverse range Covariate shift occurs when the features extracted from
of BCI applications exploiting EEG. Both invasive and non- the training differ from those of test data impacting the
invasive signal acquisitions have recently shown that reliable classification performance (Krusienski et al., 2011). Covariate
long-term (i.e., for at least several months) use of BCI systems shift is an important issue requiring the application of adaptive
is quite feasible (Saeedi et al., 2016; Sauter-Starce et al., 2019; methods for compensating feature space transitions (Jayaram
Shahriari et al., 2019; Oxley et al., 2020). et al., 2016; Saha and Baumert, 2020). The unsupervised
EEG provides relatively poor spatial resolution due to non- subspace learning method enables session-to-session and subject-
invasive scalp recordings compared to fMRI, but finer temporal to-subject information transfers, augmenting BCI performance

Frontiers in Systems Neuroscience | www.frontiersin.org 5 February 2021 | Volume 15 | Article 578875

Saha et al. Progress in Brain Computer Interface

TABLE 2 | A summary of sensorimotor rhythms and electroencephalography-based brain computer interface (BCI) studies.

References Modality Wave/Task Data analysis Context Application

Romero-Laiseca et al., 2020 EEG MI Riemannian geometry, Motor learning & Lower limb
LDA plasticity induction rehabilitation
Song and Kim, 2019 EEG MI Spectral averaging Visuo-Tactile MI enhancement
stimulation
Saha et al., 2019a EEG MI wMEM Transfer learning Inter-subject BCI
Toriyama et al., 2018 EEG MI, ME LDA MI & ME resemblance Corticospinal excitability
Yu et al., 2018 EEG MI CSP, LDA Assistive technology Wheelchair navigation
Darvishi et al., 2017 EEG MI LF Motor training by Upper limb rehabilitation
proprioceptive feedback
Park et al., 2016 EEG MI, ME ICA, CAR Variable lesion Stroke rehabilitation
characteristics
Kasahara et al., 2015 EEG MI LF, AR Neuroanatomical predictor Motor rehabilitation
Edelman et al., 2015 EEG MI WMNE Decoding complex MI Higher degrees
of freedom
Leamy et al., 2014 EEG ME, MI Filter bank CSP Plasticity induction Stroke rehabilitation
Ramos-Murguialday et al., 2013 EEG ME, MI AR Stroke rehabilitation, Hand orthosis control
assitive technology
LaFleur et al., 2013 EEG MI AR Assistive technology, Quadcopter control
robotics

EEG, electroencephalography; ME, motor execution; MI, motor imagery; wMEM, wavelet-based maximum entropy on the mean; LF, Laplacian filter; AR, autoregressive model; CSP,
common spatial pattern; ICA, independent component analysis; CAR, common average reference; WMNE, weighted minimum norm estimate; LDA, linear discriminant analysis.

TABLE 3 | A summary of non-sensorimotor rhythms and electroencephalography-based brain computer interface (BCI) studies.

References Modality Wave\Task Data analysis Context Application

Jin et al., 2020 EEG P300 LDA Vibrotactile stimuli Spinal cord injury
rehabilitation
Cecotti, 2020 EEG SSVEP CCA Generalized framework BCI literacy
Halder et al., 2019 EEG P300 LDA Auditory stimuli BCI literacy
Wei et al., 2018 EEG Continuous PCA, LDA, Driving scenario in Drowsiness detection
(alert vs. SVM virtual reality
drowsy)
Guy et al., 2018 EEG P300 LDA Amyotrophic lateral Visual P300 speller
sclerosis
Alcaide-Aguirre et al., 2017 EEG P300 LDA Cerebral palsy Cognitive assessment
Waytowich et al., 2016 EEG P300 Information Transfer learning Inter-subject BCI
geometry
Norton et al., 2015 EEG SSVEP, Averaging Soft, curved Text speller
P300 electrode systems
Botrel et al., 2015 EEG P300 LDA Amyotrophic lateral Brain painting
sclerosis
Farwell et al., 2014 EEG P300 Bootstrapping Brain fingerprinting Lie detection
Petrov et al., 2014 EEG VEP Naive Bayes High-density Epilepsy or other
EEG systems BCI applications
Chi et al., 2011 EEG SSVEP CCA Dry and non-contact Typical BCI application
sensors
Iturrate et al., 2009 EEG P300 LDA Assistive Technology Wheelchair control

EEG, electroencephalography; P300, an event-related potential; SSVEP, steady-state visual evoked potential; VEP, visual evoked potential; LDA, linear discriminant analysis; CCA,
canonical correlation analysis; PCA, principal component analysis; SVM, support vector machine.

(Samek et al., 2013; Jayaram et al., 2016; Saha et al., 2018). with such a data-driven technique is over-fitting of the
The common spatial pattern, a supervised method, has been model parameters based on training sets, causing unreliable
extensively used in EEG-based online and offline BCI settings prediction on the test data (Sannelli et al., 2016). Recent studies
(Ramoser et al., 2000; Wu et al., 2017a). A common problem integrated diverse methods into potential transfer learning

Frontiers in Systems Neuroscience | www.frontiersin.org 6 February 2021 | Volume 15 | Article 578875

Saha et al. Progress in Brain Computer Interface

TABLE 4 | A summary of multimodal and hybrid brain computer interface (BCI) studies.

References Modality Wave\Task Data analysis Context Application

Zuo et al., 2020 EEG MI+P300 CSP, LDA Hybrid BCI Post-stroke rehabilitation
Rezazadeh Sereshkeh et al., 2019 EEG+fNIRS Imagined speech WT, LDA Multimodal BCI Robotic control
Corsi et al., 2019 EEG+MEG MI LDA Multimodal BCI Motor rehabilitation
Chiarelli et al., 2018 EEG+fNIRS MI Deep learning Multimodal BCI Motor rehabilitation
Ge et al., 2017 EEG+fNIRS MI CSP, SVM Multimodal BCI Motor rehabilitation
Zhao et al., 2016 EEG+FES SSVEP CSP Motor plasticity Paretic limb rehabilitation
Combaz and Van Hulle, 2015 EEG P300+SSVEP SVM Hybrid BCI Assistive control
Rao et al., 2014 EEG+TMS Visuomotor LF Hyperinteraction Brain-to-brain interface
Grau et al., 2014 EEG+TMS MI Spatial filter, Hyperinteraction Brain-to-brain interface
re-referencing
Choi and Jo, 2013 EEG P300+SSVEP+MI CSP, CCA Hybrid BCI-based Humanoid robot control
assistive
technology
EEG SSVEP+MI LDA Hybrid BCI Motor rehabilitation,
assistive technology
Kauhanen et al., 2006 EEG+MEG ME Particle filters Multimodal BCI Spinal cord injury

EEG, electroencephalography; fNIRS, functional near infrared spectroscopy; MEG, magnetoencephelography; FES, functional electrical stimulation; TMS, transcranial magnetic
stimulation; P300, an event-related potential; SSVEP, steady-state visual evoked potential; MI, motor imagery; ME, motor execution; WT, wavelet transform; CSP, common spatial
pattern; CCA, canonical correlation analysis; SVM, support vector machine; LDA, linear discriminant analysis; LF, Laplacian filter.

TABLE 5 | A summary of brain computer interface (BCI) studies involving invasive procedures.

References Modality Wave\Task Data analysis Context Application

Oxley et al., 2020 Stentrode SSSEP Spectral analysis Catheter angiography- Minimally invasive BCI:
Oxley et al., 2017 guided implantation human clinical trial in
Oxley et al., 2016 progress
Sauter-Starce et al., 2019 ECoG SSEP Spectral analysis WIMAGINE (Wireless Intracranial BCI: validation
Mestais et al., 2014 Implantable Multi-channel on in vivo sheep model
Acquisition system for
Generic Interface
with Neurons)
Sand et al., 2017 EEG DBS response Averaging Parkinson’s disease Motor rehabilitation
Kaiju et al., 2017 ECoG SEP Wavelet transform Finger stimulation Motor learning/ rehabilitation
Vansteensel et al., 2016 iMEA Visuomotor Autoregression Late-stage amyotrophic Motor rehabilitation
filter lateral sclerosis
Downey et al., 2016 iMEA ME Firing rate Vision-guided Robotic prosthetics control
estimator assistive technology
Pahwa et al., 2015 ECoG Sleep-Wake Welch’s Assistive technology Neuroprosthetic control
periodogram,
logistic regression
Yin et al., 2014 iMEA ME, Visuomotor, PCA Full-spectrum BCI, diagnostics,
Sleep-Wake electrophysiological therapeutic treatments
recording
Keefer et al., 2008 iMEA Video watching Spectrogram Carbon nanotube- Neural decoding
coated electrodes and stimulation
Friehs et al., 2006 iMEA MI Not specified Assistive technology/ Cursor control/
clinical use of BCI epilepsy monitoring

ECoG, electrocorticography; iMEA, intracortical microelectrode array; Stentrode, stent-electrode recording array; EEG, electroencephelography; ME, motor execution; SEP,
somatosensory evoked potential; MI, motor imagery; SSSEP, steady-state somatosensory evoked potential; DBS, deep brain stimulation; PCA, principal component analysis.

frameworks for BCI including spatial filters (e.g., common 3. NEUROPLASTICITY, SENSORS, SIGNAL
spatial pattern), Riemannian geometry, Euclidean alignment PROCESSING, MODELING, AND
and subspace adaptation and deep learning-based techniques
APPLICATIONS
(Barachant et al., 2011; Congedo et al., 2013, 2017; Marathe
et al., 2015; Wu et al., 2016, 2017b; Wu et al., 2020; He Exploiting neuroplasticity, designing hi-fidelity and customized
and Wu, 2019, 2020; Kwon et al., 2019; Zhang and Wu, neural sensors, applying advanced signal processing, and
2020). machine learning techniques are the key aspects of an effective

Frontiers in Systems Neuroscience | www.frontiersin.org 7 February 2021 | Volume 15 | Article 578875

Saha et al. Progress in Brain Computer Interface

BCI design. Tables 2–5 highlight diverse characteristics of design of feedback modality that has explicit association with the
BCI components and applications including signal acquisition neural signal classification performance, (3) the consideration of
modality, experimental paradigm, data analysis and pattern application-specific feedback delays, and (4) the utilization of a
recognition, application field, and significance. Notably, there are suitable feedback modality (Grosse-Wentrup et al., 2011). Neural
no specific selection criteria for studies in Tables 2–5 due to the ensemble recordings using signal acquisition modalities such as
broad spectrum of topics covered in this review; however, they EEG, MEG, fNIRS, and fMRI have become dominant over single
are summarized such that critical advances over the last several unit recordings. Behavioral activities are likely to be distributed
years can be appreciated. across three-dimensional cortical-subcortical networks and that
cannot be captured within single unit recordings (Nicolelis and
Lebedev, 2009).
3.1. Neuroplasticity and Cognitive Rehabilitative BCI can be designed either by attaching neural
Rehabilitation prostheses to the impaired body parts or by re-stimulating
The time-variant behavior of synapses within complex neural the damaged synaptic networks; in any of the cases, the
networks underpins the plastic characteristics of the brain and idea is to exploit and promote neuroplasticity (Wang et al.,
was first illustrated by Donald O. Hebb in 1949 (Brown and 2010; Ramos-Murguialday et al., 2013; Park et al., 2016;
Milner, 2003). Neuroplasticity not only helps to assist cognitive Darvishi et al., 2017; Toriyama et al., 2018; Song and Kim,
and perceptual learning but also is the main ingredient for 2019; Romero-Laiseca et al., 2020). In stroke patients with
neurorehabilitation. How plastic a particular brain area is, paretic muscles without residual finger movement, increased
may define the effectiveness of a neurofeedback strategy to electromyographic activity post rehabilitation by BCI-driven
induce specific activity patterns. Studies have shown visual orthoses exhibits increased neuromuscular coherence that is
cortices are plastic enough to produce robust neural signals essential for restoring movement control (Pfurtscheller et al.,
for post-neurofeedback perceptual learning (Shibata et al., 2000; Ramos-Murguialday et al., 2013). Explicit application
2011; Amano et al., 2016). Another study has demonstrated of functional electrical stimulation regulated by EEG-based
if right/left hemispheric differences in neurofeedback-induced movement-related signatures further suggests a role of BCI in
alpha activities are associated with visual information processing rehabilitation (Zhao et al., 2016). Increased electromyographic
and motor behaviors, and, thus, control spatial attention activity in paretic muscles is indicative of plasticity induced by
(Jones and Sliva, 2020). fMRI-based neurofeedback training electrical stimulation (De Marchis et al., 2016). For BCI-based
sessions induce the plasticity of attention-related behavior. rehabilitation in a real-life environment, differentiating between
Implications suggest that neurofeedback can offer rehabilitation task-induced activities and resting state activities is a key factor
of attentional deficit (Megan et al., 2015). A recent study has used for controlling the prosthesis or stimulation modality (Pahwa
neurofeedback to generate robust somatosensory oscillations et al., 2015).
associated with human perception (Brickwedde et al., 2019). Externally stimulating the affected brain areas by electric
Closed-loop BCI with neurofeedback is assumed to contribute or magnetic fields holds promise for stroke rehabilitation. A
to the reorganization of cortical-subcortical neural networks recent study demonstrated the induction of neuroplasticity in
and assist subjects in self-regulating specific brain rhythms; white matter and cortical functions in chronic stroke patients
notwithstanding, the underlying mechanisms that alter neural by motor imagery-based BCI and transcranial direct current
substrates are still not fully-understood (Sitaram et al., 2017). stimulation applied to targeted brain areas (Hong et al., 2017).
For example, BCI-based covert visuomotor training modulates Magnetic stimulation of brain areas driven by BCI increases
associated neural substrates, where the effects of modulated cortical activation in stroke patients (Johnson et al., 2018).
neural substrates are observed while performing that particular The level of neuroplasticity achieved post-rehabilitation varies
movement-related task (Vyas et al., 2018). Substantial changes across subjects and, thus, an individual-specific training session
in overt movement-related task following BCI-driven training is necessary (Leamy et al., 2014). The use of BCI-based motor
induced learning suggest a critical role of BCI in enhanced rehabilitation for locked-in patients is limited because they are
motor learning for proficiently controlling neuroprosthetics unable to fully interact with the system (Birbaumer and Cohen,
(Orsborn et al., 2014), i.e., devices that can enhance or repair 2007). Other examples of BCI-driven rehabilitations include
the output of the nervous system. For example, intracortical optimizing the parameters for deep brain stimulation applied
electrodes may be used to stimulate specific brain regions to into the subthalamic nucleus in patients with Parkinson’s disease
regain motor control (Oxley et al., 2016; Sand et al., 2017). (Sand et al., 2017) and treating major depressive disorder by
BCI may augment training-induced plasticity during therapeutic BCI-driven transcranial magnetic stimulation (Ray et al., 2015).
motor rehabilitation and, thus, re-excite corresponding neural Either by providing direct control of assistive technologies or
substrates to regain control by means of neuroprosthetics or by direct neurostimulation, BCI can help patients who may suffer
upper limb functions (Dobkin, 2007). Other examples include from amyotrophic lateral sclerosis, cerebral palsy, brainstem
BCI-driven exoskeletons to enhance human working capacity stroke, spinal cord injuries, muscular dystrophies, or chronic
(Benabid et al., 2019). peripheral neuropathies (Kauhanen et al., 2006; Iturrate et al.,
The extent of BCI-induced plasticity entails several factors, 2009; Mak and Wolpaw, 2009; Allison et al., 2010; Ramos-
including (1) the selection of the signal acquisition modality, Murguialday et al., 2013; Leamy et al., 2014; Botrel et al., 2015;
which plays an important role in diagnosing neural states, (2) the Combaz and Van Hulle, 2015; Edelman et al., 2015; Park et al.,

Frontiers in Systems Neuroscience | www.frontiersin.org 8 February 2021 | Volume 15 | Article 578875

Saha et al. Progress in Brain Computer Interface

2016; Zhao et al., 2016; Alcaide-Aguirre et al., 2017; Ge et al., Since the first publication in 2000, common spatial pattern is
2017; Chiarelli et al., 2018; Guy et al., 2018; Yu et al., 2018; still one of the most popular methods to represent multichannel
Rezazadeh Sereshkeh et al., 2019; Jin et al., 2020; Zuo et al., 2020). EEG signals by corresponding spatial contents (Ramoser et al.,
Providing auxiliary degrees of freedom improves the quality of 2000). As a data-driven method, it requires a significant
life of people with disabilities significantly. Brain signals can be number of training samples to model the filtering parameters.
translated to drive wheelchairs (Galán et al., 2008; Iturrate et al., In case of small training trials, regularizing the covariance
2009; Perdikis et al., 2017; Tonin and Millán, 2020). Integration estimation works better than the traditional algorithm (Lotte
of BCI with a vision-guided autonomous system was shown and Guan, 2010). Other modifications in spatial filtering
to effectively perform the grasping task using a prosthetic arm include projecting EEG by using sparse representation and filter
in a tetraplegic patient (Downey et al., 2016). An implanted bank spectral division of raw signals (Arvaneh et al., 2014).
microelectrode array has been proposed to operate a three- Generally, spatial filtering is applicable in subject-specific BCI
dimensional neuroprosthetic device (Taylor et al., 2002). development although recent studies have proposed estimating
the filter coefficients from a subject and applied that filter to
another subject, which contributed no training sample (Saha
3.2. Signal Acquisition, Signal Processing, et al., 2017, 2018, 2019a). Other popular data-driven methods
and Modeling include linear discriminant analysis, support vector machine and
A significant number of studies are now involved in combining principal component analysis (Lotte et al., 2007, 2018). With
multimodal signal acquisition modalities to augment current the exceptional advancements in computational facilities in the
BCI systems. For example, simultaneous EEG and fMRI last decade, deep learning-based BCI paradigms by allowing the
yield complementary features by exploiting good temporal evaluation of large datasets could soon become a trend in the
resolution of EEG and good spatial resolution of fMRI community (Chiarelli et al., 2018; Kwon et al., 2019; Nagel and
(Debener et al., 2006). Enhanced multiclass sensorimotor Spüler, 2019).
tasks classification performance using hybrid EEG and fNIRS On the other hand, independent component analysis is a blind
signals implicates the importance of features extracted from source separation method requiring no training. The estimation
both hemodynamic and electrical activities (Buccino et al., of independent components is based on statistical properties of
2016). MEG is another potential tool to combine with EEG, the signals (Bell and Sejnowski, 1995). However, modeling the
as it captures radially/tangentially dipole sources in cortical- actual cortical sources as dipoles in the complex brain anatomy
subcortical networks and adds complementary information to from the scalp EEG recordings seeks to solve the so-called inverse
EEG signals (Kauhanen et al., 2006). Skepticism might still problem (Qin et al., 2004; Kamousi et al., 2005; Wronkiewicz
present about the detection of brain activities originated from et al., 2015; Saha et al., 2019a). More recent source localization
subcortical areas; however, an increasing number of studies methods such as wavelet-based maximum entropy on the mean
argue that EEG and MEG could capture subcortical activities represent EEG/MEG signals as relevant time-frequency contents
(Andersen et al., 2019; Min et al., 2020; Piastra et al., 2020). and finally transform them into spatial representations (Lina
A recent trend is to combine different signal acquisition et al., 2012; Saha et al., 2019a). Notably, different inverse
modalities together to improve BCI efficiency. Table 4 highlights methods and toolboxes demonstrate considerable variability in
multimodal and hybrid BCI applications. localized sources (Mahjoory et al., 2017). Even it is not very
The combination of signal processing and machine learning straightforward to know the exact sources, which are to be
approaches plays critical role in translating any brain signal modeled using EEG/MEG. For example, the ground truth defined
to a command for a computer or other external devices. by implanted electrodes might not be 100% reliable because of
Tables 2–5 highlight different signal processing and machine sparse (spatial) sampling. In the case of fMRI, the measurement
learning techniques. Representing signals in the time-frequency- of neural activity is indirect. Notwithstanding, inverse methods
space is necessary to obtain physiological correlates of BCI have shown promise for designing various BCI models (Qin et al.,
outcomes (McFarland et al., 2006; Bashashati et al., 2007). 2004; Kamousi et al., 2005; Wronkiewicz et al., 2015; Saha et al.,
Fourier transform (FT) and autoregressive models are examples 2019a).
of time domain representations of brain signals while short
time FT and wavelet transform are examples of time-frequency 3.3. Neurosensors: The-State-of-the-Art
representations (McFarland et al., 2006; Bashashati et al., 2007). Deeper regions of the brain, e.g., subcortical and cerebellar
In case of spatial filtering, the most popular filtering approaches regions, contribute to various neuronal activities (Müller et al.,
are common spatial pattern, independent component analysis 2002; Wardman et al., 2014). Interpreting the genesis of
and the Laplacian filter. A diverse range of inverse models allow cortical sources from cellular to scalp levels and RSNs spanned
to discern the actual sources projected on three-dimensional throughout the three-dimensional brain space can guide BCI
cortical-subcortical networks (Wronkiewicz et al., 2015; Saha development (Donoghue, 2008). Sensors with customized design
et al., 2019a). Extracted features can be translated using various are developed to advance brain signal acquisition modalities.
linear and non-linear classification algorithms. Examples of Neurosensors can be constructed in different forms like electrical,
linear and non-linear classifier models are linear discriminant optical, chemical and biological (Deisseroth and Schnitzer,
analysis and non-linear kernel-based support vector machines 2013). Dry EEG electrodes are convenient, but assumed to
(Lotte et al., 2007, 2018). provide lower signal-to-noise ratio compared to conventional

Frontiers in Systems Neuroscience | www.frontiersin.org 9 February 2021 | Volume 15 | Article 578875

Saha et al. Progress in Brain Computer Interface

wet electrodes. Wet electrodes may cause inconvenience to Besides large-scale recording modalities like EEG and MEG,
users as they use conductive gel and require proper skin very small-scale recordings of neuronal activities are crucial
preparation for minimizing the skin-electrode impedance (Liao for understanding brain circuits’ functions and intra- and
et al., 2012). However, a study on dry electrodes-based BCI inter-neuron interactions. Representation of any cognitive task
suggested that dry electrode could be used to collect good as functions of both small-scale and large-scale neuronal
quality signals by designing the circuits carefully (Chi et al., interactions is crucial for the advancement of neuroengineering
2011). Further studies support dry electrodes with wireless and BCI. In this regard, a high-density neurosensor array made
systems that could offer comparable signal quality as of wet from silicon probes combined with optogenetics enables single
electrodes, but with more convenience (Di Flumeri et al., 2019; unit recordings (Buzsáki et al., 2015). Yang et al. proposed
Marini et al., 2019; Hinrichs et al., 2020). While utilizing the a novel multi-plane two-photon microscope that can be used
advantages of both dry and wet electrodes, quasi-dry electrodes to capture multi-layer neuronal structure and mechanism with
exploiting the mechanical properties of polymer can capture cellular resolution (Yang et al., 2016). Other potential imaging
signals as comparable to commercial Ag/AgCl electrodes (Mota methods for investigating cell signaling include calcium imaging
et al., 2013). To increase the spatial resolution of EEG, Petrov (Grienberger and Konnerth, 2012) and advanced microscope
et al. have proposed an ultra-dense sensor array of 700–800 with chronically implanted lenses (Resendez et al., 2016).
electrodes (Petrov et al., 2014). The signal-to-noise ratio was Designer receptor exclusively activated by designer drugs,
twice as high as for high-density EEG that has up to 256 provides a chemogenetic tool to understand cell-signaling
gold-coated electrodes. An auricle electrode with stretchable including electrical activities in molecularly clustered cell groups
connector was proposed that not only can increase portability (Sternson and Roth, 2014; Roth, 2016). A new ultrasonic-based
but also can offer a comfortable alternative for long term wireless system, called neural dust, enables the recording of
recordings (Norton et al., 2015). The electrode is flexible electromyogram and electroneurogram on the millimeter scale
with the alterations of electrical and mechanical properties (Seo et al., 2016).
of skin.
Invasive sensors must be biocompatible. A novel organic
electrochemical transistor-based sensor enables to collect neural
signals directly from the brain surface (Khodagholy et al., 3.4. Affective Computing, Gaming,
2013). This sensor is biocompatible and mechanically flexible, Robotics, and Miscellaneous Applications
and the transistor-based design amplifies captured signals Future computers are assumed to have emotional and perceptual
locally, thus providing much better signal-to-noise ratio than capabilities, which could extend the use not only to assisting
conventional ECoG. To enhance the signal quality, carbon humans but also to making decisions (Picard, 2000). Computers
nanotube coating can decrease the electrode impedance and, might able to recognize and interpret underlying affective states
thus, increase the charge transfer (Keefer et al., 2008). Another based on physiological and behavioral variables. Recent studies
invasive biocompatible sensor, designed for recording previously demonstrated BCI is a potential tool to investigate affective states,
inaccessible spectra of large neuron populations, includes data expanding the applications into psychology (Piho and Tjahjadi,
transmission for use in natural environments (Yin et al., 2018; Song et al., 2018; Huang et al., 2019). Huang et al. have
2014). With the outstanding progress of nanotechnology, proposed an EEG-based BCI to detect positive and negative
nanowire Field Effect Transistor and other p/n junction devices emotions induced by video stimulus (Huang et al., 2019).
have potential for neuro-sensing modalities for intracellular Integration of arts into BCI is referred to as artistic BCI
recordings, even in the deep brain regions (Kruskal et al., 2015). (Andujar et al., 2015). In the late 1960s, David Rosenboom
Oxley et al. have proposed stent-electrode array (stentrode) began experimenting with ways to link brain functions with
that involves minimal invasiveness (Oxley et al., 2016, 2017). musical production, perception of musical forms and musical
Using computer-guided catheter angiography, the stentrode can proprioception (Rosenboom, 2014). Other examples of artistic
be placed within arteries or veins located inside the brain BCI include affective states detection, playing video games and
anatomy. Capturing high-fidelity cortical signals, this technology controlling virtual/augmented reality environment. Studies have
will significantly reduce the risk factors of craniotomy. A follow- demonstrated that a user can fully operate video games by
up study has recently demonstrated successful implantation SSVEP-BCI (van Vliet et al., 2012; Filiz and Arslan, 2020).
of the strentrode in humans for long-term neural signal Other studies have proposed how multiple users can participate
recording (Oxley et al., 2020). The information transfer rate for in a collaborative game, in which joint decision making is
strentrode-based BCI was comparable to the landmark study by required to control the gaming environment (Nijholt and Poel,
Vansteensel et al. with implanted electrodes (Vansteensel et al., 2016; Sekhavat, 2020). Another study previously suggested
2016). Another implanted ECoG recorder called as WIMAGINE the aggregation of information from two intelligence analysts’
(Wireless Implantable Multi-channel Acquisition system for brain signals may lead to better decision making than one’s
Generic Interface with Neurons) allows wireless neural data brain signals (Stoica, 2012). The underlying cause could be
access (Mestais et al., 2014). The WIMAGINE has recently been explained by inter-individual differences in human cognitive
tested for long-term reliability of data acquisition and any risk and perceptual skills (Kleinschmidt et al., 2012). Collaboration
associated with craniotomy (Kruskal et al., 2015; Sauter-Starce between users might assist an individual’s decision making by
et al., 2019). diversity inclusion. A modified setup could investigate how

Frontiers in Systems Neuroscience | www.frontiersin.org 10 February 2021 | Volume 15 | Article 578875

Saha et al. Progress in Brain Computer Interface

people interact in different social contexts, extending BCI games (Stocco et al., 2015). Figure 2 illustrates a timeline for
applications in sociology (Amaral et al., 2017). current advances of BCI in diverse applications.
Virtual/augmented reality (V/AR) technologies together with
BCI could offer immersive experiences and have many potential
applications including arts and neurofeedback (Andujar et al., 4. ETHICAL CONCERNS AND
2015; Tremmel et al., 2019; Putze et al., 2020). Brain painting SOCIOECONOMIC CONTEXTS
allows a user to draw lines in a virtual canvas by brain signals,
which gives an alternative communication channel for people Irrespective of the scientific breakthroughs in BCI field, there are
with paretic motor functions (Botrel et al., 2015). McClinton key factors pertaining to safety, ethics, privacy protection and
et al. have developed a brain painting application using VR data confidentiality, community acceptance and socioeconomic
environment (McClinton et al., 2019). Another work has evinced aspects that should be considered with adequate precautions
VR-BCI to measure cognitive workload that can contribute to to maximize users’ benefits and social impacts (Illes and Bird,
neuroergonomics (Tremmel et al., 2019). Studies have also used 2006; Bostrom and Sandberg, 2009; Jebari, 2013; McCullagh
immersive VR as a better neurofeedback option as compared et al., 2014). Obtaining an ethically sound informed consent
to the computer screen leading to increased BCI accuracy (Luu from a BCI worn patient may be challenging for BCI researchers
et al., 2016; Škola et al., 2019; Vourvopoulos et al., 2019; Juliano due to difficulty in communicating and the lack of alternatives.
et al., 2020). Vourvopoulos et al. have integrated the principles However, more awareness and attention to ethics policies
of VR and BCI into a platform called REINVENT for motor are recommended to improve the chance for patients to get
rehabilitation (Vourvopoulos et al., 2019). Likewise, BCI with AR adequate information.
can be used to remotely control a robot for rehabilitating children Physical and mental safety of BCI users is important. Invasive
with attention-deficit/hyperactivity disorder (Arpaia et al., 2020). procedures such as deep brain stimulation and intracortical
While BCI-driven robotic controllers can offer advanced microelectrode array may cause postoperative psychological and
assistive technology for people with mobility constraint, it neurological side effects (Jotterand and Giordano, 2011; Gilbert,
may also augment human ergonomic performance for healthy 2015; Maslen et al., 2015). Additionally, bleeding and infections
subjects (Millan et al., 2004; Gandhi et al., 2014; Tidoni et al., are infrequent but do occur and may require removal or further
2016; Perdikis et al., 2017; Spataro et al., 2017; Yuan and Li, 2018; maintenance of the implanted electrodes. Guidelines are required
Deng et al., 2019; Tonin et al., 2019; Tonin and Millán, 2020). to safely advance neurotechnologies (Goering and Yuste, 2016),
EEG-based BCI-driven controller of mobile robot or wheelchair because BCI devices can alter behavior and, thus, introduce
has demonstrated the possibility of this technology in robotics potential threats to one’s emotions, personality and memories;
industry (Millan et al., 2004; Tidoni et al., 2016; Perdikis et al., more generally one’s mind. For human brain-to-brain interface
2017; Yuan and Li, 2018; Deng et al., 2019; Tonin et al., 2019). BCI applications (Rao et al., 2014; Stocco et al., 2015), one may define
can also be used for controlling humanoid robots remotely using an upper bound for research depths keeping in mind the necessity
EEG (Spataro et al., 2017), suitable in hazardous environments, of ethical utilization of this technology. Because both sender
for example by sending a robot in a coal mine for executing and receiver play complicated roles, more specifically, sender’s
a task that is potentially unsafe for a human. In space, BCI intentional manipulative control over neural signals might alter
can be used to monitor astronauts’ working capacity and to the anticipated outcome. Altering human cognitive and possibly
drive an exoskeleton (Menon et al., 2009; de Negueruela et al., moral capacity raise a serious ethical question and it is not
2011). In the absence of gravity, working becomes tedious and predictable if the cognitive changes reversible and efficacious
inconvenient. Furthermore, astronauts’ working time is precious. (Nakazawa et al., 2016).
BCI-driven systems could be practical for improving astronauts’ A user’s expectations of achieving extended or auxiliary
functionality, efficiency and safety (Summerer et al., 2009; Farwell degree of freedom may not be fulfilled, and even the unfamiliar
et al., 2014; Botrel et al., 2015; Ortiz et al., 2016; Wang et al., risk factors can diminish the accomplished advantage of using
2016a; Vourvopoulos et al., 2019; Singh et al., 2020). BCI (Clausen, 2011; Schicktanz et al., 2015). Creating broad
Recently, brain-to-brain interface (BBI) experiments that awareness of BCI technology and its pros and cons would educate
involve decoding sender’s cognitive intentions, translate them people, who fear unnecessary technological dependency (Hobson
into commands for stimulating receiver’s brain, have been et al., 2017). However, successful clinical trials of sophisticated
explored (Pais-Vieira et al., 2013; Rao et al., 2014; Jiang devices such as strentrode or WIMAGINE are essential to
et al., 2019). In 2013, researchers implemented a direct BBI demonstrate potential advantages, especially for people suffering
system in which one rat was able to share sensorimotor from any form of cognitive disability (Sauter-Starce et al., 2019;
information to another rat (Pais-Vieira et al., 2013). Intracortical Oxley et al., 2020). In the case of healthy users, it should not
microstimulation was used to stimulate the receiver’s target brain be too difficult to create acceptance to a broader community
areas. An early attempt to develop sensorimotor rhythm-based when dry electrodes could offer the long-term operation of a BCI
BBI between two human subjects used non-invasive EEG and application with little maintenance effort (Di Flumeri et al., 2019;
transcranial magnetic stimulation has been proposed by Rao et al. Marini et al., 2019; Hinrichs et al., 2020).
(2014). Other total non-invasive BBI experiments have proposed It is critical to introduce a suitable act for lawful utilization
sharing pseudo-random binary streams encoded words between of BCI and preservation of privacy and confidentiality of
human subjects (Grau et al., 2014) and playing collaborative stored data. Recent studies have demonstrated decoding of

Frontiers in Systems Neuroscience | www.frontiersin.org 11 February 2021 | Volume 15 | Article 578875

Saha et al. Progress in Brain Computer Interface

FIGURE 2 | A schematic illustration of the evolution of the brain computer interface (BCI) applications: Cognitive & Perceptual Learning/Rehabilitation (McMillan et al.,
1995); Orthosis Control (Pfurtscheller et al., 2000); Music BCI (Rosenboom, 2014); Robotics (Millan et al., 2004); Wheelchair Control (Iturrate et al., 2009); Drowsiness
Detection (Lin et al., 2008); Affective Computing (Zander et al., 2009); Brain Racers (Perdikis et al., 2017); Multiplayer Gaming (Nijholt and Poel, 2016); Brain-to-Brain
Interface (Rao et al., 2014).

password or recognizing faces utilizing consumer-grade BCI advancing the knowledge of neuroscience, for example, by
successfully, prompting a potential concern of any illegitimate strengthening efforts to fund neuroscience research projects
access to users’ raw data and their further exploitation (Martin (Grillner et al., 2016). The European Union along with its
et al., 2016; Alomari et al., 2019). For example, affective states partner universities have initiated the Human Brain Project.
define users’ moral judgment and emotional traits. Thus, it In addition, the Brain Initiative has been announced by the
is critical to limit the applications of affective BCI while White House. In our opinion, advanced understanding of basic
preserving sensitive information (Steinert and Friedrich, 2020). neuroscientific phenomena will determine the structure, efficacy
Necessary precursor initiatives should propose application- and applications of futuristic BCI.
specific BCI frameworks, which can restrict unauthorized access
to stored data or the system (Ienca and Haselager, 2016).
For example, illicit access to a wireless BCI-driven limb and 5. CONCLUSION
manipulative reprogramming of a computer-guided neuro-
Numerous groundbreaking advances in neurosensors and
stimulation have demonstrated the importance of establishing
computational tools herald great promise for more sophisticated
resilient safeguards to BCI use (Denning et al., 2009). Agarwal
and user friendly BCI systems requiring no or little maintenance.
et al. have proposed cryptographic protocols as integrated
In addition to hi-fidelity signal acquisition, significant
parts of BCI to preserve the privacy of a user by keeping
progress in signal processing and machine learning tools,
confidential information obscure to others (Agarwal et al., 2019).
their complementary roles, and high computation power and
Without evaluating socioeconomic, ethical and policy issues, the
increased mobility of computers have significantly contributed
commercialization of BCI would hinder the progress in this
in the emergence of BCI technologies. The future of BCI
field (Eaton and Illes, 2007).
technology will rely greatly on addressing the following
By creating a common networking platform for BCI
key aspects:
researchers worldwide, the immediate proposition of a
comprehensive list of universal guidelines is key to sustainable • Elucidating the underlying psychophysiological
advancements of the field (Vaadia and Birbaumer, 2009). Various and neurological factors that potentially influence
alliance-based projects are running as common platforms for BCI performance.

Frontiers in Systems Neuroscience | www.frontiersin.org 12 February 2021 | Volume 15 | Article 578875

Saha et al. Progress in Brain Computer Interface

• Designing less invasive sensors with reliable signal in the discussion and commented on the draft. AHK and
acquisition and resolution, while considering portability, MB provided further insight and helped SS to finalize the
easy maintenance, and affordability. structure and materials. All authors read and approved the
• Modeling session-to-session and subject-to-subject final paper.
information transfer for the proposition of more generalized
BCI models with insignificant or no calibration requirement. ACKNOWLEDGMENTS
• Establishing broad consensus on ethical issues and beneficial
socioeconomic application of this technology. This manuscript has been released as a pre-print at https://arxiv.
org/ (Saha et al., 2019b). Authors would like to thank Prof.
AUTHOR CONTRIBUTIONS Moritz Grosse-Wentrup for providing his valuable feedback. This
work was partially supported by a grant (Award No. RC2-2018-
SS conceived the initial idea, wrote the first draft, and generated 022 (HEIC) and KKJRC-2019-Health 2) from Khalifa University,
all figures and tables. KM, KA, RM, GN, and SD participated Abu Dhabi, UAE.

REFERENCES control environment. Front. Hum. Neurosci. 10:539. doi: 10.3389/fnhum.2016.

00539
Abrams, D. A., Ryali, S., Chen, T., Chordia, P., Khouzam, A., Levitin, D. J., et al. Arpaia, P., Duraccio, L., Moccaldi, N., and Rossi, S. (2020). Wearable
(2013). Inter-subject synchronization of brain responses during natural music brain-computer interface instrumentation for robot-based rehabilitation
listening. Eur. J. Neurosci. 37, 1458–1469. doi: 10.1111/ejn.12173 by augmented reality. IEEE Trans. Instrument. Meas. 69, 6362–6371.
Abu-Alqumsan, M., and Peer, A. (2016). Advancing the detection of steady-state doi: 10.1109/TIM.2020.2970846
visual evoked potentials in brain-computer interfaces. J. Neural Eng. 13:036005. Arvaneh, M., Guan, C., Ang, K. K., and Quek, C. (2014). Mutual information-
doi: 10.1088/1741-2560/13/3/036005 based optimization of sparse spatio-spectral filters in brain-computer interface.
Acqualagna, L., Botrel, L., Vidaurre, C., Kübler, A., and Blankertz, B. Neural Comput. Appl. 25, 625–634. doi: 10.1007/s00521-013-1523-7
(2016). Large-scale assessment of a fully automatic co-adaptive motor Barachant, A., Bonnet, S., Congedo, M., and Jutten, C. (2011). Multiclass brain-
imagery-based brain computer interface. PLoS ONE 11:e0148886. computer interface classification by riemannian geometry. IEEE Trans. Biomed.
doi: 10.1371/journal.pone.0148886 Eng. 59, 920–928. doi: 10.1109/TBME.2011.2172210
Agarwal, A., Dowsley, R., McKinney, N. D., Wu, D., Lin, C.-T., De Bashashati, A., Fatourechi, M., Ward, R. K., and Birch, G. E. (2007). A survey of
Cock, M., et al. (2019). Protecting privacy of users in brain-computer signal processing algorithms in brain-computer interfaces based on electrical
interface applications. IEEE Trans. Neural Syst. Rehabil. Eng. 27, 1546–1555. brain signals. J. Neural Eng. 4:R32. doi: 10.1088/1741-2560/4/2/R03
doi: 10.1109/TNSRE.2019.2926965 Baxter, B. S., Edelman, B. J., Sohrabpour, A., and He, B. (2017). Anodal transcranial
Ahn, M., and Jun, S. C. (2015). Performance variation in motor imagery direct current stimulation increases bilateral directed brain connectivity
brain-computer interface: a brief review. J. Neurosci. Methods 243, 103–110. during motor-imagery based brain-computer interface control. Front. Neurosci.
doi: 10.1016/j.jneumeth.2015.01.033 11:691. doi: 10.3389/fnins.2017.00691
Alcaide-Aguirre, R., Warschausky, S., Brown, D., Aref, A., and Huggins, J. (2017). Bell, A. J., and Sejnowski, T. J. (1995). An information-maximization approach
Asynchronous brain-computer interface for cognitive assessment in people to blind separation and blind deconvolution. Neural Comput. 7, 1129–1159.
with cerebral palsy. J. Neural Eng. 14:066001. doi: 10.1088/1741-2552/aa7fc4 doi: 10.1162/neco.1995.7.6.1129
Allison, B. Z., Brunner, C., Kaiser, V., Müller-Putz, G. R., Neuper, C., and Benabid, A. L., Costecalde, T., Eliseyev, A., Charvet, G., Verney, A., Karakas, S.,
Pfurtscheller, G. (2010). Toward a hybrid brain-computer interface based et al. (2019). An exoskeleton controlled by an epidural wireless brain-machine
on imagined movement and visual attention. J. Neural Eng. 7:026007. interface in a tetraplegic patient: a proof-of-concept demonstration. Lancet
doi: 10.1088/1741-2560/7/2/026007 Neurol. 18, 1112–1122. doi: 10.1016/S1474-4422(19)30321-7
Alomari, R., Martin, M. V., MacDonald, S., Maraj, A., Liscano, R., and Bellman, C. Berger, H. (1929). Über das elektroenkephalogramm des menschen. Arch.
(2019). Inside out-a study of users’ perceptions of password memorability and Psychiatr. Nervenkrankheiten 87, 527–570. doi: 10.1007/BF01797193
recall. J. Inform. Security Appl. 47, 223–234. doi: 10.1016/j.jisa.2019.05.009 Bhattacharyya, S., Clerc, M., and Hayashibe, M. (2019). Augmenting
Amano, K., Shibata, K., Kawato, M., Sasaki, Y., and Watanabe, T. (2016). Learning motor imagery learning for brain-computer interfacing using electrical
to associate orientation with color in early visual areas by associative decoded stimulation as feedback. IEEE Trans. Med. Robot. Bionics 1, 247–255.
fMRI neurofeedback. Curr. Biol. 26, 1861–1866. doi: 10.1016/j.cub.2016.05.014 doi: 10.1109/TMRB.2019.2949854
Amaral, C. P., Simões, M. A., Mouga, S., Andrade, J., and Castelo-Branco, M. Birbaumer, N., and Cohen, L. G. (2007). Brain-computer interfaces:
(2017). A novel brain computer interface for classification of social joint communication and restoration of movement in paralysis. J. Physiol. 579,
attention in autism and comparison of 3 experimental setups: a feasibility study. 621–636. doi: 10.1113/jphysiol.2006.125633
J. Neurosci. Methods 290, 105–115. doi: 10.1016/j.jneumeth.2017.07.029 Birbaumer, N., Ghanayim, N., Hinterberger, T., Iversen, I., Kotchoubey, B., Kübler,
Andersen, L. M., Jerbi, K., and Dalal, S. S. (2019). Can electro-and A., et al. (1999). A spelling device for the paralysed. Nature 398, 297–298.
magnetoencephalography detect signals from the human cerebellum? PeerJ doi: 10.1038/18581
7:e27901. doi: 10.7287/peerj.preprints.27901 Blankertz, B., Sanelli, C., Halder, S., Hammer, E., Kübler, A., Müller, K.-R., et al.
Andersen, L. M., Jerbi, K., and Dalal, S. S. (2020). Can EEG and MEG (2009). Predicting BCI performance to study BCI illiteracy. BMC Neurosci.
detect signals from the human cerebellum? Neuroimage 215:116817. 10(Suppl. 1):P84. doi: 10.1186/1471-2202-10-S1-P84
doi: 10.1016/j.neuroimage.2020.116817 Bockbrader, M., Annetta, N., Friedenberg, D., Schwemmer, M., Skomrock, N.,
Andujar, M., Crawford, C. S., Nijholt, A., Jackson, F., and Gilbert, J. Colachis, I. V., et al. (2019). Clinically significant gains in skillful grasp
E. (2015). Artistic brain-computer interfaces: the expression and coordination by an individual with tetraplegia using an implanted brain-
stimulation of the user’s affective state. Brain Comput. Interfaces 2, 60–69. computer interface with forearm transcutaneous muscle stimulation. Arch.
doi: 10.1080/2326263X.2015.1104613 Phys. Med. Rehabil. 100, 1201–1217. doi: 10.1016/j.apmr.2018.07.445
Aricó, P., Borghini, G., Di Flumeri, G., Colosimo, A., Bonelli, S., Golfetti, A., Bostrom, N., and Sandberg, A. (2009). Cognitive enhancement:
et al. (2016). Adaptive automation triggered by EEG-based mental workload methods, ethics, regulatory challenges. Sci. Eng. Ethics 15, 311–341.
index: a passive brain-computer interface application in realistic air traffic doi: 10.1007/s11948-009-9142-5

Frontiers in Systems Neuroscience | www.frontiersin.org 13 February 2021 | Volume 15 | Article 578875

Saha et al. Progress in Brain Computer Interface

Botrel, L., Holz, E., and Kübler, A. (2015). Brain painting v2: evaluation of Debener, S., Ullsperger, M., Siegel, M., and Engel, A. K. (2006). Single-trial EEG-
p300-based brain-computer interface for creative expression by an end-user fMRI reveals the dynamics of cognitive function. Trends Cogn. Sci. 10, 558–563.
following the user-centered design. Brain Comput. Interfaces 2, 135–149. doi: 10.1016/j.tics.2006.09.010
doi: 10.1080/2326263X.2015.1100038 Deisseroth, K., and Schnitzer, M. J. (2013). Engineering approaches
Brickwedde, M., Krüger, M. C., and Dinse, H. R. (2019). Somatosensory to illuminating brain structure and dynamics. Neuron 80, 568–577.
alpha oscillations gate perceptual learning efficiency. Nat. Commun. 10, 1–9. doi: 10.1016/j.neuron.2013.10.032
doi: 10.1038/s41467-018-08012-0 Deng, X., Yu, Z. L., Lin, C., Gu, Z., and Li, Y. (2019). A bayesian shared control
Brown, R. E., and Milner, P. M. (2003). The legacy of Donald O. Hebb: more than approach for wheelchair robot with brain machine interface. IEEE Trans.
the Hebb synapse. Nat. Rev. Neurosci. 4, 1013–1019. doi: 10.1038/nrn1257 Neural Syst. Rehabil. Eng. 28, 328–338. doi: 10.1109/TNSRE.2019.2958076
Brunner, C., Birbaumer, N., Blankertz, B., Guger, C., Kübler, A., Mattia, D., et al. Denning, T., Matsuoka, Y., and Kohno, T. (2009). Neurosecurity:
(2015). BNCI horizon 2020: towards a roadmap for the BCI community. Brain security and privacy for neural devices. Neurosurg. Focus 27:E7.
Comput. Interfaces 2, 1–10. doi: 10.1080/2326263X.2015.1008956 doi: 10.3171/2009.4.FOCUS0985
Buccino, A. P., Keles, H. O., and Omurtag, A. (2016). Hybrid EEG-fNIRS Di Flumeri, G., Aricó, P., Borghini, G., Sciaraffa, N., Di Florio, A., and Babiloni,
asynchronous brain-computer interface for multiple motor tasks. PLoS ONE F. (2019). The dry revolution: evaluation of three different EEG dry electrode
11:e0146610. doi: 10.1371/journal.pone.0146610 types in terms of signal spectral features, mental states classification and
Buzsáki, G., Stark, E., Berényi, A., Khodagholy, D., Kipke, D. R., Yoon, E., et al. usability. Sensors 19:1365. doi: 10.3390/s19061365
(2015). Tools for probing local circuits: high-density silicon probes combined Dobkin, B. H. (2007). Brain-computer interface technology as a tool to augment
with optogenetics. Neuron 86, 92–105. doi: 10.1016/j.neuron.2015.01.028 plasticity and outcomes for neurological rehabilitation. J. Physiol. 579, 637–642.
Calhoun, V. D., and Adali, T. (2016). Time-varying brain connectivity doi: 10.1113/jphysiol.2006.123067
in fMRI data: whole-brain data-driven approaches for capturing and Donoghue, J. P. (2008). Bridging the brain to the world: a perspective on neural
characterizing dynamic states. IEEE Signal Process. Mag. 33, 52–66. interface systems. Neuron 60, 511–521. doi: 10.1016/j.neuron.2008.10.037
doi: 10.1109/MSP.2015.2478915 Downey, J. E., Weiss, J. M., Muelling, K., Venkatraman, A., Valois, J.-S., Hebert,
Cecotti, H. (2020). Adaptive time segment analysis for steady-state visual evoked M., et al. (2016). Blending of brain-machine interface and vision-guided
potential based brain-computer interfaces. IEEE Trans. Neural Syst. Rehabil. autonomous robotics improves neuroprosthetic arm performance during
Eng. 28, 552–560. doi: 10.1109/TNSRE.2020.2968307 grasping. J. Neuroeng. Rehabil. 13, 1–12. doi: 10.1186/s12984-016-0134-9
Chen, X., Wang, Y., Nakanishi, M., Gao, X., Jung, T.-P., and Gao, S. (2015). High- Eaton, M. L., and Illes, J. (2007). Commercializing cognitive neurotechnology–the
speed spelling with a noninvasive brain-computer interface. Proc. Natl. Acad. ethical terrain. Nat. Biotechnol. 25, 393–397. doi: 10.1038/nbt0407-393
Sci. U.S.A. 112, E6058–E6067. doi: 10.1073/pnas.1508080112 Edelman, B. J., Baxter, B., and He, B. (2015). EEG source imaging enhances the
Chi, Y. M., Wang, Y.-T., Wang, Y., Maier, C., Jung, T.-P., and Cauwenberghs, decoding of complex right-hand motor imagery tasks. IEEE Trans. Biomed.
G. (2011). Dry and noncontact EEG sensors for mobile brain- Eng. 63, 4–14. doi: 10.1109/TBME.2015.2467312
computer interfaces. IEEE Trans. Neural Syst. Rehabil. Eng. 20, 228–235. Faress, A., and Chau, T. (2013). Towards a multimodal brain-computer interface:
doi: 10.1109/TNSRE.2011.2174652 combining fNIRS and ftcd measurements to enable higher classification
Chiarelli, A. M., Croce, P., Merla, A., and Zappasodi, F. (2018). Deep learning accuracy. Neuroimage 77, 186–194. doi: 10.1016/j.neuroimage.2013.03.028
for hybrid EEG-fNIRS brain-computer interface: application to motor imagery Farwell, L. A., and Donchin, E. (1988). Talking off the top of your head: toward
classification. J. Neural Eng. 15:036028. doi: 10.1088/1741-2552/aaaf82 a mental prosthesis utilizing event-related brain potentials. Electroencephalogr.
Choi, B., and Jo, S. (2013). A low-cost EEG system-based hybrid brain-computer Clin. Neurophysiol. 70, 510–523. doi: 10.1016/0013-4694(88)90149-6
interface for humanoid robot navigation and rECOGnition. PLoS ONE Farwell, L. A., Richardson, D. C., Richardson, G. M., and Furedy, J. J.
8:e74583. doi: 10.1371/journal.pone.0074583 (2014). Brain fingerprinting classification concealed information test detects
Clausen, J. (2011). Conceptual and ethical issues with brain-hardware interfaces. us navy military medical information with p300. Front. Neurosci. 8:410.
Curr. Opin. Psychiatry 24, 495–501. doi: 10.1097/YCO.0b013e32834bb8ca doi: 10.3389/fnins.2014.00410
Combaz, A., and Van Hulle, M. M. (2015). Simultaneous detection of p300 and Fazli, S., Mehnert, J., Steinbrink, J., Curio, G., Villringer, A., Müller, K.-R.,
steady-state visually evoked potentials for hybrid brain-computer interface. et al. (2012). Enhanced performance by a hybrid NIRS-EEG brain computer
PLoS ONE 10:e0121481. doi: 10.1371/journal.pone.0121481 interface. Neuroimage 59, 519–529. doi: 10.1016/j.neuroimage.2011.07.084
Congedo, M., Barachant, A., and Andreev, A. (2013). A new generation of Filiz, E., and Arslan, R. B. (2020). “Design and implementation of steady state
brain-computer interface based on riemannian geometry. arXiv preprint visual evoked potential based brain computer interface video game,” in 2020
arXiv:1310.8115. IEEE 20th Mediterranean Electrotechnical Conference (MELECON) (Palermo:
Congedo, M., Barachant, A., and Bhatia, R. (2017). Riemannian geometry for IEEE), 335–338. doi: 10.1109/MELECON48756.2020.9140710
EEG-based brain-computer interfaces; a primer and a review. Brain Comput. Friehs, G., Penn, R. D., Park, M. C., Goldman, M., Zerris, V. A., Hochberg, L. R.,
Interfaces 4, 155–174. doi: 10.1080/2326263X.2017.1297192 et al. (2006). Initial surgical experience with an intracortical microelectrode
Corsi, M.-C., Chavez, M., Schwartz, D., Hugueville, L., Khambhati, A. N., Bassett, array for brain-computer interface applications: 881. Neurosurgery 59:481.
D. S., et al. (2019). Integrating EEG and MEG signals to improve motor imagery doi: 10.1227/00006123-200608000-00119
classification in brain-computer interface. Int. J. Neural Syst. 29:1850014. Fukuma, R., Yanagisawa, T., Saitoh, Y., Hosomi, K., Kishima, H., Shimizu,
doi: 10.1142/S0129065718500144 T., et al. (2016). Real-time control of a neuroprosthetic hand by
da Silva, F. L. (2013). EEG and MEG: relevance to neuroscience. Neuron 80, magnetoencephalographic signals from paralysed patients. Sci. Rep. 6,
1112–1128. doi: 10.1016/j.neuron.2013.10.017 1–14. doi: 10.1038/srep21781
Darvishi, S., Gharabaghi, A., Boulay, C. B., Ridding, M. C., Abbott, D., and Galán, F., Nuttin, M., Lew, E., Ferrez, P. W., Vanacker, G., Philips, J., et al. (2008).
Baumert, M. (2017). Proprioceptive feedback facilitates motor imagery-related A brain-actuated wheelchair: asynchronous and non-invasive brain-computer
operant learning of sensorimotor β-band modulation. Front. Neurosci. 11:60. interfaces for continuous control of robots. Clin. Neurophysiol. 119, 2159–2169.
doi: 10.3389/fnins.2017.00060 doi: 10.1016/j.clinph.2008.06.001
De Marchis, C., Monteiro, T. S., Simon-Martinez, C., Conforto, S., and Gandhi, V., Prasad, G., Coyle, D., Behera, L., and McGinnity, T. M.
Gharabaghi, A. (2016). Multi-contact functional electrical stimulation for (2014). EEG-based mobile robot control through an adaptive brain-
hand opening: electrophysiologically driven identification of the optimal robot interface. IEEE Trans. Syst. Man Cybernet. Syst. 44, 1278–1285.
stimulation site. J. Neuroeng. Rehabil. 13, 1–9. doi: 10.1186/s12984-016- doi: 10.1109/TSMC.2014.2313317
0129-6 Gao, Z., Wang, X., Yang, Y., Mu, C., Cai, Q., Dang, W., et al. (2019).
de Negueruela, C., Broschart, M., Menon, C., and Millán, J. d. R. (2011). Brain- EEG-based spatio-temporal convolutional neural network for driver
computer interfaces for space applications. Pers. Ubiquit. Comput. 15, 527–537. fatigue evaluation. IEEE Trans. Neural Netw. Learn. Syst. 30, 2755–2763.
doi: 10.1007/s00779-010-0322-8 doi: 10.1109/TNNLS.2018.2886414

Frontiers in Systems Neuroscience | www.frontiersin.org 14 February 2021 | Volume 15 | Article 578875

Saha et al. Progress in Brain Computer Interface

Ge, S., Yang, Q., Wang, R., Lin, P., Gao, J., Leng, Y., et al. (2017). A brain-computer from structural connectivity. Proc. Natl. Acad. Sci. U.S.A. 106, 2035–2040.
interface based on a few-channel EEG-fNIRS bimodal system. IEEE Access 5, doi: 10.1073/pnas.0811168106
208–218. doi: 10.1109/ACCESS.2016.2637409 Honey, C. J., Thivierge, J.-P., and Sporns, O. (2010). Can structure
Gilbert, F. (2015). Self-estrangement and deep brain stimulation: predict function in the human brain? Neuroimage 52, 766–776.
ethical issues related to forced explantation. Neuroethics 8, 107–114. doi: 10.1016/j.neuroimage.2010.01.071
doi: 10.1007/s12152-014-9224-1 Hong, X., Lu, Z. K., Teh, I., Nasrallah, F. A., Teo, W. P., Ang, K. K., et al. (2017).
Gilja, V., Chestek, C. A., Diester, I., Henderson, J. M., Deisseroth, K., and Brain plasticity following MI-BCI training combined with tdcs in a randomized
Shenoy, K. V. (2011). Challenges and opportunities for next-generation trial in chronic subcortical stroke subjects: a preliminary study. Sci. Rep. 7, 1–12.
intracortically based neural prostheses. IEEE Trans. Biomed. Eng. 58, doi: 10.1038/s41598-017-08928-5
1891–1899. doi: 10.1109/TBME.2011.2107553 Horki, P., Solis-Escalante, T., Neuper, C., and Müller-Putz, G. (2011). Combined
Goering, S., and Yuste, R. (2016). On the necessity of ethical guidelines motor imagery and ssvep based BCI control of a 2 DOF artificial upper limb.
for novel neurotechnologies. Cell 167, 882–885. doi: 10.1016/j.cell.2016. Med. Biol. Eng. Comput. 49, 567–577. doi: 10.1007/s11517-011-0750-2
10.029 Huang, H., Xie, Q., Pan, J., He, Y., Wen, Z., Yu, R., et al. (2019). An EEG-
Gonçalves, S. I., De Munck, J. C., Pouwels, P. J., Schoonhoven, R., Kuijer, J. based brain computer interface for emotion rECOGnition and its application
P., Maurits, N. M., et al. (2006). Correlating the alpha rhythm to bold using in patients with disorder of consciousness. IEEE Trans. Affect. Comput.
simultaneous EEG/fMRI: inter-subject variability. Neuroimage 30, 203–213. doi: 10.1109/TAFFC.2019.2901456
doi: 10.1016/j.neuroimage.2005.09.062 Ienca, M., and Haselager, P. (2016). Hacking the brain: brain-computer interfacing
Grau, C., Ginhoux, R., Riera, A., Nguyen, T. L., Chauvat, H., Berg, M., et al. technology and the ethics of neurosecurity. Ethics Inform. Technol. 18, 117–129.
(2014). Conscious brain-to-brain communication in humans using non- doi: 10.1007/s10676-016-9398-9
invasive technologies. PLoS ONE 9:e105225. doi: 10.1371/journal.pone.0105225 Illes, J., and Bird, S. J. (2006). Neuroethics: a modern context for ethics in
Grienberger, C., and Konnerth, A. (2012). Imaging calcium in neurons. Neuron 73, neuroscience. Trends Neurosci. 29, 511–517. doi: 10.1016/j.tins.2006.07.002
862–885. doi: 10.1016/j.neuron.2012.02.011 Iturrate, I., Antelis, J. M., Kubler, A., and Minguez, J. (2009). A noninvasive
Grillner, S., Ip, N., Koch, C., Koroshetz, W., Okano, H., Polachek, M., et al. (2016). brain-actuated wheelchair based on a p300 neurophysiological
Worldwide initiatives to advance brain research. Nat. Neurosci. 19, 1118–1122. protocol and automated navigation. IEEE Trans. Robot. 25, 614–627.
doi: 10.1038/nn.4371 doi: 10.1109/TRO.2009.2020347
Grosse-Wentrup, M., Mattia, D., and Oweiss, K. (2011). Using brain-computer Jayaram, V., Alamgir, M., Altun, Y., Scholkopf, B., and Grosse-Wentrup, M. (2016).
interfaces to induce neural plasticity and restore function. J. Neural Eng. Transfer learning in brain-computer interfaces. IEEE Comput. Intell. Mag. 11,
8:025004. doi: 10.1088/1741-2560/8/2/025004 20–31. doi: 10.1109/MCI.2015.2501545
Guy, V., Soriani, M.-H., Bruno, M., Papadopoulo, T., Desnuelle, C., and Clerc, M. Jebari, K. (2013). Brain machine interface and human enhancement-an ethical
(2018). Brain computer interface with the p300 speller: usability for disabled review. Neuroethics 6, 617–625. doi: 10.1007/s12152-012-9176-2
people with amyotrophic lateral sclerosis. Ann. Phys. Rehabil. Med. 61, 5–11. Jensen, O., Bahramisharif, A., Oostenveld, R., Klanke, S., Hadjipapas, A.,
doi: 10.1016/j.rehab.2017.09.004 Okazaki, Y. O., et al. (2011). Using brain-computer interfaces and brain-state
Halder, S., Leinfelder, T., Schulz, S. M., and Kübler, A. (2019). Neural mechanisms dependent stimulation as tools in cognitive neuroscience. Front. Psychol. 2:100.
of training an auditory event-related potential task in a brain-computer doi: 10.3389/fpsyg.2011.00100
interface context. Hum. Brain Mapp. 40, 2399–2412. doi: 10.1002/hbm.24531 Jiang, L., Stocco, A., Losey, D. M., Abernethy, J. A., Prat, C. S., and Rao, R. P. (2019).
Hammer, E. M., Halder, S., Blankertz, B., Sannelli, C., Dickhaus, T., Kleih, S., et al. Brainnet: a multi-person brain-to-brain interface for direct collaboration
(2012). Psychological predictors of SMR-BCI performance. Biol. Psychol. 89, between brains. Sci. Rep. 9, 1–11. doi: 10.1038/s41598-019-41895-7
80–86. doi: 10.1016/j.biopsycho.2011.09.006 Jin, J., Chen, Z., Xu, R., Miao, Y., yu Wang, X., and Jung, T.-P. (2020). Developing a
Han, C.-H., Müller, K.-R., and Hwang, H.-J. (2020). Brain-switches for novel tactile p300 brain-computer interface with a cheeks-stim paradigm. IEEE
asynchronous brain-computer interfaces: a systematic review. Electronics 9:422. Trans. Biomed. Eng. 67, 2585–2593. doi: 10.1109/TBME.2020.2965178
doi: 10.3390/electronics9030422 Johnson, N., Carey, J., Edelman, B., Doud, A., Grande, A., Lakshminarayan,
Hasson, U., Nir, Y., Levy, I., Fuhrmann, G., and Malach, R. (2004). Intersubject K., et al. (2018). Combined rtms and virtual reality brain-computer
synchronization of cortical activity during natural vision. Science 303, interface training for motor recovery after stroke. J. Neural Eng. 15:016009.
1634–1640. doi: 10.1126/science.1089506 doi: 10.1088/1741-2552/aa8ce3
He, B., Yang, L., Wilke, C., and Yuan, H. (2011). Electrophysiological imaging Jones, S. R., and Sliva, D. D. (2020). Is alpha asymmetry a byproduct or
of brain activity and connectivity-challenges and opportunities. IEEE Trans. cause of spatial attention? New evidence alpha neurofeedback controls
Biomed. Eng. 58, 1918–1931. doi: 10.1109/TBME.2011.2139210 measures of spatial attention. Neuron 105, 404–406. doi: 10.1016/j.neuron.2019.
He, H., and Wu, D. (2019). Transfer learning for brain-computer interfaces: 12.033
a Euclidean space data alignment approach. IEEE Trans. Biomed. Eng. 67, Jotterand, F., and Giordano, J. (2011). Transcranial magnetic stimulation,
399–410. doi: 10.1109/TBME.2019.2913914 deep brain stimulation and personal identity: ethical questions, and
He, H., and Wu, D. (2020). Different set domain adaptation for brain-computer neuroethical approaches for medical practice. Int. Rev. Psychiatry 23, 476–485.
interfaces: a label alignment approach. IEEE Trans. Neural Syst. Rehabil. Eng.28, doi: 10.3109/09540261.2011.616189
1091–1108. doi: 10.1109/TNSRE.2020.2980299 Juliano, J. M., Spicer, R. P., Vourvopoulos, A., Lefebvre, S., Jann, K., Ard,
Hinrichs, H., Scholz, M., Baum, A. K., Kam, J. W., Knight, R. T., and Heinze, T., et al. (2020). Embodiment is related to better performance on a brain-
H.-J. (2020). Comparison between a wireless dry electrode EEG system with computer interface in immersive virtual reality: a pilot study. Sensors 20:1204.
a conventional wired wet electrode EEG system for clinical applications. Sci. doi: 10.3390/s20041204
Rep. 10, 1–14. doi: 10.1038/s41598-020-62154-0 Kaas, A. L., Goebel, R., Valente, G., and Sorger, B. (2019). Topographic
Hobson, E. V., Fazal, S., Shaw, P. J., and McDermott, C. J. (2017). somatosensory imagery for real-time fMRI brain-computer interfacing. Front.
“Anything that makes life’s journey better.” Exploring the use of Hum. Neurosci. 13:427. doi: 10.3389/fnhum.2019.00427
digital technology by people living with motor neurone disease. Kaiju, T., Yokota, M., Watanabe, K., Inoue, M., Ando, H., Takahashi, K., et al.
Amyotrophic Lateral Sclerosis Frontotemporal Degeneration 18, 378–387. (2017). High spatiotemporal resolution ECOG recording of somatosensory
doi: 10.1080/21678421.2017.1288253 evoked potentials with flexible micro-electrode arrays. Front. Neural Circ.
Hochberg, L. R. (2013). Intracortical brain-computer interfaces for the 11:20. doi: 10.3389/fncir.2017.00020
restoration of communication and mobility. Biophys. J. 104:376. Kamousi, B., Liu, Z., and He, B. (2005). Classification of motor imagery
doi: 10.1016/j.bpj.2012.11.2094 tasks for brain-computer interface applications by means of two equivalent
Honey, C. J., Sporns, O., Cammoun, L., Gigandet, X., Thiran, J.-P., Meuli, dipoles analysis. IEEE Trans. Neural Syst. Rehabil. Eng. 13, 166–171.
R., et al. (2009). Predicting human resting-state functional connectivity doi: 10.1109/TNSRE.2005.847386

Frontiers in Systems Neuroscience | www.frontiersin.org 15 February 2021 | Volume 15 | Article 578875

Saha et al. Progress in Brain Computer Interface

Kasahara, K., DaSalla, C. S., Honda, M., and Hanakawa, T. (2015). Lina, J.-M., Chowdhury, R., Lemay, E., Kobayashi, E., and Grova,
Neuroanatomical correlates of brain-computer interface performance. C. (2012). Wavelet-based localization of oscillatory sources from
Neuroimage 110, 95–100. doi: 10.1016/j.neuroimage.2015.01.055 magnetoencephalography data. IEEE Trans. Biomed. Eng. 61, 2350–2364.
Käthner, I., Wriessnegger, S. C., Müller-Putz, G. R., Kübler, A., and Halder, doi: 10.1109/TBME.2012.2189883
S. (2014). Effects of mental workload and fatigue on the p300, alpha Lotte, F., Bougrain, L., Cichocki, A., Clerc, M., Congedo, M., Rakotomamonjy,
and theta band power during operation of an ERP (p300) brain- A., et al. (2018). A review of classification algorithms for EEG-based
computer interface. Biol. Psychol. 102, 118–129. doi: 10.1016/j.biopsycho.2014. brain-computer interfaces: a 10 year update. J. Neural Eng. 15:031005.
07.014 doi: 10.1088/1741-2552/aab2f2
Kaufmann, T., Vögele, C., Sütterlin, S., Lukito, S., and Kübler, A. (2012). Effects Lotte, F., Congedo, M., Lécuyer, A., Lamarche, F., and Arnaldi, B. (2007). A review
of resting heart rate variability on performance in the p300 brain-computer of classification algorithms for EEG-based brain-computer interfaces. J. Neural
interface. Int. J. Psychophysiol. 83, 336–341. doi: 10.1016/j.ijpsycho.2011.11.018 Eng. 4:R1. doi: 10.1088/1741-2560/4/2/R01
Kauhanen, L., Nykopp, T., Lehtonen, J., Jylanki, P., Heikkonen, J., Lotte, F., and Guan, C. (2010). Regularizing common spatial patterns to improve
Rantanen, P., et al. (2006). EEG and MEG brain-computer interface for BCI designs: unified theory and new algorithms. IEEE Trans. Biomed. Eng. 58,
tetraplegic patients. IEEE Trans. Neural Syst. Rehabil. Eng. 14, 190–193. 355–362. doi: 10.1109/TBME.2010.2082539
doi: 10.1109/TNSRE.2006.875546 Lu, J., Mamun, K. A., and Chau, T. (2015). Pattern classification to optimize
Keefer, E. W., Botterman, B. R., Romero, M. I., Rossi, A. F., and Gross, G. the performance of transcranial doppler ultrasonography-based brain machine
W. (2008). Carbon nanotube coating improves neuronal recordings. Nat. interface. Pattern Recogn. Lett. 66, 135–143. doi: 10.1016/j.patrec.2015.07.020
Nanotechnol. 3, 434–439. doi: 10.1038/nnano.2008.174 Luu, T. P., He, Y., Brown, S., Nakagome, S., and Contreras-Vidal, J. L. (2016).
Khalaf, A., Sejdic, E., and Akcakaya, M. (2019). A novel motor Gait adaptation to visual kinematic perturbations using a real-time closed-loop
imagery hybrid brain computer interface using EEG and functional brain-computer interface to a virtual reality avatar. J. Neural Eng. 13:036006.
transcranial doppler ultrasound. J. Neurosci. Methods 313, 44–53. doi: 10.1088/1741-2560/13/3/036006
doi: 10.1016/j.jneumeth.2018.11.017 Lystad, R. P., and Pollard, H. (2009). Functional neuroimaging: a brief overview
Khodagholy, D., Doublet, T., Quilichini, P., Gurfinkel, M., Leleux, P., Ghestem, A., and feasibility for use in chiropractic research. J. Can. Chiropract. Assoc. 53:59.
et al. (2013). In vivo recordings of brain activity using organic transistors. Nat. Mahjoory, K., Nikulin, V. V., Botrel, L., Linkenkaer-Hansen, K., Fato, M. M.,
Commun. 4, 1–7. doi: 10.1038/ncomms2573 and Haufe, S. (2017). Consistency of EEG source localization and connectivity
Kleih, S. C., and Kübler, A. (2013). Empathy, motivation, and p300 BCI estimates. Neuroimage 152, 590–601. doi: 10.1016/j.neuroimage.2017.02.076
performance. Front. Hum. Neurosci. 7:642. doi: 10.3389/fnhum.2013.00642 Mak, J. N., and Wolpaw, J. R. (2009). Clinical applications of brain-computer
Kleinschmidt, A., Sterzer, P., and Rees, G. (2012). Variability of perceptual interfaces: current state and future prospects. IEEE Rev. Biomed. Eng. 2,
multistability: from brain state to individual trait. Philos. Trans. R. Soc. B Biol. 187–199. doi: 10.1109/RBME.2009.2035356
Sci. 367, 988–1000. doi: 10.1098/rstb.2011.0367 Mantini, D., Perrucci, M. G., Del Gratta, C., Romani, G. L., and Corbetta,
Krusienski, D. J., Grosse-Wentrup, M., Galán, F., Coyle, D., Miller, K. J., Forney, M. (2007). Electrophysiological signatures of resting state networks
E., et al. (2011). Critical issues in state-of-the-art brain-computer interface in the human brain. Proc. Natl. Acad. Sci. U.S.A. 104, 13170–13175.
signal processing. J. Neural Eng. 8:025002. doi: 10.1088/1741-2560/8/2/0 doi: 10.1073/pnas.0700668104
25002 Marathe, A. R., Lawhern, V. J., Wu, D., Slayback, D., and Lance, B. J. (2015).
Kruskal, P. B., Jiang, Z., Gao, T., and Lieber, C. M. (2015). Beyond the Improved neural signal classification in a rapid serial visual presentation task
patch clamp: nanotechnologies for intracellular recording. Neuron 86, 21–24. using active learning. IEEE Trans. Neural Syst. Rehabil. Eng. 24, 333–343.
doi: 10.1016/j.neuron.2015.01.004 doi: 10.1109/TNSRE.2015.2502323
Kwon, O.-Y., Lee, M.-H., Guan, C., and Lee, S.-W. (2019). Subject- Marchesotti, S., Bassolino, M., Serino, A., Bleuler, H., and Blanke, O. (2016).
independent brain-computer interfaces based on deep convolutional Quantifying the role of motor imagery in brain-machine interfaces. Sci. Rep.
neural networks. IEEE Trans. Neural Netw. Learn. Syst. 31, 3839–3852. 6:24076. doi: 10.1038/srep24076
doi: 10.1109/TNNLS.2019.2946869 Marchesotti, S., Martuzzi, R., Schurger, A., Blefari, M. L., del Millán, J. R.,
LaFleur, K., Cassady, K., Doud, A., Shades, K., Rogin, E., and He, B. Bleuler, H., et al. (2017). Cortical and subcortical mechanisms of brain-machine
(2013). Quadcopter control in three-dimensional space using a noninvasive interfaces. Hum. Brain Mapp. 38, 2971–2989. doi: 10.1002/hbm.23566
motor imagery-based brain-computer interface. J. Neural Eng. 10:046003. Marini, F., Lee, C., Wagner, J., Makeig, S., and Gola, M. (2019). A
doi: 10.1088/1741-2560/10/4/046003 comparative evaluation of signal quality between a research-grade and
Lajoie, G., Krouchev, N. I., Kalaska, J. F., Fairhall, A. L., and Fetz, E. E. (2017). a wireless dry-electrode mobile EEG system. J. Neural Eng. 16:054001.
Correlation-based model of artificially induced plasticity in motor cortex by doi: 10.1088/1741-2552/ab21f2
a bidirectional brain-computer interface. PLoS Comput. Biol. 13:e1005343. Martin, M. V., Cho, V., and Aversano, G. (2016). Detection of subconscious face
doi: 10.1371/journal.pcbi.1005343 rECOGnition using consumer-grade brain-computer interfaces. ACM Trans.
Leamy, D. J., Kocijan, J., Domijan, K., Duffin, J., Roche, R. A., Commins, S., Appl. Percept. 14, 1–20. doi: 10.1145/2955097
et al. (2014). An exploration of EEG features during recovery following Maslen, H., Pugh, J., and Savulescu, J. (2015). The ethics of deep brain
stroke-implications for BCI-mediated neurorehabilitation therapy. J. Neuroeng. stimulation for the treatment of anorexia nervosa. Neuroethics 8, 215–230.
Rehabil. 11:9. doi: 10.1186/1743-0003-11-9 doi: 10.1007/s12152-015-9240-9
Lebedev, M. A., and Nicolelis, M. A. (2017). Brain-machine interfaces: from basic Matthews, F., Pearlmutter, B. A., Wards, T. E., Soraghan, C., and Markham, C.
science to neuroprostheses and neurorehabilitation. Physiol. Rev. 97, 767–837. (2007). Hemodynamics for brain-computer interfaces. IEEE Signal Process.
doi: 10.1152/physrev.00027.2016 Mag. 25, 87–94. doi: 10.1109/MSP.2008.4408445
Lécuyer, A., Lotte, F., Reilly, R. B., Leeb, R., Hirose, M., and Slater, M. (2008). Brain- McCane, L. M., Heckman, S. M., McFarland, D. J., Townsend, G., Mak,
computer interfaces, virtual reality, and videogames. Computer 41, 66–72. J. N., Sellers, E. W., et al. (2015). P300-based brain-computer interface
doi: 10.1109/MC.2008.410 (BCI) event-related potentials (ERPs): people with amyotrophic lateral
Liao, L.-D., Lin, C.-T., McDowell, K., Wickenden, A. E., Gramann, K., sclerosis (ALS) vs. age-matched controls. Clin. Neurophysiol. 126, 2124–2131.
Jung, T.-P., et al. (2012). Biosensor technologies for augmented brain- doi: 10.1016/j.clinph.2015.01.013
computer interfaces in the next decades. Proc. IEEE 100, 1553–1566. McClinton, W., Garcia, S., and Andujar, M. (2019). “An immersive brain painting:
doi: 10.1109/JPROC.2012.2184829 the effects of brain painting in a virtual reality environment,” in International
Lin, C.-T., Chen, Y.-C., Huang, T.-Y., Chiu, T.-T., Ko, L.-W., Liang, S.- Conference on Human-Computer Interaction (Copenhagen: Springer), 436–
F., et al. (2008). Development of wireless brain computer interface with 445. doi: 10.1007/978-3-030-22419-6_31
embedded multitask scheduling and its application on real-time driver’s McCullagh, P., Lightbody, G., Zygierewicz, J., and Kernohan, W. G. (2014). Ethical
drowsiness detection and warning. IEEE Trans. Biomed. Eng. 55, 1582–1591. challenges associated with the development and deployment of brain computer
doi: 10.1109/TBME.2008.918566 interface technology. Neuroethics 7, 109–122. doi: 10.1007/s12152-013-9188-6

Frontiers in Systems Neuroscience | www.frontiersin.org 16 February 2021 | Volume 15 | Article 578875

Saha et al. Progress in Brain Computer Interface

McFarland, D. J., Anderson, C. W., Muller, K.-R., Schlogl, A., and Krusienski, Nijboer, F., Birbaumer, N., and Kubler, A. (2010). The influence of psychological
D. J. (2006). Bci meeting 2005-workshop on BCI signal processing: feature state and motivation on brain-computer interface performance in patients
extraction and translation. IEEE Trans. Neural Syst. Rehabil. Eng. 14, 135–138. with amyotrophic lateral sclerosis-A longitudinal study. Front. Neurosci. 4:55.
doi: 10.1109/TNSRE.2006.875637 doi: 10.3389/fnins.2010.00055
McMillan, G. R., Calhoun, G., Middendorf, M., Schnurer, J., Ingle, D., and Nasman, Nijholt, A., and Poel, M. (2016). “Multi-brain BCI: Characteristics and social
V. (1995). “Direct brain interface utilizing self-regulation of steady-state visual interactions,” in International Conference on Augmented Cognition (Las Vegas,
evoked response (SSVER),” in Proc. RESNA 95 Annual Conf. (Vancouver, BC), NV: Springer), 79–90. doi: 10.1007/978-3-319-39955-3_8
693–695. Norton, J. J., Lee, D. S., Lee, J. W., Lee, W., Kwon, O., Won, P., et al.
Megan, T. D., Cohen, J. D., Lee, R. F., Norman, K. A., and Turk-Browne, N. B. (2015). Soft, curved electrode systems capable of integration on the auricle
(2015). Closed-loop training of attention with real-time brain imaging. Nat. as a persistent brain-computer interface. Proc. Natl. Acad. Sci. U.S.A. 112,
Neurosci. 18, 470–475. doi: 10.1038/nn.3940 3920–3925. doi: 10.1073/pnas.1424875112
Mellinger, J., Schalk, G., Braun, C., Preissl, H., Rosenstiel, W., Birbaumer, N., Orsborn, A. L., Moorman, H. G., Overduin, S. A., Shanechi, M. M., Dimitrov,
et al. (2007). An MEG-based brain-computer interface (BCI). Neuroimage 36, D. F., and Carmena, J. M. (2014). Closed-loop decoder adaptation shapes
581–593. doi: 10.1016/j.neuroimage.2007.03.019 neural plasticity for skillful neuroprosthetic control. Neuron 82, 1380–1393.
Menon, C., De Negueruela, C., Millán, J. d. R., Tonet, O., Carpi, F., Broschart, M., doi: 10.1016/j.neuron.2014.04.048
et al. (2009). Prospects of brain-machine interfaces for space system control. Ortiz, F., González, J., Montes, A., González, N., and Peña, A. (2016).
Acta Astronaut. 64, 448–456. doi: 10.1016/j.actaastro.2008.09.008 Induction of emotional states in people with disabilities through film
Mestais, C. S., Charvet, G., Sauter-Starace, F., Foerster, M., Ratel, D., and Benabid, clips using brain computer interfaces. IEEE Latin Am. Trans. 14, 563–568.
A. L. (2014). Wimagine: wireless 64-channel ECOG recording implant for long doi: 10.1109/TLA.2016.7437193
term clinical applications. IEEE Trans. Neural Syst. Rehabil. Eng. 23, 10–21. Oxley, T. J., Opie, N. L., John, S. E., Rind, G. S., Ronayne, S. M., Burkitt, A.
doi: 10.1109/TNSRE.2014.2333541 N., et al. (2017). “A minimally invasive endovascular stent-electrode array for
Milan, J. D. R., and Carmena, J. M. (2010). Invasive or noninvasive: understanding chronic recordings of cortical neural activity,” in Brain-Computer Interface
brain-machine interface technology [conversations in BME]. IEEE Eng. Med. Research, eds C. Guger, B. Allison, and M. Lebedev (Cham: Springer), 55–63.
Biol. Mag. 29, 16–22. doi: 10.1109/MEMB.2009.935475 doi: 10.1007/978-3-319-64373-1_6
Millan, J. R., Renkens, F., Mourino, J., and Gerstner, W. (2004). Noninvasive brain- Oxley, T. J., Opie, N. L., John, S. E., Rind, G. S., Ronayne, S. M., Wheeler, T.
actuated control of a mobile robot by human EEG. IEEE Trans. Biomed. Eng. L., et al. (2016). Minimally invasive endovascular stent-electrode array for
51, 1026–1033. doi: 10.1109/TBME.2004.827086 high-fidelity, chronic recordings of cortical neural activity. Nat. Biotechnol. 34,
Min, B.-K., Hämäläinen, M. S., and Pantazis, D. (2020). New cognitive 320–327. doi: 10.1038/nbt.3428
neurotechnology facilitates studies of cortical-subcortical interactions. Trends Oxley, T. J., Yoo, P. E., Rind, G. S., Ronayne, S. M., Lee, C. S., Bird, C., et al.
Biotechnol. 38, 952–962. doi: 10.1016/j.tibtech.2020.03.003 (2020). Motor neuroprosthesis implanted with neurointerventional surgery
Min, B.-K., Marzelli, M. J., and Yoo, S.-S. (2010). Neuroimaging-based improves capacity for activities of daily living tasks in severe paralysis: first
approaches in the brain-computer interface. Trends Biotechnol. 28, 552–560. in-human experience. J. NeuroIntervent. Surg. doi: 10.1136/neurintsurg-2020-
doi: 10.1016/j.tibtech.2010.08.002 016862
Mota, A. R., Duarte, L., Rodrigues, D., Martins, A., Machado, A., Vaz, F., et al. Pahwa, M., Kusner, M., Hacker, C. D., Bundy, D. T., Weinberger, K. Q., and
(2013). Development of a quasi-dry electrode for EEG recording. Sensors Actuat Leuthardt, E. C. (2015). Optimizing the detection of wakeful and sleep-like
A Phys. 199, 310–317. doi: 10.1016/j.sna.2013.06.013 states for future electrocorticographic brain computer interface applications.
Mudgal, S. K., Sharma, S. S. K., Chaturvedi, I., and Sharma, A. (2020). PLoS ONE 10:e0142947. doi: 10.1371/journal.pone.0142947
Brain computer interface advancement in neurosciences: applications Pais-Vieira, M., Lebedev, M., Kunicki, C., Wang, J., and Nicolelis, M. A. (2013). A
and issues. Interdiscipl. Neurosurg. 20:100694. doi: 10.1016/j.inat.2020.1 brain-to-brain interface for real-time sharing of sensorimotor information. Sci.
00694 Rep. 3:1319. doi: 10.1038/srep01319
Müller, J. L., Röder, C. H., Schuierer, G., and Klein, H. (2002). Motor- Pandarinath, C., Nuyujukian, P., Blabe, C. H., Sorice, B. L., Saab, J., Willett,
induced brain activation in cortical, subcortical and cerebellar regions F. R., et al. (2017). High performance communication by people with
in schizophrenic inpatients. A whole brain fMRI fingertapping paralysis using an intracortical brain-computer interface. eLife 6:e18554.
study. Prog. Neuropsychopharmacol. Biol. Psychiatry 26, 421–426. doi: 10.7554/eLife.18554
doi: 10.1016/S0278-5846(01)00271-8 Park, W., Kwon, G. H., Kim, Y.-H., Lee, J.-H., and Kim, L. (2016). EEG response
Muller-Putz, G. R., Scherer, R., Neuper, C., and Pfurtscheller, G. (2006). Steady- varies with lesion location in patients with chronic stroke. J. Neuroeng. Rehabil.
state somatosensory evoked potentials: suitable brain signals for brain- 13, 1–10. doi: 10.1186/s12984-016-0120-2
computer interfaces? IEEE Trans. Neural Syst. Rehabil. Eng. 14, 30–37. Perdikis, S., Tonin, L., and Millan, J., d. R. (2017). Brain racers. IEEE Spectrum 54,
doi: 10.1109/TNSRE.2005.863842 44–51. doi: 10.1109/MSPEC.2017.8012239
Murovec, N., Heilinger, A., Xu, R., Ortner, R., Spataro, R., La Bella, V., et al. Petrov, Y., Nador, J., Hughes, C., Tran, S., Yavuzcetin, O., and Sridhar, S.
(2020). Effects of a vibro-tactile p300 based brain-computer interface on the (2014). Ultra-dense EEG sampling results in two-fold increase of functional
coma recovery scale-revised in patients with disorders of consciousness. Front. brain information. Neuroimage 90, 140–145. doi: 10.1016/j.neuroimage.2013.
Neurosci. 14:294. doi: 10.3389/fnins.2020.00294 12.041
Nagel, S., and Spüler, M. (2019). World’s fastest brain-computer interface: Pfurtscheller, G., Guger, C., Müller, G., Krausz, G., and Neuper, C. (2000). Brain
combining EEG2code with deep learning. PLoS ONE 14:e0221909. oscillations control hand orthosis in a tetraplegic. Neurosci. Lett. 292, 211–214.
doi: 10.1371/journal.pone.0221909 doi: 10.1016/S0304-3940(00)01471-3
Nakazawa, E., Yamamoto, K., Tachibana, K., Toda, S., Takimoto, Y., and Pfurtscheller, G., Solis-Escalante, T., Ortner, R., Linortner, P., and Muller-
Akabayashi, A. (2016). Ethics of decoded neurofeedback in clinical Putz, G. R. (2010). Self-paced operation of an ssvep-based orthosis with
research, treatment, and moral enhancement. AJOB Neurosci. 7, 110–117. and without an imagery-based “brain switch:” a feasibility study towards
doi: 10.1080/21507740.2016.1172134 a hybrid BCI. IEEE Trans. Neural Syst. Rehabil. Eng. 18, 409–414.
Naseer, N., and Hong, K.-S. (2015). fNIRS-based brain-computer interfaces: a doi: 10.1109/TNSRE.2010.2040837
review. Front. Hum. Neurosci. 9:3. doi: 10.3389/fnhum.2015.00003 Piastra, M. C., Nüßing, A., Vorwerk, J., Clerc, M., Engwer, C., and Wolters,
Nicolas-Alonso, L. F., and Gomez-Gil, J. (2012). Brain computer interfaces, a C. H. (2020). A comprehensive study on electroencephalography and
review. Sensors 12, 1211–1279. doi: 10.3390/s120201211 magnetoencephalography sensitivity to cortical and subcortical sources. Hum.
Nicolelis, M. A., and Lebedev, M. A. (2009). Principles of neural ensemble Brain Mapp. doi: 10.1002/hbm.25272
physiology underlying the operation of brain-machine interfaces. Nat. Rev. Picard, R. W. (2000). Affective Computing. Cambridge, MA; London: MIT Press.
Neurosci. 10, 530–540. doi: 10.1038/nrn2653 doi: 10.7551/mitpress/1140.001.0001

Frontiers in Systems Neuroscience | www.frontiersin.org 17 February 2021 | Volume 15 | Article 578875

Saha et al. Progress in Brain Computer Interface

Piho, L., and Tjahjadi, T. (2018). A mutual information based adaptive windowing Samek, W., Meinecke, F. C., and Müller, K.-R. (2013). Transferring subspaces
of informative EEG for emotion recognition. IEEE Trans. Affect. Comput. 11, between subjects in brain-computer interfacing. IEEE Trans. Biomed. Eng. 60,
722–735. doi: 10.1109/TAFFC.2018.2840973 2289–2298. doi: 10.1109/TBME.2013.2253608
Putze, F., Vourvopoulos, A., Lécuyer, A., Krusienski, D., i Badia, S. B., Mullen, T., Sand, D., Peremen, Z., Haor, D., Arkadir, D., Bergman, H., and Geva, A.
et al. (2020). Brain-computer interfaces and augmented/virtual reality. Front. (2017). Optimization of deep brain stimulation in stn among patients with
Hum. Neurosci. 14:144. doi: 10.3389/fnhum.2020.00144 Parkinson’s disease using a novel EEG-based tool. Brain Stimulat. 10:510.
Qin, L., Ding, L., and He, B. (2004). Motor imagery classification by means of doi: 10.1016/j.brs.2017.01.490
source analysis for brain-computer interface applications. J. Neural Eng. 1:135. Sannelli, C., Vidaurre, C., Müller, K.-R., and Blankertz, B. (2016). Ensembles of
doi: 10.1088/1741-2560/1/3/002 adaptive spatial filters increase BCI performance: an online evaluation. J. Neural
Ramoser, H., Muller-Gerking, J., and Pfurtscheller, G. (2000). Optimal spatial Eng. 13:046003. doi: 10.1088/1741-2560/13/4/046003
filtering of single trial EEG during imagined hand movement. IEEE Trans. Sauter-Starce, F., Ratel, D., Cretallaz, C., Foerster, M., Lambert, A., Gaude,
Rehabil. Eng. 8, 441–446. doi: 10.1109/86.895946 C., et al. (2019). Long-term sheep implantation of wimagine R , a
Ramos-Murguialday, A., Broetz, D., Rea, M., Läer, L., Yilmaz, Ö., Brasil, F. L., et al. wireless 64-channels electrocorticogram recorder. Front. Neurosci. 13:847.
(2013). Brain-machine interface in chronic stroke rehabilitation: a controlled doi: 10.3389/fnins.2019.00847
study. Ann. Neurol. 74, 100–108. doi: 10.1002/ana.23879 Schalk, G. (2010). Can electrocorticography (ECOG) support robust
Rao, R. P., Stocco, A., Bryan, M., Sarma, D., Youngquist, T. M., Wu, J., et al. and powerful brain-computer interfaces? Front. Neuroeng. 3:9.
(2014). A direct brain-to-brain interface in humans. PLoS ONE 9:e111332. doi: 10.3389/fneng.2010.00009
doi: 10.1371/journal.pone.0111332 Schaworonkow, N., Triesch, J., Ziemann, U., and Zrenner, C. (2019). EEG-
Rashid, M., Sulaiman, N., P. P., Abdul Majeed, A., Musa, R. M., Ab Nasir, A. F., triggered tms reveals stronger brain state-dependent modulation of motor
et al. (2020). Current status, challenges, and possible solutions of EEG-based evoked potentials at weaker stimulation intensities. Brain Stimulat. 12, 110–118.
brain-computer interface: a comprehensive review. Front. Neurorobot. 14:25. doi: 10.1016/j.brs.2018.09.009
doi: 10.3389/fnbot.2020.00025 Schicktanz, S., Amelung, T., and Rieger, J. W. (2015). Qualitative assessment
Ray, A. M., Sitaram, R., Rana, M., Pasqualotto, E., Buyukturkoglu, K., Guan, C., of patients’ attitudes and expectations toward BCIs and implications
et al. (2015). A subject-independent pattern-based brain-computer interface. for future technology development. Front. Syst. Neurosci. 9:64.
Front. Behav. Neurosci. 9:269. doi: 10.3389/fnbeh.2015.00269 doi: 10.3389/fnsys.2015.00064
Resendez, S. L., Jennings, J. H., Ung, R. L., Namboodiri, V. M. K., Zhou, Z. C., Scholkmann, F., Kleiser, S., Metz, A. J., Zimmermann, R., Pavia, J. M.,
Otis, J. M., et al. (2016). Visualization of cortical, subcortical and deep brain Wolf, U., et al. (2014). A review on continuous wave functional
neural circuit dynamics during naturalistic mammalian behavior with head- near-infrared spectroscopy and imaging instrumentation and
mounted microscopes and chronically implanted lenses. Nat. Protoc. 11:566. methodology. Neuroimage 85, 6–27. doi: 10.1016/j.neuroimage.2013.
doi: 10.1038/nprot.2016.021 05.004
Rezazadeh Sereshkeh, A., Yousefi, R., Wong, A. T., Rudzicz, F., and Chau, Schreuder, M., Blankertz, B., and Tangermann, M. (2010). A new auditory
T. (2019). Development of a ternary hybrid fNIRS-EEG brain-computer multi-class brain-computer interface paradigm: spatial hearing as an
interface based on imagined speech. Brain Comput. Interfaces 6, 128–140. informative cue. PLoS ONE 5:e9813. doi: 10.1371/journal.pone.00
doi: 10.1080/2326263X.2019.1698928 09813
Romero-Laiseca, M. A., Delisle-Rodriguez, D., Cardoso, V., Gurve, D., Loterio, Schwartz, A. B., Cui, X. T., Weber, D. J., and Moran, D. W. (2006). Brain-controlled
F., Nascimento, J. H. P., et al. (2020). A low-cost lower-limb brain- interfaces: movement restoration with neural prosthetics. Neuron 52, 205–220.
machine interface triggered by pedaling motor imagery for post-stroke doi: 10.1016/j.neuron.2006.09.019
patients rehabilitation. IEEE Trans. Neural Syst. Rehabil. Eng. 28, 988–996. Sekhavat, Y. A. (2020). Battle of minds: a new interaction approach in BCI games
doi: 10.1109/TNSRE.2020.2974056 through competitive reinforcement. Multimedia Tools Appl. 79, 3449–3464.
Rosenboom, D. (2014). Active imaginative listening–A neuromusical critique. doi: 10.1007/s11042-019-07963-w
Front. Neurosci. 8:251. doi: 10.3389/fnins.2014.00251 Seo, D., Neely, R. M., Shen, K., Singhal, U., Alon, E., Rabaey, J. M.,
Rosenfeld, J. V., and Wong, Y. T. (2017). Neurobionics and the brain-computer et al. (2016). Wireless recording in the peripheral nervous system with
interface: current applications and future horizons. Med. J. Australia 206, ultrasonic neural dust. Neuron 91, 529–539. doi: 10.1016/j.neuron.2016.
363–368. doi: 10.5694/mja16.01011 06.034
Roth, B. L. (2016). Dreadds for neuroscientists. Neuron 89, 683–694. Shahriari, Y., Vaughan, T. M., McCane, L., Allison, B. Z., Wolpaw, J. R., and
doi: 10.1016/j.neuron.2016.01.040 Krusienski, D. J. (2019). An exploration of BCI performance variations in
Saeedi, S., Chavarriaga, R., and Millán, J. d. R. (2016). Long-term stable people with amyotrophic lateral sclerosis using longitudinal EEG data. J. Neural
control of motor-imagery BCI by a locked-in user through adaptive Eng. 16:056031. doi: 10.1088/1741-2552/ab22ea
assistance. IEEE Trans. Neural Syst. Rehabil. Eng. 25, 380–391. Shibata, K., Watanabe, T., Sasaki, Y., and Kawato, M. (2011). Perceptual learning
doi: 10.1109/TNSRE.2016.2645681 incepted by decoded fMRI neurofeedback without stimulus presentation.
Saha, S., Ahmed, K. I., Mostafa, R., Khandoker, A. H., and Hadjileontiadis, Science 334, 1413–1415. doi: 10.1126/science.1212003
L. (2017). Enhanced inter-subject brain computer interface with Shih, J. J., Krusienski, D. J., and Wolpaw, J. R. (2012). “Brain-computer
associative sensorimotor oscillations. Healthcare Technol. Lett. 4, 39–43. interfaces in medicine,” in Mayo Clinic Proceedings (Elsevier), 268–279.
doi: 10.1049/htl.2016.0073 doi: 10.1016/j.mayocp.2011.12.008
Saha, S., Ahmed, K. I. U., Mostafa, R., Hadjileontiadis, L., and Khandoker, A. Singh, A. K., Wang, Y.-K., King, J.-T., and Lin, C.-T. (2020). Extended
(2018). Evidence of variabilities in EEG dynamics during motor imagery-based interaction with a BCI video game changes resting-state brain activity.
multiclass brain-computer interface. IEEE Trans. Neural Syst. Rehabil. Eng. 26, IEEE Trans. Cogn. Dev. Syst. 12, 809–823. doi: 10.1109/TCDS.2020.29
371–382. doi: 10.1109/TNSRE.2017.2778178 85102
Saha, S., and Baumert, M. (2020). Intra-and inter-subject variability in EEG-based Sitaram, R., Ros, T., Stoeckel, L., Haller, S., Scharnowski, F., Lewis-Peacock, J.,
sensorimotor brain computer interface: a review. Front. Comput. Neurosci. et al. (2017). Closed-loop brain training: the science of neurofeedback. Nat. Rev.
13:87. doi: 10.3389/fncom.2019.00087 Neurosci. 18, 86–100. doi: 10.1038/nrn.2016.164
Saha, S., Hossain, M., Ahmed, K. I. U., Mostafa, R., Hadjileontiadis, Sitaram, R., Weiskopf, N., Caria, A., Veit, R., Erb, M., and Birbaumer, N.
L. J., Khandoker, A. H., et al. (2019a). Wavelet entropy-based inter- (2007). fMRI brain-computer interfaces. IEEE Signal Process. Mag. 25, 95–106.
subject associative cortical source localization for sensorimotor BCI. Front. doi: 10.1109/MSP.2008.4408446
Neuroinform. 13:47. doi: 10.3389/fninf.2019.00047 Škola, F., Tinková, S., and Liarokapis, F. (2019). Progressive training for
Saha, S., Mamun, K. A., Ahmed, K., Mostafa, R., Naik, G. R., Khandoker, A., et al. motor imagery brain-computer interfaces using gamification and virtual
(2019b). Progress in brain computer interfaces: challenges and trends. arXiv reality embodiment. Front. Hum. Neurosci. 13:329. doi: 10.3389/fnhum.2019.
preprint arXiv:1901.03442. 00329

Frontiers in Systems Neuroscience | www.frontiersin.org 18 February 2021 | Volume 15 | Article 578875

Saha et al. Progress in Brain Computer Interface

Song, M., and Kim, J. (2019). A paradigm to enhance motor imagery using Vidaurre, C., and Blankertz, B. (2010). Towards a cure for BCI illiteracy. Brain
rubber hand illusion induced by visuo-tactile stimulus. IEEE Trans. Neural Syst. Topogr. 23, 194–198. doi: 10.1007/s10548-009-0121-6
Rehabil. Eng. 27, 477–486. doi: 10.1109/TNSRE.2019.2895029 Vourvopoulos, A., Pardo, O. M., Lefebvre, S., Neureither, M., Saldana, D., Jahng,
Song, T., Zheng, W., Song, P., and Cui, Z. (2018). EEG emotion recognition using E., et al. (2019). Effects of a brain-computer interface with virtual reality (VR)
dynamical graph convolutional neural networks. IEEE Trans. Affect. Comput. neurofeedback: a pilot study in chronic stroke patients. Front. Hum. Neurosci.
11, 532–541. doi: 10.1109/TAFFC.2018.2817622 13:210. doi: 10.3389/fnhum.2019.00210
Spataro, R., Chella, A., Allison, B., Giardina, M., Sorbello, R., Tramonte, S., Vyas, S., Even-Chen, N., Stavisky, S. D., Ryu, S. I., Nuyujukian, P., and Shenoy,
et al. (2017). Reaching and grasping a glass of water by locked-in als patients K. V. (2018). Neural population dynamics underlying motor learning transfer.
through a BCI-controlled humanoid robot. Front. Hum. Neurosci. 11:68. Neuron 97, 1177–1186. doi: 10.1016/j.neuron.2018.01.040
doi: 10.3389/fnhum.2017.00068 Wang, H., Chang, W., and Zhang, C. (2016a). Functional brain network and
Sporns, O. (2013). Structure and function of complex brain networks. Dialog. Clin. multichannel analysis for the p300-based brain computer interface system of
Neurosci. 15:247. doi: 10.31887/DCNS.2013.15.3/osporns lying detection. Expert Syst. Appl. 53, 117–128. doi: 10.1016/j.eswa.2016.01.024
Steinert, S., and Friedrich, O. (2020). Wired emotions: ethical issues Wang, H., McIntosh, A. R., Kovacevic, N., Karachalios, M., and Protzner, A.
of affective brain-computer interfaces. Sci. Eng. Ethics 26, 351–367. B. (2016b). Age-related multiscale changes in brain signal variability in
doi: 10.1007/s11948-019-00087-2 pre-task versus post-task resting-state EEG. J. Cogn. Neurosci. 28, 971–984.
Sternson, S. M., and Roth, B. L. (2014). Chemogenetic tools to doi: 10.1162/jocn_a_00947
interrogate brain functions. Annu. Rev. Neurosci. 37, 387–407. Wang, W., Collinger, J. L., Perez, M. A., Tyler-Kabara, E. C., Cohen, L. G.,
doi: 10.1146/annurev-neuro-071013-014048 Birbaumer, N., et al. (2010). Neural interface technology for rehabilitation:
Stocco, A., Prat, C. S., Losey, D. M., Cronin, J. A., Wu, J., Abernethy, J. A., exploiting and promoting neuroplasticity. Phys. Med. Rehabil. Clin. 21,
et al. (2015). Playing 20 questions with the mind: collaborative problem 157–178. doi: 10.1016/j.pmr.2009.07.003
solving by humans using a brain-to-brain interface. PLoS ONE 10:e0137303. Wardman, D. L., Gandevia, S. C., and Colebatch, J. G. (2014). Cerebral,
doi: 10.1371/journal.pone.0137303 subcortical, and cerebellar activation evoked by selective stimulation of
Stoica, A. (2012). “Multimind: multi-brain signal fusion to exceed the power of muscle and cutaneous afferents: an fMRI study. Physiol. Rep. 2:e00270.
a single brain,” in 2012 Third International Conference on Emerging Security doi: 10.1002/phy2.270
Technologies (Lisbon: IEEE), 94–98. doi: 10.1109/EST.2012.47 Waytowich, N. R., Lawhern, V. J., Bohannon, A. W., Ball, K. R., and
Summerer, L., Izzo, D., and Rossini, L. (2009). Brain-machine interfaces for space Lance, B. J. (2016). Spectral transfer learning using information geometry
applications-research, technological development, and opportunities. Int. Rev. for a user-independent brain-computer interface. Front. Neurosci. 10:430.
Neurobiol. 86, 213–223. doi: 10.1016/S0074-7742(09)86016-9 doi: 10.3389/fnins.2016.00430
Taylor, D. M., Tillery, S. I. H., and Schwartz, A. B. (2002). Direct Wei, C.-S., Wang, Y.-T., Lin, C.-T., and Jung, T.-P. (2018). Toward
cortical control of 3d neuroprosthetic devices. Science 296, 1829–1832. drowsiness detection using non-hair-bearing EEG-based brain-
doi: 10.1126/science.1070291 computer interfaces. IEEE Trans. Neural Syst. Rehabil. Eng. 26, 400–406.
Tidoni, E., Abu-Alqumsan, M., Leonardis, D., Kapeller, C., Fusco, G., Guger, C., doi: 10.1109/TNSRE.2018.2790359
et al. (2016). Local and remote cooperation with virtual and robotic agents: a Wronkiewicz, M., Larson, E., and Lee, A. K. (2015). Leveraging anatomical
p300 BCI study in healthy and people living with spinal cord injury. IEEE Trans. information to improve transfer learning in brain-computer interfaces. J.
Neural Syst. Rehabil. Eng. 25, 1622–1632. doi: 10.1109/TNSRE.2016.2626391 Neural Eng. 12:046027. doi: 10.1088/1741-2560/12/4/046027
Tonin, L., Bauer, F. C., and Millán, J. d. R. (2019). The role of the control framework Wu, D., King, J.-T., Chuang, C.-H., Lin, C.-T., and Jung, T.-P. (2017a). Spatial
for continuous teleoperation of a brain-machine interface-driven mobile robot. filtering for EEG-based regression problems in brain-computer interface
IEEE Trans. Robot. 36, 78–91. doi: 10.1109/TRO.2019.2943072 (BCI). IEEE Trans. Fuzzy Syst. 26, 771–781. doi: 10.1109/TFUZZ.2017.26
Tonin, L., and Millán, J. d. R. (2020). Noninvasive brain-machine 88423
interfaces for robotic devices. Annu. Rev. Control Robot. Auton. Syst. 4. Wu, D., Lance, B. J., Lawhern, V. J., Gordon, S., Jung, T.-P., and Lin, C.-
doi: 10.1146/annurev-control-012720-093904 T. (2017b). EEG-based user reaction time estimation using riemannian
Toriyama, H., Ushiba, J., and Ushiyama, J. (2018). Subjective vividness of geometry features. IEEE Trans. Neural Syst. Rehabil. Eng. 25, 2157–2168.
kinesthetic motor imagery is associated with the similarity in magnitude of doi: 10.1109/TNSRE.2017.2699784
sensorimotor event-related desynchronization between motor execution and Wu, D., Lawhern, V. J., Hairston, W. D., and Lance, B. J. (2016). Switching EEG
motor imagery. Front. Hum. Neurosci. 12:295. doi: 10.3389/fnhum.2018.00295 headsets made easy: reducing offline calibration effort using active weighted
Tremmel, C., Herff, C., Sato, T., Rechowicz, K., Yamani, Y., and Krusienski, adaptation regularization. IEEE Trans. Neural Syst. Rehabil. Eng. 24, 1125–1137.
D. J. (2019). Estimating cognitive workload in an interactive virtual doi: 10.1109/TNSRE.2016.2544108
reality environment using EEG. Front. Hum. Neurosci. 13:401. Wu, D., Xu, Y., and Lu, B. (2020). Transfer learning for EEG-based brain-computer
doi: 10.3389/fnhum.2019.00401 interfaces: a review of progress made since 2016. IEEE Trans. Cogn. Dev. Syst.
Vaadia, E., and Birbaumer, N. (2009). Grand challenges of brain doi: 10.1109/TCDS.2020.3007453
computer interfaces in the years to come. Front. Neurosci. 3:15. Yang, W., Miller, J. K., Carrillo-Reid, L., Pnevmatikakis, E., Paninski, L., Yuste, R.,
doi: 10.3389/neuro.01.015.2009 et al. (2016). Simultaneous multi-plane imaging of neural circuits. Neuron 89,
van Vliet, M., Robben, A., Chumerin, N., Manyakov, N. V., Combaz, A., 269–284. doi: 10.1016/j.neuron.2015.12.012
and Van Hulle, M. M. (2012). “Designing a brain-computer interface Yin, E., Zeyl, T., Saab, R., Chau, T., Hu, D., and Zhou, Z. (2015). A hybrid
controlled video-game using consumer grade EEG hardware,” in 2012 ISSNIP brain-computer interface based on the fusion of p300 and SSVEP scores. IEEE
Biosignals and Biorobotics Conference: Biosignals and Robotics for Better Trans. Neural Syst. Rehabil. Eng. 23, 693–701. doi: 10.1109/TNSRE.2015.24
and Safer Living (BRC) (Manaus: IEEE), 1–6. doi: 10.1109/BRC.2012.62 03270
22186 Yin, M., Borton, D. A., Komar, J., Agha, N., Lu, Y., Li, H., et al.
Vansteensel, M. J., Pels, E. G., Bleichner, M. G., Branco, M. P., Denison, (2014). Wireless neurosensor for full-spectrum electrophysiology recordings
T., Freudenburg, Z. V., et al. (2016). Fully implanted brain-computer during free behavior. Neuron 84, 1170–1182. doi: 10.1016/j.neuron.2014.
interface in a locked-in patient with als. N. Engl. J. Med. 375, 2060–2066. 11.010
doi: 10.1056/NEJMoa1608085 Yu, Y., Liu, Y., Jiang, J., Yin, E., Zhou, Z., and Hu, D. (2018). An asynchronous
Vasilyev, A., Liburkina, S., Yakovlev, L., Perepelkina, O., and Kaplan, A. control paradigm based on sequential motor imagery and its application in
(2017). Assessing motor imagery in brain-computer interface training: wheelchair navigation. IEEE Trans. Neural Syst. Rehabil. Eng. 26, 2367–2375.
psychological and neurophysiological correlates. Neuropsychologia 97, 56–65. doi: 10.1109/TNSRE.2018.2881215
doi: 10.1016/j.neuropsychologia.2017.02.005 Yuan, W., and Li, Z. (2018). Brain teleoperation control of a nonholonomic mobile
Vidal, J. J. (1973). Toward direct brain-computer communication. Annu. Rev. robot using quadrupole potential function. IEEE Trans. Cogn. Dev. Syst. 11,
Biophys. Bioeng. 2, 157–180. doi: 10.1146/annurev.bb.02.060173.001105 527–538. doi: 10.1109/TCDS.2018.2869903

Frontiers in Systems Neuroscience | www.frontiersin.org 19 February 2021 | Volume 15 | Article 578875

Saha et al. Progress in Brain Computer Interface

Zander, T. O., Kothe, C., Welke, S., and Rötting, M. (2009). “Utilizing Zuo, C., Jin, J., Yin, E., Saab, R., Miao, Y., Wang, X., et al. (2020). Novel
secondary input from passive brain-computer interfaces for enhancing hybrid brain-computer interface system based on motor imagery
human-machine interaction,” in International Conference on Foundations and p300. Cogn. Neurodyn. 14, 253–265. doi: 10.1007/s11571-019-0
of Augmented Cognition (Las Vegas, NV: Springer), 759–771. 9560-x
doi: 10.1007/978-3-642-02812-0_86
Zhang, R., Xu, P., Chen, R., Li, F., Guo, L., Li, P., et al. (2015). Predicting Conflict of Interest: The authors declare that the research was conducted in the
inter-session performance of SMR-based brain-computer interface using absence of any commercial or financial relationships that could be construed as a
the spectral entropy of resting-state EEG. Brain Topogr. 28, 680–690. potential conflict of interest.
doi: 10.1007/s10548-015-0429-3
Zhang, W., and Wu, D. (2020). Manifold embedded knowledge transfer for brain- Copyright © 2021 Saha, Mamun, Ahmed, Mostafa, Naik, Darvishi, Khandoker and
computer interfaces. IEEE Trans. Neural Syst. Rehabil. Eng. 28, 1117–1127. Baumert. This is an open-access article distributed under the terms of the Creative
doi: 10.1109/TNSRE.2020.2985996 Commons Attribution License (CC BY). The use, distribution or reproduction in
Zhao, X., Chu, Y., Han, J., and Zhang, Z. (2016). SSVEP-based brain- other forums is permitted, provided the original author(s) and the copyright owner(s)
computer interface controlled functional electrical stimulation system for are credited and that the original publication in this journal is cited, in accordance
upper extremity rehabilitation. IEEE Trans. Syst. Man Cybernet. Syst. 46, with accepted academic practice. No use, distribution or reproduction is permitted
947–956. doi: 10.1109/TSMC.2016.2523762 which does not comply with these terms.

Frontiers in Systems Neuroscience | www.frontiersin.org 20 February 2021 | Volume 15 | Article 578875