Recent Advances on Brain‑Computer Interface Applications and

Challenges in Stroke Rehabilitation

Haoqi Wang

Collefe of Life Sciences and Engineering, South West Jiaotong University, Chengdu, China

Keywords: Brain‑Computer Interface (BCI), Stroke Rehabilitation, Motor Recovery.

Abstract: Stroke is the severe diseases for patients and their families. For many patients, the post-stroke motor disability

means a really tough way to go back to the normal lives and a heavy burden for the patients’ families. To treat

the motor disability, the main strategy for increasing the primary motor cortex's activity through both

medication and physical training is active motor training. However, those patients with severe motor disability

may face difficulties when doing rehabilitation training as it is more difficult to track the rehabilitation process

and observe outward improvements. Nowadays, the problem could be solved step by step with the assist of

Brain-Computer Interface (BCI). Through translating the brain activity into specific signals and commands

that guide the external devices, BCI can enhance the process of Motor Imagery and assist patients with specific

feedback. This review summarize recent advances in stroke treatment with BCI and more applications.

1 INTRODUCTION

Brain-Computer Interface (BCI) is a recently

developed biotechnology that has gained attention

in recent years. It serves as a communication

pathway between the brain and electronic devices

such as computers. BCI measures, decodes, and

translates brain activity into electrical and magnetic

signals, then outputs this transformed information

to external machines to execute corresponding

actions. Due to its ability to convert neural signals,

BCI has the potential to aid in the treatment of

neurological diseases. Unlike the traditional

pathways for brain signal output (peripheral nerves

and muscle tissues), for individuals with significant

motor dysfunction, BCI can offer alternate

communication channels, improving their ability to

communicate with their environment. The signal

acquisition unit, signal processing unit, control unit,

and application unit are the four primary parts of

BCI. Microelectrodes, optoprobes, and

magnetoprobes are the three types of probes utilized

in BCI. Based on their degree of invasiveness and

signal processing synchronization, BCIs are further

divided into intrusive, somewhat invasive, and

noninvasive categories (Awuah et al. 2024). The

invasive type is the most precise, as it directly

interacts with intracortical electrodes.

Consequently, invasive BCIs are currently a major

focus in BCI research. However, the implantation

process may cause brain tissue damage or incorrect

information transmission. BCI has the remarkable

potential to translate brain activity, enabling people

to control external devices through their immediate

thoughts. In the medical field, BCI offers

significant advantages by bypassing the normal

pathways of neural signal output and allowing

patients with severe motor dysfunction to

communicate directly with the real world (Wen et

al. 2021). As a result, BCI is considered a valuable

tool in assisting patients with daily communication

and neurorehabilitation.

The diminution or total loss of function in one

or more bodily parts is referred to as a motor

disability. Many neurological diseases, such as

stroke, spinal cord injury, brain injury, Parkinson's

disease, and cerebral palsy, can cause motor

disabilities. The most crucial strategy for treating

motor impairments after neurological disorders is

rehabilitation training. In physical and occupational

therapy, active motor training is a popular

technique for promoting motor recovery in patients

by increasing primary motor cortex activity.

However, active motor training may not be

effective for patients with severe motor disabilities.

For example, individuals with paralysis or

paraplegia may not be able to benefit from

254

Wang, H.

Recent Advances on Brain-Computer Interface Applications and Challenges in Stroke Rehabilitation.

DOI: 10.5220/0014486100004933

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

In Proceedings of the 1st International Conference on Biomedical Engineering and Food Science (BEFS 2025), pages 254-260

ISBN: 978-989-758-789-4

rehabilitation training through physical and

occupational therapy (Chen et al. 2023). The

process of Motor Imagery alters brain neuronal

connections to adapt to the physical movement

information that patients rehearse in their minds.

This process is known as neuroplasticity. In motor

rehabilitation, BCI is essential for supporting

neuroplasticity because it gives the central nervous

system useful feedback. By monitoring changes in

brain activity in response to a stimulus or during

voluntary movement practice, it aids in directing

plasticity (Chen et al. 2023). This, in turn, helps

patients access their motor systems and facilitates

rehabilitation across all stages of motor recovery.

BCIs are currently developing in two main

directions. The first is enhancing the abilities of

healthy individuals by integrating AI within their

brains or externally on their bodies, allowing them

to perform tasks that would otherwise be impossible

without BCI. The second is their application in the

treatment of neurological diseases, enabling

patients to regain the ability to express their

thoughts like able-bodied individuals. This review

primarily focuses on the application of BCI in

motor rehabilitation for patients who have suffered

from strokes.

2 THE SIGNIFICANCE AND

CURRENT APPLICATIONS OF

BRAIN-COMPUTER

INTERFACE

Brain Computer interface is an emerging technique

that facilitates the communication between Human

Brain and Artificial Intelligence devices. The

emergence of BCI has changed a large quantity of

industries including entertainment, gaming,

automation, education, medical field and so on

(Maiseli et al. 2023). Connecting to AI through our

brain means a real dramatic change in our lives. AI

possesses implausibly powerful functions that never

could be achieved by human beings. They contain all

the knowledge that are input by people, they have the

ability to study by themselves and improve their study

abilities, they are also able to work as calculators to

aid people shorten their time in meaningful and time-

consuming things. Therefore, BCI, which is the

bridge between our brain and AI, works as a built -in

AI that aids our brain and enhance our abilities as a

average person. The applications of Brain-Computer

Interfaces can be seen in many fields. In clinical

fields, BCI can treat with many neurodiseases and

help those patients back to the society better. For

example, post-stroke rehabilitation with BCI may

help patients gain motor and sensor ability (Yang et

al. 2021). However, recent advances in non-invasive

and portable brain imaging techniques related to

EEG, have also facilitated the development of novel

applications outside the medical and scientific areas.

BCI has had a try in video game fields. Many famous

games have been introduced to people with BCI like

“Pacman” and “World of warcraft”, people may

obtain better game experience through BCI as they

can enhance our specific perception (Ahn et al. 2014).

Beside that, fields of biometrics Authentification and

civil and military aviation fields can also be highly

integrated with BCI. In Biometrics authentification

field, based on the research of

electroencephalography, BCI is constructed for

authentification. Brain-computer interface (BCI)

systems establish direct human-machine

communication by circumventing traditional motor

pathways. These systems rely on the extraction of

distinctive neural patterns from

electroencephalogram (EEG) signals instead of motor

characteristics, which are subsequently classified into

specific cognitive states. These states are mapped to

predefined machine commands. To address inter-

subject variability, the extracted neural features must

exhibit cross-user generalizability. In contrast, EEG-

based identity recognition systems operate under an

opposing principle: their goal is to distinguish

individuals even when performing identical tasks.

Here, inter-individual differences in neural feature

patterns become advantageous, serving as

discriminative markers for personal identification

(Chan et al. 2018). Nowadays, BCI do have great

potential in many fields and improve our experiences

in our daily lives, it has the strength to change our

world in the near future.

3 RECENT ADVANCEMENTS

AND APPLICATIONS OF

BRAIN-COMPUTER

INTERFACE IN STROKE

REHABILITATION

Stroke is a quite severe neurodiseases for human

beings, many patients loss their motor ability and

sensory ability after stroke, which leads to unabling

to connect with the society and live by the patients

selves. As a result, stroke not only causes self-

Recent Advances on Brain-Computer Interface Applications and Challenges in Stroke Rehabilitation

255

disability in patients, but also imposes burdens to the

patients’ family. The patients family needs to pay

attention to take care of the patients, it’s a cost of

time, money and energy. The recovery process also

means a tough way as there is no fully valid way to

help those patients recover through traditional

rehabilitation methods. However, Brain-Computer

Interface seems to have the ability to help patients to

recover through connecting the motor or sensory

signals with the realistic motion and feeling. In this

way, patients of stroke may have the chances to

recover and return to normal lives. BCI’s applications

after stroke focus on several aspects, such as motor

rehabilitation and sensory rehabilitation according to

the recovery aspects, or focus on the upper, lower

limbs and the hand part according to the position that

requires to be treated, non-invasive therapy and

invasive therapy according to their position

comparing to the brain. One of the applications of

BCI in stroke rehabilitation is the combination of BCI

and the exoskeletons. Individuals suffering from

severe muscle paralysis or impaired motion function

can realize movement through the Brain/Exoskeleton

devices by transforming their thoughts in the brain

into the command in the exoskeleton. The B/NE

devices can trigger motor rehabilitation after repeated

use over several weeks. Also, BCI could also be used

to drive electric stimulator to activate peripheral

muscles in the form of functional electric stimulation

to help those patients gain the ability to achieve

movement and recover their motor ability through

constant movements controlled by the brain (Colucci

et al. 2022). The BCI-FES therapy help with

promoting functional recovery and purposeful

plasticity by activating the body's natural input and

output pathways. Therefore, the motor rehabilitation

and neural reactivation are facilitated (Yang et al.

2021). BCI-VR is also an emerging technique in

stroke rehabilitation, comparing to traditional

rehabilitation method, BCI-VR method shows more

attraction for the patients to try rehabilitation as

normal method may require a lot of vigor and energy,

which make it boring and fatigued for patients. With

VR, patients have more motivations when doing

exercises, this will shorten the rehabilitation cycle

and provide more meaningful feedbacks (Wen et al.

2021). Through motor training, BCI-VR systems

could track and support cortical reorganization

(Bermúdez i Badia et al. 2013). The benefit also

shows when VR is used as an auxiliary equipment,

the treatment time and the therapeutic effect can all

be improved a lot (Vourvopoulos et al. 2019). As a

result, the BCI shows great potential in rehabilitation

in post-stroke treatment and will be applied more

widely in the future.

4 MOTOR REHABILITATION

AFTER STROKE

Motor disability is a very important pattern after

stroke, which means the partially or completely loss

of motor function at specific parts in patients’ bodies.

Motor disability may cause it harder for patients to

achieve movement casually according to their

thoughts. Reduced muscular function, impaired

motor coordination, or even paralysis can result from

a motor impairment (Chen et al. 2023). In order to

help individuals with motor disabilities return to their

regular life, rehabilitation training is one of the most

crucial treatments. The primary method for

promoting motor rehabilitation in patients is active

motor training. Occupational and physical therapy as

well as pharmaceutical interventions are the primary

training approaches (Khan et al. 2020). The basic idea

behind these techniques is that they might increase

primary motor cortex activation. However, those

patients with severe motor disability like paralysis in

limbs may face difficulties when doing rehabilitation

training as It is more difficult to track the healing

process and observe improvements on the

outside.Facing these problems, BCI technique

provides an alternative avenue for motor

rehabilitation. With the ability to bypass the normal

output pathway of neuro signals, BCI measures,

translates and transform electromagnetic or brain

activity into command that controls external devices.

This process cross the normal pathway that deliver

signals form brain to specific body part. Instead, BCI

recognizes and translates the brain activity into

specific command and instruct the external devices to

operate.

In Motor rehabilitation, motor imagery is an

important part that can rehearse movement in the

patients’ brains, which is meaningful to the

neuroplasticity (Belda-Lois et al. 2011). There are

two steps involved in improving neuroplasticity.

The functional plasticity linked to the synaptic

efficiency alterations occurs during the first phase

on a time scale ranging from a few minutes to a few

days. Changes in synapse strength are strongly

linked to learning new abilities and creating

memories (Nicoll 2017). Functional plasticity

enhances learning consolidation in the latter stage

through changes in brain structure. New synapses

and axons may form, axons and dendrites branches

BEFS 2025 - International Conference on Biomedical Engineering and Food Science

256

will also change, these changes are long-term

modifications by particular intervention (Rossini et

al. 2012). BCI can stimulate neuroplasticity to

improve patients’ movement ability in four

different mechanisms. The first is called the

neurofeedback training, when the patients get visual

or auditory representation that they are interested

in, their brain activity will be more active, then BCI

will provide meaningful feedback to central nerve

in order to help neuroplasticity. The second

mechanism is reinforecement-based operant

conditioning, which synchronize the motor imagery

and actual movement by external devices. During

this mechanism, correct and successful imagery

will get positive feedback and real movement in

specific part while wrong trial will not get real

movement, even negative feedback will be used

(Remsik et al. 2018). The third mechanism is

repetitive engagement. When the patients achieve

movement repetitively, Repetitive activation of

related stroke-affected neural circuits may improve

the axon connection and its current creation,

thereby curing motor impairment (Mrachacz-

Kersting et al. 2021). The last mechanism is derived

from the Hebbian learning principle. The synaptic

strength between neurons is strengthened when they

are occasionally activated. The lack of motor

control after stroke cause the difference between

motor intension and execution, the difference is

caused by lacking in the input feedback of

movement execution, which can decreases the

inhibitory drive on the motor system. BCI works by

providing sensory feedback after movement

execution (Ang et al. 2014).

5 TECHNIQUES APPLYING IN

MOTOR-REHABILITATION OF

BRAIN-COMPUTER

INTERFACE

The Brain-Computer Interface (BCI) functions works

in motor rehabilitation mainly depend on four main

components: signal reception, signal processing,

generation of a specific response in a machine and

providing feedback to the central nerve system to

fully achieve treatment effect through the four

mechanism. The signal reception means to collect

brain activity into the BCI system. There are two

ways: invasive and non-invasive. Invasive shows

high precision but also high risk in clinical

application. Cortical surface microelectrodes, cortical

penetrating microelectrodes, and profoundly

penetrating electrodes are invasive electrodes that

record distinct aspects of the brain action potential,

resulting in a more accurate outcome. The EEG and

fNIRS are the primary components of the non-

invasive techniques. Using many electrodes applied

to the scalp, EEG captures electrical signals that show

the activity of neuronal populations during a brief

period of time. fNIRS tracks the hemodynamic

activity of the brain by measuring variations in the

intensity of near-infrared light that has passed through

the scalp and brain (Jöbsis 1977). Both of them

possess good non-invasive property and are portable,

they can also make up each others’ flaws. Therefore,

combining them together is becoming a potential

method. The second process contains specific

techniques to filter received brain signals and explain.

Filtering and explanation are done via sensorimotor

rhythms, slow cortical potentials, event-related

potentials, and visual evoked potentials. These

filtered signals are converted into voltage/time

frequencies using methods like Fourier, common

spatial filter, and wavelet transform. These signals are

then subjected to additional analysis using

classification algorithms before being output as a

specific command for the external devices (Burns et

al. 2014, Cervera et al. 2018). Then about the

generation of a specific response in a machine, the

devices are programmed to receive command and

execute functions like basic movement to help

rehabilitation and improve life quality. The last part

is providing feedback to the nerve system and the

brain. Brain signals are converted by BCI algorithms

into both devices that provide real-time feedback and

control commands for external movement execution

devices. The patient represents or tries to produce

passive limb movement, which is carried out by an

orthosis, robot, or exoskeleton arm in the BCI loop.

Prior RCTs have most frequently employed this

kinaesthetic form of input, sometimes in conjunction

with visual feedback (Frolov et al. 2018). The

functional electrical stimulation (FES) in the BCI

loop is considered the most preferable in physiology.

When executing FES, more motor and sensory axons

are depolarized, more powerful signals from muscles

spindles are delivered to the central nerve system, the

pulses from the muscle spindles can activate motor

neurons simultaneously with the descending cortical

command when representing a movement, thus

inducing Hebbian association (Fu et al. 2022). With

these useful techniques, motor rehabilitation have

more chances to be treated or cured.

Recent Advances on Brain-Computer Interface Applications and Challenges in Stroke Rehabilitation

257

6 CHALLENGES AND

OPPORTUNITIES: THE

FUTURE OF BRAIN-

COMPUTER INTERFACE IN

NEUROREHABILITATION

The post-stroke recovery phase has historically been

separated into three categories: acute, functional or

subacute, and chronic or plateau. BCI that combining

with other specific techniques not only benefit

subacute phase, but also benefits the chronic phase.

Comparing to the traditional rehabilitation methods,

the time scale of rehabilitation is small and may not

that useful when facing severe patuents (Marín-

Medina et al. 2024). The future of BCI devices for

neurorehabilitation is “BCI+X” mode. Many BCI

devices show more possibilities cooperating with

other techniques. “BCI+VR” mode make it easier for

patients to take treatment with more attraction as it is

a fatigued and painful training, the rehabilitation

cycle will be shorter and the feedback would be more

meaningful. Also, interaction between human brain

and external BCI would provide a better working

direction towards motor rehabilitation field.

The challenges also exist in the future of

application of BCI. Though BCI show its great

potential and power in motor rehabilitation, the field

of post-stroke cognitive and speech rehabilitation

using BCI still requires more efforts as it is at the first

step of research. However, this field have not got

focus from the whole society. The speech and

cognitive function exist positive interactions, the

cognitive function also will affect the motor

rehabilitation, the success in cognitive and speech

rehabilitation will also have a positive effect on the

motor rehabilitation. As a result, more attention

should be paid to cognitive and speech rehabilitation.

In the future, recognizing different aspects of

rehabilitation of post-stroke as an entirety should be

the definite goal to aid more patients away from the

stroke and return to normal lives. Another challenge

is the difference in individual’s ability to use Motor

Imagery to control the non-invasive Brain-Computer

Interface, some patients may still have low or

unstable control quality, which leads to low

motivation to participate rehabilitation. This situation

requires more intelligent signal processing algorithm

and more specific adaptation to different degrees of

cerebral injury. Another practical issue is the patients'

fatigue during therapy. One common symptom

following a stroke is fatigue (Alghamdi et al. 2021).

It requires focusing on the rehabilitation for such a

long time that make it not that easy for patients to

overcome. As a result, more rehabilitation forms like

combining with AI should be presented to the patients

to make them have more attention on rehabilitation.

Though there are still many questions that exists in

the future, it is no denying that BCI has the potential

to fully cure neurodiseases like stroke.

7 CONCLUSION

This review discusses recent advances of BCI’s

applications in the field of post-stroke rehabilitation,

especially the motor rehabilitation and relative

cutting-edge technique combined with BCI. Research

shows, motor rehabilitation with BCI has a large

quantity of advantages than traditional “active motor

training” when treating motor disability after stroke.

As an alternative avenue, BCI bypasses the normal

output pathway through translating the brain activity

into command and signals to guide external devices

to give patients abilities to achieve movement

passively. Then the BCI works to build connections

in neurons again to help the motor rehabilitation.

Neuroplasticity is the important parts in motor

rehabilitation. BCI could stimulate neuroplasticity

with four mechanisms with its specific abilities like

feedback offering, external devices assistance and so

on. Techniques combined with EEG, fNIRS and VR

are widely used to improve the treatment effect.

These breakthroughs not only deepen our

understanding of motor rehabilitation mechanisms,

but also provides new therapy for the personalized

medicine.

Though many advancements have been achieved,

BCI therapy for motor rehabilitation also faces a large

quantities of problems. For invasive BCI, it means a

precise detector of our brain activity but with high

surgical risk and may not that suitable for many

patients. For non-invasive BCI, though it means

lower risk, however, it may detect imprecisely and

affect the translation part and subsequent feedback

part and forms wrong connections. Also, brain

science is still a subject that requires more research

into it. As a result, translating the brain activity may

also a tough issue for BCI, simple motor

rehabilitation may works, but more elaborate

functions and movement that enable patients to go

back to normal lives still requires more knowledge of

brain science. The future research should focus on the

development of safer and more economical BCI

therapy for stroke patients. Exploring the combined

therapy’s potential like “BCI+X” mode. Otherwise,

BEFS 2025 - International Conference on Biomedical Engineering and Food Science

258

the further study of the brain activity, further

improvement of artificial intelligence and the

advancement in detectors like EEG and fNIRS will

also benefit the optimizing of the current therapy. In

conclusion, the rapid development of the BCI therapy

for the post-stroke motor rehabilitation brings new

hopes for the patients. The BCI therapy is a complex

therapy that contains a huge range of knowledge from

different fields. With the further development of the

technique and the research, BCI therapy will find a

better way to cure patients with pain.

REFERENCES

Ahn, M., Lee, M., & Choi, J., et al. 2014. A review of brain-

computer interface games and an opinion survey from

researchers, developers and users. Sensors 14(8):

14601-14633.

Alghamdi, I., Ariti, C., & Williams, A., et al. 2021.

Prevalence of fatigue after stroke: A systematic review

and meta-analysis. European Stroke Journal 6(4): 319-

332.

Ang, K. K., Guan, C., & Phua, K. S., et al. 2014. Brain-

computer interface-based robotic end effector system

for wrist and hand rehabilitation: results of a three-

armed randomized controlled trial for chronic stroke.

Frontiers in Neuroengineering 7: 30.

Awuah, W. A., Ahluwalia, A., & Darko, K., et al. 2024.

Bridging Minds and Machines: The Recent Advances

of Brain-Computer Interfaces in Neurological and

Neurosurgical Applications. World Neurosurgery 189:

138-153.

Belda-Lois, J. M., Mena-del Horno, S., & Bermejo-Bosch,

I., et al. 2011. Rehabilitation of gait after stroke: a

review towards a top-down approach. Journal of

Neuroengineering and Rehabilitation 8: 66.

Bermúdez i Badia, S.,García Morgade, A., & Samaha, H.,

et al. 2013. Using a hybrid brain computer interface and

virtual reality system to monitor and promote cortical

reorganization through motor activity and motor

imagery training. IIEEE Transactions on Neural

Systems and Rehabilitation Engineering 21(2): 174-

181.

Burns, A., Adeli, H., & Buford, J. A. 2014. Brain-computer

interface after nervous system injury. The

Neuroscientist 20(6): 639-651.

Cervera, M. A., Soekadar, S. R., & Ushiba, J., et al. 2018.

Brain-computer interfaces for post-stroke motor

rehabilitation: a meta-analysis. Annals of Clinical and

Translational Neurology 5(5): 651-663.

Chan, H. L., Kuo, P. C., & Cheng, C. Y., et al. 2018.

Challenges and Future Perspectives on

Electroencephalogram-Based Biometrics in Person

Recognition. Frontiers in Neuroinformatics 12: 66.

Chen, J., Xia, Y., & Zhou, X., et al. 2023. fNIRS-EEG BCIs

for Motor Rehabilitation: A Review. Bioengineering

10(12) :1393.

Colucci, A., Vermehren, M., & Cavallo, A., et al. 2022.

Brain-Computer Interface-Controlled Exoskeletons in

Clinical Neurorehabilitation: Ready or Not?

Neurorehabilitation and Neural Repair 36(12): 747-

756.

Frolov, A. A., Bobrov, P. D., & Biryukova, E. V., et al.

2018. Electrical, Hemodynamic, and Motor Activity in

BCI Post-stroke Rehabilitation: Clinical Case Study.

Frontiers in Neurology 9: 1135.

Fu, J., Chen, S., & Jia, J. 2022. Sensorimotor Rhythm-

Based Brain-Computer Interfaces for Motor Tasks

Used in Hand Upper Extremity Rehabilitation after

Stroke: A Systematic Review. Brain Science 13(1) :56.

Jöbsis, F. F. 1977. Noninvasive, infrared monitoring of

cerebral and myocardial oxygen sufficiency and

circulatory parameters. Science 198(4323): 1264-1267.

Khan, M. A., Das, R., & Iversen, H. K., et al. 2020. Review

on motor imagery based BCI systems for upper limb

post-stroke neurorehabilitation: From designing to

application. Computers in Biology and Medicine 123:

103843.

Maiseli, B., Abdalla, A. T., & Massawe, L. V., et al. 2023.

Brain-computer interface: trend, challenges, and

threats. Brain Informatics 10(1): 20.

Marín-Medina, D. S., Arenas-Vargas, P. A., & Arias-

Botero, J. C., et al. 2024. New approaches to recovery

after stroke. Neurological Sciences 45(1): 55-63.

Mrachacz-Kersting, N., Ibáñez, J., & Farina, D. 2021.

Towards a mechanistic approach for the development

of non-invasive brain-computer interfaces for motor

rehabilitation. Journal of Physiology 599(9): 2361-

2374.

Nicoll, R. A. 2017. A Brief History of Long-Term

Potentiation. Neuron 93(2): 281-290.

Remsik, A. B., Dodd, K., & Williams, L., et al. 2018.

Behavioral Outcomes Following Brain-Computer

Interface Intervention for Upper Extremity

Rehabilitation in Stroke: A Randomized Controlled

Trial. Frontiers in Neuroscience 12: 752.

Rossini, P. M., Noris Ferilli, M. A., & Ferreri, F. 2012.

Cortical plasticity and brain computer interface.

European Journal of Physical and Rehabilitation

Medicine 48(2): 307-312.

Vourvopoulos, A., Pardo, O. M., & Lefebvre, S., et al.

2019. Effects of a Brain-Computer Interface With

Virtual Reality (VR) Neurofeedback: A Pilot Study in

Chronic Stroke Patients. Frontiers in Human

Neuroscience 13: 210.

Wen, D., Fan, Y., & Hsu, S. H., et al. 2021. Combining

brain-computer interface and virtual reality for

rehabilitation in neurological diseases: A narrative

review. Annales de Physique Médicale et de

Réadaptation 64(1): 101404.

Recent Advances on Brain-Computer Interface Applications and Challenges in Stroke Rehabilitation

259

Yang, S., Li, R., & Li, H., et al. 2021. Exploring the Use of

Brain-Computer Interfaces in Stroke

Neurorehabilitation. BioMed Research International

2021: 9967348.

BEFS 2025 - International Conference on Biomedical Engineering and Food Science

260