Introduction
Brain Computer Interface (BCI) systems depend on machine
learning algorithms to decode the brain activity [5].
Artificial intelligence (AI) algorithms have very important
role in the development of brain-computer interfaces. AI
algorithms such as neural networks play powerful role in
brain-computer interfaces [6]. Recent advances in the
computational innovation have led to significant developments
in BCI [7].
BCI translate brain activity patterns into commands which can
be executed by an artificial device. Motor BCIs can be used
for the augmentation of motor function by activating the
neuromuscular circuit [8]. In future, BCIs may become an
important communication and control technology for people with
disabilities. The current range of applications also targets
cognitive impairments and opportunities for human enhancement
[9].
Deep Learning Technology in BCI
Big data analysis is essential to understand the complexities
of brain activity with regard to BCI performance [3].
Advancements in deep learning have led many researchers to
adopt deep neural networks to extract features from brain
signals [10].
Deep learning can solve complex tasks using EEG data.
Researchers are now doing a lot of work on deep learning-based
technology in the BCI field. Important deep learning
algorithms employed in EEG-based BCI applications are
convolutional neural network (CNN), long short-term memory,
recurrent neural network, stacked autoencoder and variational
autoencoder. CNN has considerable success in several research
areas making it an ideal option. Many BCI techniques produce
two-dimensional visuals that can be processed by CNNs.
Convolutional neural network is the most frequent deep
learning algorithm used in EEG-based BCI applications [7].
BCIs based on steady-state visual evoked potentials, have the
highest information transfer rate. Deep learning has provided
an effective solution for solving complex classification
problems and several researchers are now using deep learning
to classify steady-state visual evoked potential signals [11].
Five categories of deep learning techniques used in
steady-state visual evoked potential based BCI applications
are convolutional neural network, recurrent neural network,
deep neural network, long short-term memory and restricted
Boltzmann machine [12].
BCI based on functional near-infrared spectroscopy (fNIRS) is
a promising application. fNIRS measures functional changes in
cerebral hemodynamics. Deep learning methods can be used in
fNIRS decoding. In the study by Qin Y et al, an end-to-end
hybrid neural network was proposed for feature extraction of
fNIRS. The method uses a spatial-temporal convolutional layer
and a spatial attention mechanism. A temporal convolutional
network is used to utilize the temporal information of fNIRS.
This deep learning method has high accuracy and provides an
important reference for development of BCI [1].
In functional near-infrared spectroscopy brain-computer
interface (fNIRS-BCI) systems, deep learning algorithms have
very important role in enhancing accuracy. Deep learning
neural networks automatically extract hidden features within a
dataset to classify the data. Integrated contextual gate
network (ICGN) algorithm applied to the dataset yielded
significantly higher classification accuracy compared to long
short-term memory (LSTM) and bidirectional long short-term
memory (Bi-LSTM) [13]. Enhanced performance of fNIRS-BCI in
terms of classification accuracy can be achieved using deep
learning algorithms, including convolutional neural networks
[14].
Conclusion
Brain-computer interface is a highly promising human-computer
interaction method. It is an advanced multidisciplinary domain
that has attracted great deal of research in recent years.
Brain-computer interfaces are used in a variety of application
areas. BCI research is expanding in the breadth of its
applications. For improving BCI performance, collaboration is
essential between clinicians, scientists and experts in data
analysis. Researchers are now working on deep learning-based
approaches in the BCI field. Deep learning algorithms have
shown promising results in BCI research.
References
1. Qin Y, Li B, Wang W, Shi X, Peng C, Lu Y. Classification
Algorithm for fNIRS-based Brain Signals Using Convolutional
Neural Network with Spatiotemporal Feature Extraction
Mechanism. Neuroscience. 2024 Mar 26;542:59-68. doi:
10.1016/j.neuroscience.2024.02.011. Epub 2024 Feb 17. PMID:
38369007
2. Chen Y, Wang F, Li T, Zhao L, Gong A, Nan W, Ding P, Fu Y.
Considerations and discussions on the clear definition and
definite scope of brain-computer interfaces. Front Neurosci.
2024 Aug 5;18:1449208. doi: 10.3389/fnins.2024.1449208. PMID:
39161655; PMCID: PMC11330831.
3. Huggins JE, Guger C, Ziat M, Zander TO, Taylor D,
Tangermann M, Soria-Frisch A, Simeral J, Scherer R, Rupp R,
Ruffini G, Robinson DKR, Ramsey NF, Nijholt A, Muller-Putz G,
McFarland DJ, Mattia D, Lance BJ, Kindermans PJ, Iturrate I,
Herff C, Gupta D, Do AH, Collinger JL, Chavarriaga R, Chase
SM, Bleichner MG, Batista A, Anderson CW, Aarnoutse EJ.
Workshops of the Sixth International Brain-Computer Interface
Meeting: brain-computer interfaces past, present, and future.
Brain Comput Interfaces (Abingdon). 2017;4(1-2):3-36. doi:
10.1080/2326263X.2016.1275488. Epub 2017 Jan 30. PMID:
29152523; PMCID: PMC5693371.
4. Lv Z, Qiao L, Wang Q, Piccialli F. Advanced
Machine-Learning Methods for Brain-Computer Interfacing.
IEEE/ACM Trans Comput Biol Bioinform. 2021
Sep-Oct;18(5):1688-1698. doi: 10.1109/TCBB.2020.3010014. Epub
2021 Oct 7. PMID: 32750892.
5. Iturrate I, Chavarriaga R, Millan JDR. General principles
of machine learning for brain-computer interfacing. Handb Clin
Neurol. 2020;168:311-328. doi:
10.1016/B978-0-444-63934-9.00023-8. PMID: 32164862.
6. He R. Perspective of Signal Processing-Based on
Brain-Computer Interfaces Using Machine Learning Methods. Stud
Health Technol Inform. 2023 Nov 23;308:295-302. doi:
10.3233/SHTI230853. PMID: 38007753.
7. Hossain KM, Islam MA, Hossain S, Nijholt A, Ahad MAR.
Status of deep learning for EEG-based brain-computer interface
applications. Front Comput Neurosci. 2023 Jan 16;16:1006763.
doi: 10.3389/fncom.2022.1006763. PMID: 36726556; PMCID:
PMC9885375.
8. Wang J, Bi L, Fei W. EEG-Based Motor BCIs for Upper Limb
Movement: Current Techniques and Future Insights. IEEE Trans
Neural Syst Rehabil Eng. 2023;31:4413-4427. doi:
10.1109/TNSRE.2023.3330500. Epub 2023 Nov 10. PMID: 37930905.
9. Huggins JE, Krusienski D, Vansteensel MJ, Valeriani D,
Thelen A, Stavisky S, Norton JJS, Nijholt A, Muller-Putz G,
Kosmyna N, Korczowski L, Kapeller C, Herff C, Halder S, Guger
C, Grosse-Wentrup M, Gaunt R, Dusang AN, Clisson P,
Chavarriaga R, Anderson CW, Allison BZ, Aksenova T, Aarnoutse
E. Workshops of the Eighth International Brain-Computer
Interface Meeting: BCIs: The Next Frontier. Brain Comput
Interfaces (Abingdon). 2022;9(2):69-101. doi:
10.1080/2326263X.2021.2009654. Epub 2022 Feb 8. PMID:
36908334; PMCID: PMC9997957.
10. Ahn M, Jun SC, Yeom HG, Cho H. Editorial: Deep Learning in
Brain-Computer Interface. Front Hum Neurosci. 2022 May
19;16:927567. doi: 10.3389/fnhum.2022.927567. PMID: 35664345;
PMCID: PMC9161139.
11. Xu D, Tang F, Li Y, Zhang Q, Feng X. An Analysis of Deep
Learning Models in SSVEP-Based BCI: A Survey. Brain Sci. 2023
Mar 13;13(3):483. doi: 10.3390/brainsci13030483. PMID:
36979293; PMCID: PMC10046535.
12. Albahri AS, Al-Qaysi ZT, Alzubaidi L, Alnoor A, Albahri
OS, Alamoodi AH, Bakar AA. A Systematic Review of Using Deep
Learning Technology in the Steady-State Visually Evoked
Potential-Based Brain-Computer Interface Applications: Current
Trends and Future Trust Methodology. Int J Telemed Appl. 2023
Apr 30;2023:7741735. doi:10.1155/2023/7741735. PMID: 37168809;
PMCID: PMC10164869.
13. Akhter J, Naseer N, Nazeer H, Khan H, Mirtaheri P.
Enhancing Classification Accuracy with Integrated Contextual
Gate Network: Deep Learning Approach for Functional
Near-Infrared Spectroscopy Brain-Computer Interface
Application. Sensors (Basel). 2024 May 10;24(10):3040. doi:
10.3390/s24103040. PMID: 38793895; PMCID: PMC11125334.
14. Hamid H, Naseer N, Nazeer H, Khan MJ, Khan RA, Shahbaz
Khan U. Analyzing Classification Performance of fNIRS-BCI for
Gait Rehabilitation Using Deep Neural Networks. Sensors
(Basel). 2022 Mar 1;22(5):1932. doi: 10.3390/s22051932. PMID:
35271077; PMCID: PMC8914987.