sensors

Editorial
EEG Signal Processing Techniques and Applications
Yifan Zhao 1, * , Fei He 2 and Yuzhu Guo 3

1 School of Aerospace, Transport and Manufacturing, Cranfield University, Cranfield MK43 0AL, UK
2 Research Centre for Computational Science and Mathematical Modelling, Coventry University,
Coventry CV1 5FB, UK; [email protected]
3 School of Automation Science and Electrical Engineering, Beihang University, Beijing 100191, China;
[email protected]
* Correspondence: [email protected]

1. Background
Electroencephalography (EEG) is a widely recognised non-invasive method for cap-
turing brain electrophysiological activity. It stands out for its cost-effectiveness, portability,
ease of administration, and widespread availability in most hospital settings. Unlike other
neuroimaging modalities focused on anatomical structure, such as MRI, CT, and fMRI,
EEG excels in providing ultra-high time resolution, a crucial asset for in-depth insights into
brain functioning [1].
The empirical interpretation of EEG data predominantly relies on the identification of
abnormal frequency patterns in distinct biological states (e.g., wakefulness versus sleep [2])
and the spatial-temporal and morphological characteristics of paroxysmal [3] and per-
sistent discharges [4]. Reactivity to external stimuli and activation procedures, such as
intermittent photic stimulation or hyperventilation, also plays a significant role in EEG anal-
ysis [5,6]. While these practical approaches are valuable in many cases, they often fall short
of capturing the intricate, dynamic, and nonlinear interactions among various anatomical
constituents of the brain networks. These interactions frequently remain hidden within the
EEG recordings, surpassing the observational capabilities of even highly trained physicians
in the field. This oversight is supported by substantial evidence across various neurolog-
ical conditions, including epilepsy, neurodegenerative dementias, neuropsychiatric and
movement disorders, as well as normal cognitive paradigms [7].
Moreover, EEG data are inherently nonstationary and susceptible to various sources
Citation: Zhao, Y.; He, F.; Guo, Y.
of noise, notably frequency interference. Consequently, the effective removal of noise from
EEG Signal Processing Techniques
raw EEG data is imperative to extract meaningful information that accurately reflects brain
and Applications. Sensors 2023, 23, activity and states [8]. In recent years, approaches based on machine learning have attracted
9056. https://doi.org/10.3390/ considerable attention due to their exceptional capability to unveil underlying patterns
s23229056 within noisy EEG recordings for various applications.
This Special Issue serves as a platform for the dissemination of original high-quality
Received: 17 October 2023
research in EEG signal pre-processing, modelling, analysis, and their applications, with a
Accepted: 6 November 2023
particular focus on the utilisation of machine learning and deep learning techniques. The
Published: 9 November 2023
range of applications covered includes the following:
• Healthcare applications, including epilepsy (contributions 1–3) and anaesthesia (con-
tribution 4);
Copyright: © 2023 by the authors. • Studies related to emotion (contributions 5–7);
Licensee MDPI, Basel, Switzerland. • Research on motor imagery (contributions 8–10);
This article is an open access article
• Investigations into external stimulations (contributions 11–13);
distributed under the terms and
• Research concerning mental workload (contributions 14–15);
conditions of the Creative Commons
• Studies in satisfaction (contribution 16).
Attribution (CC BY) license (https://
creativecommons.org/licenses/by/
4.0/).

Sensors 2023, 23, 9056. https://doi.org/10.3390/s23229056 https://www.mdpi.com/journal/sensors

Sensors 2023, 23, 9056 2 of 5

2. Overview of Contributions
Alreshidi et al. (contribution 14) reported a novel multimodal approach for mental
state detection in pilots using EEG signals. The innovative nature of this study lies in its
combination of advanced automated preprocessing techniques, Riemannian geometry-
based feature extraction, and ensemble learning models, which, together, provide a detailed
and accurate characterization of pilot mental states, ultimately leading to a safer and more
efficient aviation system.
Borra et al. (contribution 8) investigated the power and connectivity in the Alpha and
Beta bands of EEG recordings during planning goal-directed movement. It was suggested
that alpha and beta oscillations are functionally involved in the preparation of reaching
in different ways, with the former mediating the inhibition of the ipsilateral sensorimotor
areas and disinhibition of visual areas, and the latter coordinating disinhibition of the
contralateral sensorimotor and visuomotor areas. This study contributes to enriching
the description of the neural mechanisms underlying reaching movement preparation in
healthy subjects, leading to a better comprehension of the neurophysiological correlates.
Mockevičius et al. (contribution 12) produced a methodology for determining the
individual gamma frequency from EEG data where subjects received auditory stimulation
consisting of clicks with varying inter-click periods. This work demonstrates that the
estimation of individual gamma frequency is possible using a limited number of both the
gel and dry electrodes from responses to click-based chirp-modulated sounds.
Oikonomou et al. (contribution 13) proposed a novel framework to recognise the cog-
nitive and affective processes of the brain during neuromarketing-based stimuli using EEG
signals. More specifically, an extension of the basic Sparse Representation Classification
(SRC) scheme was proposed that utilises the graph properties of neuroimaging data. The
experimental analysis provides evidence that EEG signals could be used for predicting
consumers’ preferences in neuromarketing scenarios.
Yang et al. (contribution 2) presented novel EEG–EEG or EEG–ECG transfer learning
strategies to explore their effectiveness for the training of simple cross-domain convo-
lutional neural networks (CNNs) used in seizure prediction and sleep staging systems,
respectively. It was concluded that transfer learning from an EEG model to produce person-
alised models for a more convenient signal can both reduce the training time and increase
the accuracy; moreover, challenges such as data insufficiency, variability, and inefficiency
can be effectively overcome.
Abdel-Hamid (contribution 5) introduced a subject-dependent emotional valence
recognition method using EEG recordings. Time and frequency features were computed
from only two channels and state-of-the-art performance was achieved and validated by
a benchmark DEAP dataset. This approach would thus be highly attractive for practical
EEG-based emotion AI systems relying on wearable EEG devices.
Shi et al. (contribution 4) proposed a deep residual shrinkage network to estimate
the depth of anesthesia (DoA) from EEG signals. The proposed procedure is not merely
feasible for estimating DoA by mimicking patient state index (PSI) values but also inspired
us to develop a precise DoA-estimation system with more convincing assessments of
anesthetisation levels.
Yuvaraj et al. (contribution 6) contributed another emotion recognition approach
that uses features including statistical features, fractal dimension (FD), Hjorth parameters,
higher order spectra (HOS), and those derived using wavelet analysis. The results of
this research may lead to the possible development of an online feature extraction frame-
work, thereby enabling the development of an EEG-based emotion recognition system in
real time.
Kim et al. (contribution 16) reported a study to use EEG measures to reflect user
satisfaction in controlling a robot hand. For the moment that dominated satisfaction,
it was observed that brain activity exhibited significant differences in satisfaction not
immediately after feeding an input but during the later stage. The other indicators exhibited
independently significant patterns in event-related spectral perturbations. The results
Sensors 2023, 23, 9056 3 of 5

reveal that regardless of subjective satisfaction, objective performance evaluation might

more fully reflect user satisfaction.
As an effort in neuromarketing, Shah et al. (contribution 7) proposed an ensemble
model for predicting emotion using EEG signals to evaluate the consumer’s opinion
toward a product. Automated features were extracted by using a long short-term memory
network (LSTM) and then concatenated with handcrafted features such as power spectral
density (PSD) and discrete wavelet transform (DWT) to create a complete feature set.
This research demonstrates that brain-imaging techniques and tools can help marketers
and advertisement agencies to improve their marketing campaigns before launching the
product in the market and also during the in-market inspection of the campaign’s success
after the launch.
Jochumsen et al. (contribution 10) implemented three performance accommodation
mechanisms (PAMs) in an online motor imagery-based EEG to aid people and evaluate
their perceived control and frustration for stroke rehabilitation. Within the different types
of PAMs, game developers can exercise tremendous artistic freedom to create engaging
interactions for Brain–Computer Interface (BCI) training that either directly manipulates
the outcomes of a single action or its effect in a bigger task context.
Hu et al. (contribution 9) proposed a novel circulant singular spectrum analysis embed-
ded common spatial pattern method for learning the optimal time–frequency–spatial fea-
tures to improve the motor imagery (MI) classification accuracy using EEG data. The results
confirm that it is a promising method for improving the performance of MI-based BCIs.
Li and Iramina (contribution 11) estimated dynamic functional connectivity between
the visual cortex and all the other areas of the brain to find which of them were influenced
by visual stimuli. They found that seeing manipulable objects and seeing tools caused
similar phenomena in both time and space. There is no evidence suggesting that seeing
a manipulable object led to a similar mu rhythm change to seeing an interaction with the
same object.
Cao et al. (contribution 15) introduced a sensor fusion method to evaluate cognitive
workload based on EEG and functional near-infrared spectroscopy (fNIRS). They explored
the classification performance of the features of bivariate functional brain connectivity in
the time and frequency domains of delta, theta, and alpha bands, with the assistance of the
fNIRS oxyhemoglobin and deoxyhemoglobin indicators.
Najafi et al. (contribution 1) explored the potential of diagnosing focal and generalised
epilepsy using EEG by extracting features from discrete wavelet transform and combining
them with an RNN-LSTM classifier. The results show that the theta frequency band was
more successful than alpha and beta in the detection procedure.
Alharthi et al. (contribution 3) presented another study on epileptic disorder detection
using EEG. The proposed system uses a wavelet decomposition technique and a simple
one-dimensional convolutional neural network, along with bidirectional long-short-term
memory and attention, to receive EEG signals as input data, pass them to various layers,
and finally make a decision via a dense layer. This model can assist neurophysiologists in
detecting seizures and significantly decrease the burden, while also increasing the efficiency.

Conflicts of Interest: The authors declare no conflict of interest.

List of Contributions:
1. Najafi, T.; Jaafar, R.; Remli, R.; Zaidi, W.A.W. A Classification Model of EEG Signals Based
on RNN-LSTM for Diagnosing Focal and Generalized Epilepsy. Sensors 2022, 22, 7269. https:
//doi.org/10.3390/s22197269.
2. Yang, C.Y.; Chen, P.C.; Huang, W.C. Cross-Domain Transfer of EEG to EEG or ECG Learning for
CNN Classification Models. Sensors 2023, 23, 2458. https://doi.org/10.3390/s23052458.
Sensors 2023, 23, 9056 4 of 5

3. Alharthi, M.K.; Moria, K.M.; Alghazzawi, D.M.; Tayeb, H.O. Epileptic Disorder Detection of
Seizures Using EEG Signals. Sensors 2022, 22, 6592. https://doi.org/10.3390/s22176592.
4. Shi, M.; Huang, Z.; Xiao, G.; Xu, B.; Ren, Q.; Zhao, H. Estimating the Depth of Anesthesia
from EEG Signals Based on a Deep Residual Shrinkage Network. Sensors 2023, 23, 1008.
https://doi.org/10.3390/s23021008.
5. Abdel-Hamid, L. An Efficient Machine Learning-Based Emotional Valence Recognition Ap-
proach Towards Wearable EEG. Sensors 2023, 23, 1255. https://doi.org/10.3390/s23031255.
6. Yuvaraj, R.; Thagavel, P.; Thomas, J.; Fogarty, J.; Ali, F. Comprehensive Analysis of Feature
Extraction Methods for Emotion Recognition from Multichannel EEG Recordings. Sensors 2023,
23, 915. https://doi.org/10.3390/s23020915.
7. Shah, S.M.A.; Usman, S.M.; Khalid, S.; Rehman, I.U.; Anwar, A.; Hussain, S.; Ullah, S.S.;
Elmannai, H.; Algarni, A.D.; Manzoor, W. An Ensemble Model for Consumer Emotion Pre-
diction Using EEG Signals for Neuromarketing Applications. Sensors 2022, 22, 9744. https:
//doi.org/10.3390/s22249744.
8. Borra, D.; Fantozzi, S.; Bisi, M.C.; Magosso, E. Modulations of Cortical Power and Connectivity
in Alpha and Beta Bands during the Preparation of Reaching Movements. Sensors 2023, 23, 3530.
https://doi.org/10.3390/s23073530.
9. Hu, H.; Pu, Z.; Li, H.; Liu, Z.; Wang, P. Learning Optimal Time-Frequency-Spatial Features
by the CiSSA-CSP Method for Motor Imagery EEG Classification. Sensors 2022, 22, 8526.
https://doi.org/10.3390/s22218526.
10. Jochumsen, M.; Hougaard, B.I.; Kristensen, M.S.; Knoche, H. Implementing Performance
Accommodation Mechanisms in Online BCI for Stroke Rehabilitation: A Study on Perceived
Control and Frustration. Sensors 2022, 22, 9051. https://doi.org/10.3390/s22239051.
11. Li, Z.; Iramina, K. Spatio-Temporal Neural Dynamics of Observing Non-Tool Manipulable
Objects and Interactions. Sensors 2022, 22, 7771. https://doi.org/10.3390/s22207771.
12. Mockevičius, A.; Yokota, Y.; Tarailis, P.; Hasegawa, H.; Naruse, Y.; Griškova-Bulanova, I.
Extraction of Individual EEG Gamma Frequencies from the Responses to Click-Based Chirp-
Modulated Sounds. Sensors 2023, 23, 2826. https://doi.org/10.3390/s23052826.
13. Oikonomou, V.P.; Georgiadis, K.; Kalaganis, F.; Nikolopoulos, S.; Kompatsiaris, I. A Sparse
Representation Classification Scheme for the Recognition of Affective and Cognitive Brain
Processes in Neuromarketing. Sensors 2023, 23, 2480. https://doi.org/10.3390/s23052480.
14. Alreshidi, I.; Moulitsas, I.; Jenkins, K.W. Multimodal Approach for Pilot Mental State Detection
Based on EEG. Sensors 2023, 23, 7350. https://doi.org/10.3390/s23177350.
15. Cao, J.; Garro, E.M.; Zhao, Y. EEG/fNIRS Based Workload Classification Using Functional Brain
Connectivity and Machine Learning. Sensors 2022, 22, 7623. https://doi.org/10.3390/s22197623.
16. Kim, H.; Miyakoshi, M.; Kim, Y.; Stapornchaisit, S.; Yoshimura, N.; Koike, Y. Electroencephalog-
raphy Reflects User Satisfaction in Controlling Robot Hand through Electromyographic Signals.
Sensors 2023, 23, 277. https://doi.org/10.3390/s23010277.

References
1. Pievani, M.; de Haan, W.; Wu, T.; Seeley, W.W.; Frisoni, G.B. Functional network disruption in the degenerative dementias. Lancet
Neurol. 2011, 10, 829–843. [CrossRef] [PubMed]
2. Lioi, G.; Bell, S.L.; Smith, D.C.; Simpson, D.M. Directional connectivity in the EEG is able to discriminate wakefulness from
NREM sleep. Physiol. Meas. 2017, 38, 1802–1820. [CrossRef] [PubMed]
3. Dash, G.K.; Rathore, C.; Jeyaraj, M.K.; Wattamwar, P.; Sarma, S.P.; Radhakrishnan, K. Interictal regional paroxysmal fast activity
on scalp EEG is common in patients with underlying gliosis. Clin. Neurophysiol. 2018, 129, 946–951. [CrossRef] [PubMed]
4. Renzel, R.; Baumann, C.R.; Mothersill, I.; Poryazova, R. Persistent generalized periodic discharges: A specific marker of fatal
outcome in cerebral hypoxia. Clin. Neurophysiol. 2017, 128, 147–152. [CrossRef] [PubMed]
5. Visani, E.; Varotto, G.; Binelli, S.; Fratello, L.; Franceschetti, S.; Avanzini, G.; Panzica, F. Photosensitive epilepsy: Spectral and
coherence analyses of EEG using 14 Hz intermittent photic stimulation. Clin. Neurophysiol. 2010, 121, 318–324. [CrossRef]
[PubMed]
6. Watanabe, H.; Terada, K.; Suzuki, N.; Ishisaka, M.; Naitoh, Y.; Ishihara, R.; Shimoeda, H.; Konagaya, T.; Inoue, Y. P1-3-10. Effect of
hyperventilation on seizures and EEG findings during routine EEG. Clin. Neurophysiol. 2018, 129, e38. [CrossRef]
Sensors 2023, 23, 9056 5 of 5

7. Cao, J.; Grajcar, K.; Shan, X.; Zhao, Y.; Zou, J.; Chen, L.; Li, Z.; Grunewald, R.; Zis, P.; De Marco, M.; et al. Using interictal
seizure-free EEG data to recognise patients with epilepsy based on machine learning of brain functional connectivity. Biomed.
Signal Process. Control 2021, 67, 102554. [CrossRef]
8. Hassani, M.; Karami, M. Noise estimation in electroencephalogram signal by using volterra series coefficients. J. Med. Signals
Sens. 2015, 5, 192–200. [CrossRef] [PubMed]

Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual
author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to
people or property resulting from any ideas, methods, instructions or products referred to in the content.