Sensors

Research

17 pages, 596 KiB

Open AccessArticle

User Evaluation of a Shared Robot Control System Combining BCI and Eye Tracking in a Portable Augmented Reality User Interface

by Arnau Dillen, Mohsen Omidi, Fakhreddine Ghaffari, Olivier Romain, Bram Vanderborght, Bart Roelands, Ann Nowé and Kevin De Pauw

Sensors 2024, 24(16), 5253; https://doi.org/10.3390/s24165253 - 14 Aug 2024

Viewed by 680

Abstract

This study evaluates an innovative control approach to assistive robotics by integrating brain–computer interface (BCI) technology and eye tracking into a shared control system for a mobile augmented reality user interface. Aimed at enhancing the autonomy of individuals with physical disabilities, particularly those [...] Read more.

This study evaluates an innovative control approach to assistive robotics by integrating brain–computer interface (BCI) technology and eye tracking into a shared control system for a mobile augmented reality user interface. Aimed at enhancing the autonomy of individuals with physical disabilities, particularly those with impaired motor function due to conditions such as stroke, the system utilizes BCI to interpret user intentions from electroencephalography signals and eye tracking to identify the object of focus, thus refining control commands. This integration seeks to create a more intuitive and responsive assistive robot control strategy. The real-world usability was evaluated, demonstrating significant potential to improve autonomy for individuals with severe motor impairments. The control system was compared with an eye-tracking-based alternative to identify areas needing improvement. Although BCI achieved an acceptable success rate of 0.83 in the final phase, eye tracking was more effective with a perfect success rate and consistently lower completion times (

p < 0.001

). The user experience responses favored eye tracking in 11 out of 26 questions, with no significant differences in the remaining questions, and subjective fatigue was higher with BCI use (

p = 0.04

). While BCI performance lagged behind eye tracking, the user evaluation supports the validity of our control strategy, showing that it could be deployed in real-world conditions and suggesting a pathway for further advancements. Full article

(This article belongs to the Special Issue EEG Signal Processing Techniques and Applications—2nd Edition)

► Show Figures

Figure 1

Figure 1
The control strategy of the MI BCI control system. The user selects an object with their gaze and uses MI to select one of the possible actions. After accepting or rejecting the decoded MI class, the robot executes the associated action or returns to the action selection stage. Full article ">Figure 2
The experimental procedure that was followed for Phase1 in (a) session 1 and (b) session 2. Full article ">Figure 3
The experimental procedure that was followed for Phase 3 in (a) session 1, (b) session 2, and (c) session 3 with options A and B. Full article ">Figure 4
(a) Success rate for individual participants together with the mean for each phase and (b) boxplots comparing the mean completion times between the eye tracking and BCI control system variants for each task. Full article ">Figure 5
UEQ questions where a significant difference was found between the participant’s answers for the eye tracking and BCI variants. A score of 1 indicates that users felt more that the top term of the label was applicable, while 7 means that the bottom term was more applicable. Full article ">Figure 6
mVAS scores at the beginning of each session and the end for the first two sessions and for sessions 3 before, between the first and second evaluation rounds, and at the end of the session, split by the control system that was used. Full article ">

19 pages, 6459 KiB

Open AccessArticle

Detection of Pilots’ Psychological Workload during Turning Phases Using EEG Characteristics

by Li Ji, Leiye Yi, Haiwei Li, Wenjie Han and Ningning Zhang

Sensors 2024, 24(16), 5176; https://doi.org/10.3390/s24165176 - 10 Aug 2024

Viewed by 926

Abstract

Pilot behavior is crucial for aviation safety. This study aims to investigate the EEG characteristics of pilots, refine training assessment methodologies, and bolster flight safety measures. The collected EEG signals underwent initial preprocessing. The EEG characteristic analysis was performed during left and right [...] Read more.

Pilot behavior is crucial for aviation safety. This study aims to investigate the EEG characteristics of pilots, refine training assessment methodologies, and bolster flight safety measures. The collected EEG signals underwent initial preprocessing. The EEG characteristic analysis was performed during left and right turns, involving the calculation of the energy ratio of beta waves and Shannon entropy. The psychological workload of pilots during different flight phases was quantified as well. Based on the EEG characteristics, the pilots’ psychological workload was classified through the use of a support vector machine (SVM). The study results showed significant changes in the energy ratio of beta waves and Shannon entropy during left and right turns compared to the cruising phase. Additionally, the pilots’ psychological workload was found to have increased during these turning phases. Using support vector machines to detect the pilots’ psychological workload, the classification accuracy for the training set was 98.92%, while for the test set, it was 93.67%. This research holds significant importance in understanding pilots’ psychological workload. Full article

(This article belongs to the Special Issue EEG Signal Processing Techniques and Applications—2nd Edition)

► Show Figures

Figure 1

22 pages, 6002 KiB

Open AccessArticle

Latent Prototype-Based Clustering: A Novel Exploratory Electroencephalography Analysis Approach

by Sun Zhou, Pengyi Zhang and Huazhen Chen

Sensors 2024, 24(15), 4920; https://doi.org/10.3390/s24154920 - 29 Jul 2024

Viewed by 570

Abstract

Electroencephalography (EEG)-based applications in brain–computer interfaces (BCIs), neurological disease diagnosis, rehabilitation, etc., rely on supervised approaches such as classification that requires given labels. However, with the ever-increasing amount of EEG data, incomplete or incorrectly labeled or unlabeled EEG data are increasing. It likely [...] Read more.

Electroencephalography (EEG)-based applications in brain–computer interfaces (BCIs), neurological disease diagnosis, rehabilitation, etc., rely on supervised approaches such as classification that requires given labels. However, with the ever-increasing amount of EEG data, incomplete or incorrectly labeled or unlabeled EEG data are increasing. It likely degrades the performance of supervised approaches. In this work, we put forward a novel unsupervised exploratory EEG analysis solution by clustering based on low-dimensional prototypes in latent space that are associated with the respective clusters. Having the prototype as a baseline of each cluster, a compositive similarity is defined to act as the critic function in clustering, which incorporates similarities on three levels. The approach is implemented with a Generative Adversarial Network (GAN), termed W-SLOGAN, by extending the Stein Latent Optimization for GANs (SLOGAN). The Gaussian Mixture Model (GMM) is utilized as the latent distribution to adapt to the diversity of EEG signal patterns. The W-SLOGAN ensures that images generated from each Gaussian component belong to the associated cluster. The adaptively learned Gaussian mixing coefficients make the model remain effective in dealing with an imbalanced dataset. By applying the proposed approach to two public EEG or intracranial EEG (iEEG) epilepsy datasets, our experiments demonstrate that the clustering results are close to the classification of the data. Moreover, we present several findings that were discovered by intra-class clustering and cross-analysis of clustering and classification. They show that the approach is attractive in practice in the diagnosis of the epileptic subtype, multiple labelling of EEG data, etc. Full article

(This article belongs to the Special Issue EEG Signal Processing Techniques and Applications—2nd Edition)

► Show Figures

Figure 1

19 pages, 2917 KiB

Open AccessArticle

An Innovative EEG-Based Pain Identification and Quantification: A Pilot Study

by Colince Meli Segning, Rubens A. da Silva and Suzy Ngomo

Sensors 2024, 24(12), 3873; https://doi.org/10.3390/s24123873 - 14 Jun 2024

Viewed by 874

Abstract

Objective: The present pilot study aimed to propose an innovative scale-independent measure based on electroencephalographic (EEG) signals for the identification and quantification of the magnitude of chronic pain. Methods: EEG data were collected from three groups of participants at rest: seven healthy participants [...] Read more.

Objective: The present pilot study aimed to propose an innovative scale-independent measure based on electroencephalographic (EEG) signals for the identification and quantification of the magnitude of chronic pain. Methods: EEG data were collected from three groups of participants at rest: seven healthy participants with pain, 15 healthy participants submitted to thermal pain, and 66 participants living with chronic pain. Every 30 s, the pain intensity score felt by the participant was also recorded. Electrodes positioned in the contralateral motor region were of interest. After EEG preprocessing, a complex analytical signal was obtained using Hilbert transform, and the upper envelope of the EEG signal was extracted. The average coefficient of variation of the upper envelope of the signal was then calculated for the beta (13–30 Hz) band and proposed as a new EEG-based indicator, namely Piq_β, to identify and quantify pain. Main results: The main results are as follows: (1) A Piq_β threshold at 10%, that is, Piq_β ≥ 10%, indicates the presence of pain, and (2) the higher the Piq_β (%), the higher the extent of pain. Conclusions: This finding indicates that Piq_β can objectively identify and quantify pain in a population living with chronic pain. This new EEG-based indicator can be used for objective pain assessment based on the neurophysiological body response to pain. Significance: Objective pain assessment is a valuable decision-making aid and an important contribution to pain management and monitoring. Full article

(This article belongs to the Special Issue EEG Signal Processing Techniques and Applications—2nd Edition)

► Show Figures

Figure 1

Figure 1
The thermal stimulus kit. Full article ">Figure 2
Experimental design—Thermal stimulus, pain score, and EEG recording during 300 s or 5 min. Full article ">Figure 3
Methodological approach from the filtering of the EEG signal, through the estimation of the coefficient of variation of the upper envelope in beta (CVUEβ), to the calculation of pain identification and quantification (Piqβ). Full article ">Figure 4
Methodological steps showing the detail of the application of the Hilbert transform until the extraction of the upper envelope. (a)—original real-valued signal, (b)—real and imaginary parts of analytic signal, (c)—superposition of real and imaginary parts of analytic signal and upper envelope of original signal. Full article ">Figure 5
Normalized mean [0–1] for all participants (n = 15) of the three variables: (1) Normalized pain score intensity (black dotted line curve), (2) normalized level of pain stimulus (grid curve), and (3) normalized pain identification and quantification in beta frequency band (Piqβ) (black curve). Full article ">Figure 6
Scatter plot—Piqβ indicator and pain score. 100% of participants living with chronic pain show a Piqβ ≥ 10%. The two solid points represent participants who reported a pain score lower than 1/10 but had a Piqβ indicator ≥10%. The hollow points represent participants whose pain scores are consistent with their Piqβ indicator values. Full article ">

17 pages, 5820 KiB

Open AccessArticle

Detection Method of Epileptic Seizures Using a Neural Network Model Based on Multimodal Dual-Stream Networks

by Baiyang Wang, Yidong Xu, Siyu Peng, Hongjun Wang and Fang Li

Sensors 2024, 24(11), 3360; https://doi.org/10.3390/s24113360 - 24 May 2024

Cited by 1 | Viewed by 818

Abstract

Epilepsy is a common neurological disorder, and its diagnosis mainly relies on the analysis of electroencephalogram (EEG) signals. However, the raw EEG signals contain limited recognizable features, and in order to increase the recognizable features in the input of the network, the differential [...] Read more.

Epilepsy is a common neurological disorder, and its diagnosis mainly relies on the analysis of electroencephalogram (EEG) signals. However, the raw EEG signals contain limited recognizable features, and in order to increase the recognizable features in the input of the network, the differential features of the signals, the amplitude spectrum and the phase spectrum in the frequency domain are extracted to form a two-dimensional feature vector. In order to solve the problem of recognizing multimodal features, a neural network model based on a multimodal dual-stream network is proposed, which uses a mixture of one-dimensional convolution, two-dimensional convolution and LSTM neural networks to extract the spatial features of the EEG two-dimensional vectors and the temporal features of the signals, respectively, and combines the advantages of the two networks, using the hybrid neural network to extract both the temporal and spatial features of the signals at the same time. In addition, a channel attention module was used to focus the model on features related to seizures. Finally, multiple sets of experiments were conducted on the Bonn and New Delhi data sets, and the highest accuracy rates of 99.69% and 97.5% were obtained on the test set, respectively, verifying the superiority of the proposed model in the task of epileptic seizure detection. Full article

(This article belongs to the Special Issue EEG Signal Processing Techniques and Applications—2nd Edition)

► Show Figures

Figure 1

29 pages, 5473 KiB

Open AccessArticle

Optimal Channel Selection of Multiclass Motor Imagery Classification Based on Fusion Convolutional Neural Network with Attention Blocks

by Joharah Khabti, Saad AlAhmadi and Adel Soudani

Sensors 2024, 24(10), 3168; https://doi.org/10.3390/s24103168 - 16 May 2024

Viewed by 820

Abstract

The widely adopted paradigm in brain–computer interfaces (BCIs) involves motor imagery (MI), enabling improved communication between humans and machines. EEG signals derived from MI present several challenges due to their inherent characteristics, which lead to a complex process of classifying and finding the [...] Read more.

The widely adopted paradigm in brain–computer interfaces (BCIs) involves motor imagery (MI), enabling improved communication between humans and machines. EEG signals derived from MI present several challenges due to their inherent characteristics, which lead to a complex process of classifying and finding the potential tasks of a specific participant. Another issue is that BCI systems can result in noisy data and redundant channels, which in turn can lead to increased equipment and computational costs. To address these problems, the optimal channel selection of a multiclass MI classification based on a Fusion convolutional neural network with Attention blocks (FCNNA) is proposed. In this study, we developed a CNN model consisting of layers of convolutional blocks with multiple spatial and temporal filters. These filters are designed specifically to capture the distribution and relationships of signal features across different electrode locations, as well as to analyze the evolution of these features over time. Following these layers, a Convolutional Block Attention Module (CBAM) is used to, further, enhance EEG signal feature extraction. In the process of channel selection, the genetic algorithm is used to select the optimal set of channels using a new technique to deliver fixed as well as variable channels for all participants. The proposed methodology is validated showing 6.41% improvement in multiclass classification compared to most baseline models. Notably, we achieved the highest results of 93.09% for binary classes involving left-hand and right-hand movements. In addition, the cross-subject strategy for multiclass classification yielded an impressive accuracy of 68.87%. Following channel selection, multiclass classification accuracy was enhanced, reaching 84.53%. Overall, our experiments illustrated the efficiency of the proposed EEG MI model in both channel selection and classification, showing superior results with either a full channel set or a reduced number of channels. Full article

(This article belongs to the Special Issue EEG Signal Processing Techniques and Applications—2nd Edition)

► Show Figures

Figure 1

20 pages, 2552 KiB

Open AccessArticle

Identifying the Effect of Cognitive Motivation with the Method Based on Temporal Association Rule Mining Concept

by Tustanah Phukhachee, Suthathip Maneewongvatana, Chayapol Chaiyanan, Keiji Iramina and Boonserm Kaewkamnerdpong

Sensors 2024, 24(9), 2857; https://doi.org/10.3390/s24092857 - 30 Apr 2024

Cited by 1 | Viewed by 840

Abstract

Being motivated has positive influences on task performance. However, motivation could result from various motives that affect different parts of the brain. Analyzing the motivation effect from all affected areas requires a high number of EEG electrodes, resulting in high cost, inflexibility, and [...] Read more.

Being motivated has positive influences on task performance. However, motivation could result from various motives that affect different parts of the brain. Analyzing the motivation effect from all affected areas requires a high number of EEG electrodes, resulting in high cost, inflexibility, and burden to users. In various real-world applications, only the motivation effect is required for performance evaluation regardless of the motive. Analyzing the relationships between the motivation-affected brain areas associated with the task’s performance could limit the required electrodes. This study introduced a method to identify the cognitive motivation effect with a reduced number of EEG electrodes. The temporal association rule mining (TARM) concept was used to analyze the relationships between attention and memorization brain areas under the effect of motivation from the cognitive motivation task. For accuracy improvement, the artificial bee colony (ABC) algorithm was applied with the central limit theorem (CLT) concept to optimize the TARM parameters. From the results, our method can identify the motivation effect with only FCz and P3 electrodes, with 74.5% classification accuracy on average with individual tests. Full article

(This article belongs to the Special Issue EEG Signal Processing Techniques and Applications—2nd Edition)

► Show Figures

Figure 1

21 pages, 840 KiB

Open AccessArticle

Using Explainable Artificial Intelligence to Obtain Efficient Seizure-Detection Models Based on Electroencephalography Signals

by Jusciaane Chacon Vieira, Luiz Affonso Guedes, Mailson Ribeiro Santos and Ignacio Sanchez-Gendriz

Sensors 2023, 23(24), 9871; https://doi.org/10.3390/s23249871 - 16 Dec 2023

Cited by 2 | Viewed by 1519

Abstract

Epilepsy is a condition that affects 50 million individuals globally, significantly impacting their quality of life. Epileptic seizures, a transient occurrence, are characterized by a spectrum of manifestations, including alterations in motor function and consciousness. These events impose restrictions on the daily lives [...] Read more.

Epilepsy is a condition that affects 50 million individuals globally, significantly impacting their quality of life. Epileptic seizures, a transient occurrence, are characterized by a spectrum of manifestations, including alterations in motor function and consciousness. These events impose restrictions on the daily lives of those affected, frequently resulting in social isolation and psychological distress. In response, numerous efforts have been directed towards the detection and prevention of epileptic seizures through EEG signal analysis, employing machine learning and deep learning methodologies. This study presents a methodology that reduces the number of features and channels required by simpler classifiers, leveraging Explainable Artificial Intelligence (XAI) for the detection of epileptic seizures. The proposed approach achieves performance metrics exceeding 95% in accuracy, precision, recall, and F1-score by utilizing merely six features and five channels in a temporal domain analysis, with a time window of 1 s. The model demonstrates robust generalization across the patient cohort included in the database, suggesting that feature reduction in simpler models—without resorting to deep learning—is adequate for seizure detection. The research underscores the potential for substantial reductions in the number of attributes and channels, advocating for the training of models with strategically selected electrodes, and thereby supporting the development of effective mobile applications for epileptic seizure detection. Full article

(This article belongs to the Special Issue EEG Signal Processing Techniques and Applications—2nd Edition)

► Show Figures

Figure 1

20 pages, 2917 KiB

Open AccessArticle

Illuminating the Neural Landscape of Pilot Mental States: A Convolutional Neural Network Approach with Shapley Additive Explanations Interpretability

by Ibrahim Alreshidi, Desmond Bisandu and Irene Moulitsas

Sensors 2023, 23(22), 9052; https://doi.org/10.3390/s23229052 - 8 Nov 2023

Cited by 2 | Viewed by 1211

Abstract

Predicting pilots’ mental states is a critical challenge in aviation safety and performance, with electroencephalogram data offering a promising avenue for detection. However, the interpretability of machine learning and deep learning models, which are often used for such tasks, remains a significant issue. [...] Read more.

Predicting pilots’ mental states is a critical challenge in aviation safety and performance, with electroencephalogram data offering a promising avenue for detection. However, the interpretability of machine learning and deep learning models, which are often used for such tasks, remains a significant issue. This study aims to address these challenges by developing an interpretable model to detect four mental states—channelised attention, diverted attention, startle/surprise, and normal state—in pilots using EEG data. The methodology involves training a convolutional neural network on power spectral density features of EEG data from 17 pilots. The model’s interpretability is enhanced via the use of SHapley Additive exPlanations values, which identify the top 10 most influential features for each mental state. The results demonstrate high performance in all metrics, with an average accuracy of 96%, a precision of 96%, a recall of 94%, and an F1 score of 95%. An examination of the effects of mental states on EEG frequency bands further elucidates the neural mechanisms underlying these states. The innovative nature of this study lies in its combination of high-performance model development, improved interpretability, and in-depth analysis of the neural correlates of mental states. This approach not only addresses the critical need for effective and interpretable mental state detection in aviation but also contributes to our understanding of the neural underpinnings of these states. This study thus represents a significant advancement in the field of EEG-based mental state detection. Full article

(This article belongs to the Special Issue EEG Signal Processing Techniques and Applications—2nd Edition)

► Show Figures

Figure 1

16 pages, 519 KiB

Open AccessFeature PaperArticle

Domain-Specific Processing Stage for Estimating Single-Trail Evoked Potential Improves CNN Performance in Detecting Error Potential

by Andrea Farabbi and Luca Mainardi

Sensors 2023, 23(22), 9049; https://doi.org/10.3390/s23229049 - 8 Nov 2023

Cited by 1 | Viewed by 942

Abstract

We present a novel architecture designed to enhance the detection of Error Potential (ErrP) signals during ErrP stimulation tasks. In the context of predicting ErrP presence, conventional Convolutional Neural Networks (CNNs) typically accept a raw EEG signal as input, encompassing both the information [...] Read more.

We present a novel architecture designed to enhance the detection of Error Potential (ErrP) signals during ErrP stimulation tasks. In the context of predicting ErrP presence, conventional Convolutional Neural Networks (CNNs) typically accept a raw EEG signal as input, encompassing both the information associated with the evoked potential and the background activity, which can potentially diminish predictive accuracy. Our approach involves advanced Single-Trial (ST) ErrP enhancement techniques for processing raw EEG signals in the initial stage, followed by CNNs for discerning between ErrP and NonErrP segments in the second stage. We tested different combinations of methods and CNNs. As far as ST ErrP estimation is concerned, we examined various methods encompassing subspace regularization techniques, Continuous Wavelet Transform, and ARX models. For the classification stage, we evaluated the performance of EEGNet, CNN, and a Siamese Neural Network. A comparative analysis against the method of directly applying CNNs to raw EEG signals revealed the advantages of our architecture. Leveraging subspace regularization yielded the best improvement in classification metrics, at up to 14% in balanced accuracy and 13.4% in F1-score. Full article

(This article belongs to the Special Issue EEG Signal Processing Techniques and Applications—2nd Edition)

► Show Figures

Figure 1

Figure 1
Pipeline of the employed methods. The novelty of our approach consists in the introduction of a Single Trial Extraction block between the EEG preprocessing stage and the NN classification stage. Full article ">Figure 2
Performance metrics on the test set when using EEGNet as a classifier. The metrics are reported for each processing technique and in both subject-wise and population-wise analysis. The balanced accuracy and F1-score are reported for completeness. Full article ">Figure 3
Utility Gain when using EEGNet as a classifier and applying different techniques for ST estimation. The metric is reported for the subject-wise and population-wise analysis for only p values over <math display="inline"><semantics> <mrow> <mn>70</mn> <mo>%</mo> </mrow> </semantics></math>, as no major differences were observed outside of this range. Full article ">Figure 4
Performance metrics on the test set when using L-CNN as a classifier. The metrics are reported for each processing technique and in both the subject-wise and population-wise analysis. Balanced accuracy and F1-score are reported for completeness. Full article ">Figure 5
Utility gain obtained using L-CNN as a classifier and applying different ST estimation techniques. The metric is reported for subject-wise and population-wise analysis and for p values over <math display="inline"><semantics> <mrow> <mn>70</mn> <mo>%</mo> </mrow> </semantics></math>, as no major differences were observed below this level. Full article ">Figure 6
Performance metrics on the test set when using a Siamese Neural Network as the classifier. The metrics are reported for each processing technique and for subject-wise and population-wise analysis. Balanced accuracy and F1-score are reported for completeness. Full article ">Figure 7
Utility gain obtained using a Siamese Neural Network as the classifier and applying different ST estimation techniques. The metric is reported for the subject-wise and population-wise analysis and for only those p values over <math display="inline"><semantics> <mrow> <mn>70</mn> <mo>%</mo> </mrow> </semantics></math>, as no major differences were observed below this level. Full article ">

21 pages, 2109 KiB

Open AccessArticle

Graph Analysis of TMS–EEG Connectivity Reveals Hemispheric Differences following Occipital Stimulation

by Ilaria Siviero, Davide Bonfanti, Gloria Menegaz, Silvia Savazzi, Chiara Mazzi and Silvia Francesca Storti

Sensors 2023, 23(21), 8833; https://doi.org/10.3390/s23218833 - 30 Oct 2023

Cited by 2 | Viewed by 1953

Abstract

(1) Background: Transcranial magnetic stimulation combined with electroencephalography (TMS–EEG) provides a unique opportunity to investigate brain connectivity. However, possible hemispheric asymmetries in signal propagation dynamics following occipital TMS have not been investigated. (2) Methods: Eighteen healthy participants underwent occipital single-pulse TMS at two [...] Read more.

(1) Background: Transcranial magnetic stimulation combined with electroencephalography (TMS–EEG) provides a unique opportunity to investigate brain connectivity. However, possible hemispheric asymmetries in signal propagation dynamics following occipital TMS have not been investigated. (2) Methods: Eighteen healthy participants underwent occipital single-pulse TMS at two different EEG sites, corresponding to early visual areas. We used a state-of-the-art Bayesian estimation approach to accurately estimate TMS-evoked potentials (TEPs) from EEG data, which has not been previously used in this context. To capture the rapid dynamics of information flow patterns, we implemented a self-tuning optimized Kalman (STOK) filter in conjunction with the information partial directed coherence (iPDC) measure, enabling us to derive time-varying connectivity matrices. Subsequently, graph analysis was conducted to assess key network properties, providing insight into the overall network organization of the brain network. (3) Results: Our findings revealed distinct lateralized effects on effective brain connectivity and graph networks after TMS stimulation, with left stimulation facilitating enhanced communication between contralateral frontal regions and right stimulation promoting increased intra-hemispheric ipsilateral connectivity, as evidenced by statistical test (p < 0.001). (4) Conclusions: The identified hemispheric differences in terms of connectivity provide novel insights into brain networks involved in visual information processing, revealing the hemispheric specificity of neural responses to occipital stimulation. Full article

(This article belongs to the Special Issue EEG Signal Processing Techniques and Applications—2nd Edition)

► Show Figures

Figure 1

Figure 1
(a) Electrode placement and schematic representation of the experimental procedure: (i) random interval (700–1000 ms), (ii) single-pulse TMS stimulation, (iii) phosphene awareness assessment (up to 2000 ms), and (iv) inter-trial interval (1300 ms). (b) Signal processing pipeline for assessing the differences between hemispheres after TMS stimulation. The proposed pipeline is divided into preprocessing, TMS-evoked potential estimation, the time-varying effective connectivity calculation, and the graph network. Full article ">Figure 2
Global mean field power calculated for TEP using the conventional averaging (CA) method (blue) and the Bayesian smoothing approach (red) obtained from 360 sweeps. The upper part referred to the left TMS stimulation and the lower part to the right TMS stimulation. Statistical significance differences between the CA method and the Bayesian smoothing are indicated by gray areas (Wilcoxon signed-rank test Bonferroni corrected, <math display="inline"><semantics> <mrow> <mi>p</mi> <mo><</mo> </mrow> </semantics></math> 0.05). Time is measured in ms and the amplitude is in <math display="inline"><semantics> <mi mathvariant="sans-serif">μ</mi> </semantics></math>V. Full article ">Figure 3
Time-varying brain connectivity analysis for stimulation sites. (a) TMS on O1; (b) TMS on O2; (c) statistical significance differences between conditions are indicated by arrows (Wilcoxon signed-rank test uncorrected <math display="inline"><semantics> <mrow> <mi>p</mi> <mo><</mo> </mrow> </semantics></math> 0.001). Red arrows indicate stronger connections following right stimulation (O2) than left stimulation (O1). By contrast, blue arrows indicate stronger connections following left stimulation (O1) than right stimulation (O2). Full article ">Figure 4
Time-varying edges betweenness centrality for stimulation sites. (a) Left (O1) TMS stimulation; (b) right (O2) TMS stimulation; (c) statistical significance differences between conditions are indicated by arrows (Wilcoxon signed-rank test uncorrected <math display="inline"><semantics> <mrow> <mi>p</mi> <mo><</mo> </mrow> </semantics></math> 0.001). Red arrows indicate stronger values of the edges following right stimulation (O2) than left stimulation (O1). By contrast, blue arrows indicate stronger values of the edge following left stimulation (O1) than right stimulation (O2). Full article ">Figure 5
(a) Electrode placement using the international 10-10 system covering the ipsilateral frontal channels (highlighted in blue), contralateral frontal channels (highlighted in green), ipsilateral occipital channels (highlighted in yellow), and contralateral occipital channels (highlighted in orange). (b) Degree metric of graph networks in response to left vs. right TMS. Statistical significant differences between conditions are indicated by asterisks above boxplots (Wilcoxon signed-rank test uncorrected <math display="inline"><semantics> <mrow> <mi>p</mi> <mo><</mo> </mrow> </semantics></math> 0.001). Full article ">

18 pages, 2099 KiB

Open AccessArticle

Depressive Disorder Recognition Based on Frontal EEG Signals and Deep Learning

by Yanting Xu, Hongyang Zhong, Shangyan Ying, Wei Liu, Guibin Chen, Xiaodong Luo and Gang Li

Sensors 2023, 23(20), 8639; https://doi.org/10.3390/s23208639 - 23 Oct 2023

Cited by 4 | Viewed by 2142

Abstract

Depressive disorder (DD) has become one of the most common mental diseases, seriously endangering both the affected person’s psychological and physical health. Nowadays, a DD diagnosis mainly relies on the experience of clinical psychiatrists and subjective scales, lacking objective, accurate, practical, and automatic [...] Read more.

Depressive disorder (DD) has become one of the most common mental diseases, seriously endangering both the affected person’s psychological and physical health. Nowadays, a DD diagnosis mainly relies on the experience of clinical psychiatrists and subjective scales, lacking objective, accurate, practical, and automatic diagnosis technologies. Recently, electroencephalogram (EEG) signals have been widely applied for DD diagnosis, but mainly with high-density EEG, which can severely limit the efficiency of the EEG data acquisition and reduce the practicability of diagnostic techniques. The current study attempts to achieve accurate and practical DD diagnoses based on combining frontal six-channel electroencephalogram (EEG) signals and deep learning models. To this end, 10 min clinical resting-state EEG signals were collected from 41 DD patients and 34 healthy controls (HCs). Two deep learning models, multi-resolution convolutional neural network (MRCNN) combined with long short-term memory (LSTM) (named MRCNN-LSTM) and MRCNN combined with residual squeeze and excitation (RSE) (named MRCNN-RSE), were proposed for DD recognition. The results of this study showed that the higher EEG frequency band obtained the better classification performance for DD diagnosis. The MRCNN-RSE model achieved the highest classification accuracy of 98.48 ± 0.22% with 8–30 Hz EEG signals. These findings indicated that the proposed analytical framework can provide an accurate and practical strategy for DD diagnosis, as well as essential theoretical and technical support for the treatment and efficacy evaluation of DD. Full article

(This article belongs to the Special Issue EEG Signal Processing Techniques and Applications—2nd Edition)

► Show Figures

Figure 1

16 pages, 4776 KiB

Open AccessArticle

Characterisation of Cognitive Load Using Machine Learning Classifiers of Electroencephalogram Data

by Qi Wang, Daniel Smythe, Jun Cao, Zhilin Hu, Karl J. Proctor, Andrew P. Owens and Yifan Zhao

Sensors 2023, 23(20), 8528; https://doi.org/10.3390/s23208528 - 17 Oct 2023

Cited by 4 | Viewed by 1978

Abstract

A high cognitive load can overload a person, potentially resulting in catastrophic accidents. It is therefore important to ensure the level of cognitive load associated with safety-critical tasks (such as driving a vehicle) remains manageable for drivers, enabling them to respond appropriately to [...] Read more.

A high cognitive load can overload a person, potentially resulting in catastrophic accidents. It is therefore important to ensure the level of cognitive load associated with safety-critical tasks (such as driving a vehicle) remains manageable for drivers, enabling them to respond appropriately to changes in the driving environment. Although electroencephalography (EEG) has attracted significant interest in cognitive load research, few studies have used EEG to investigate cognitive load in the context of driving. This paper presents a feasibility study on the simulation of various levels of cognitive load through designing and implementing four driving tasks. We employ machine learning-based classification techniques using EEG recordings to differentiate driving conditions. An EEG dataset containing these four driving tasks from a group of 20 participants was collected to investigate whether EEG can be used as an indicator of changes in cognitive load. The collected dataset was used to train four Deep Neural Networks and four Support Vector Machine classification models. The results showed that the best model achieved a classification accuracy of 90.37%, utilising statistical features from multiple frequency bands in 24 EEG channels. Furthermore, the Gamma and Beta bands achieved higher classification accuracy than the Alpha and Theta bands during the analysis. The outcomes of this study have the potential to enhance the Human–Machine Interface of vehicles, contributing to improved safety. Full article

(This article belongs to the Special Issue EEG Signal Processing Techniques and Applications—2nd Edition)

► Show Figures

Figure 1

Figure 1
A block diagram displaying the four steps of the overall methodology. Full article ">Figure 2
Images displaying the experimental setup and a map showing routes followed by participants. (a) Complete sets of sensors and their locations. (b) Driver simulator setup. (c) Display of the EEG sensor device. (d) Route map of the experiment; the red route shows the motorway route, and the blue route shows the urban route. Full article ">Figure 3
The cognitive load of driving tasks rated by participants. Full article ">Figure 4
The architecture of the DNN model. Full article ">Figure 5
Classification matrices using DNN for 10-folder cross-validation. (a) Classification matrices of High-Low two-class classification in urban environments. (b) Classification matrices of High-Low two-class classification in motorway environments. (c) Classification matrices of High-Low two-class classification combined. (d) Classification matrices of four-class classification. Full article ">Figure 6
Topographic maps of the accuracy values in each model. Full article ">Figure 7
Topographic maps displaying median means of the Theta, Alpha, Beta, and Gamma bands across each section of experiment. (a) Theta band. (b) Alpha band. (c) Beta band. (d) Gamma band. Full article ">Figure 8
Topographic maps displaying the variance in Theta, Alpha, Beta, and Gamma bands across each experiment section. (a) Theta band. (b) Alpha band. (c) Beta band. (d) Gamma band. Full article ">

13 pages, 2045 KiB

Open AccessArticle

Electrocortical Dynamics of Usual Walking and the Planning to Step over Obstacles in Parkinson’s Disease

by Rodrigo Vitório, Ellen Lirani-Silva, Diego Orcioli-Silva, Victor Spiandor Beretta, Anderson Souza Oliveira and Lilian Teresa Bucken Gobbi

Sensors 2023, 23(10), 4866; https://doi.org/10.3390/s23104866 - 18 May 2023

Cited by 2 | Viewed by 2396

Abstract

The neural correlates of locomotion impairments observed in people with Parkinson’s disease (PD) are not fully understood. We investigated whether people with PD present distinct brain electrocortical activity during usual walking and the approach phase of obstacle avoidance when compared to healthy individuals. [...] Read more.

The neural correlates of locomotion impairments observed in people with Parkinson’s disease (PD) are not fully understood. We investigated whether people with PD present distinct brain electrocortical activity during usual walking and the approach phase of obstacle avoidance when compared to healthy individuals. Fifteen people with PD and fourteen older adults walked overground in two conditions: usual walking and obstacle crossing. Scalp electroencephalography (EEG) was recorded using a mobile 64-channel EEG system. Independent components were clustered using a k-means clustering algorithm. Outcome measures included absolute power in several frequency bands and alpha/beta ratio. During the usual walk, people with PD presented a greater alpha/beta ratio in the left sensorimotor cortex than healthy individuals. While approaching obstacles, both groups reduced alpha and beta power in the premotor and right sensorimotor cortices (balance demand) and increased gamma power in the primary visual cortex (visual demand). Only people with PD reduced alpha power and alpha/beta ratio in the left sensorimotor cortex when approaching obstacles. These findings suggest that PD affects the cortical control of usual walking, leading to a greater proportion of low-frequency (alpha) neuronal firing in the sensorimotor cortex. Moreover, the planning for obstacle avoidance changes the electrocortical dynamics associated with increased balance and visual demands. People with PD rely on increased sensorimotor integration to modulate locomotion. Full article

(This article belongs to the Special Issue EEG Signal Processing Techniques and Applications—2nd Edition)

► Show Figures

Figure 1

Figure 1
Top: Topographic plots for the four clusters identified. Bottom: Bar graphs show alpha/beta ratio for people with PD (green) and healthy older adults (blue) in the usual walking and obstacle condition. Circles represent individual values. a indicates significant differences between usual walking and obstacle condition for people with PD; b indicates significant difference between people with PD and healthy older adults. Full article ">Figure 2
Top: Electrocortical clusters of independent components plotted on the MNI brain. Blue spheres represent independent components of healthy older adults and green spheres represent independent components of people with PD. Red spheres represent the centroid locations for the clusters ((A)—left sensorimotor cortex; (B)—right sensorimotor cortex). Bottom: Bar graphs show absolute power for theta, alpha, beta, and gamma bands in the usual walking and obstacle condition. Circles within the graphs represent individual values. a indicates significant difference between usual walking and obstacle condition for people with PD; c indicates main effect of condition. Full article ">Figure 3
Top: Electrocortical clusters of independent components plotted on the MNI brain. Blue spheres represent independent components of healthy older adults and green spheres represent independent components of people with PD. Red spheres represent the centroid locations for the clusters ((A)—middle premotor and supplementary motor area; (B)—visual cortex). Bottom: Bar graphs show absolute power for theta, alpha, beta, and gamma bands in the usual walking and obstacle condition. Circles within the graphs represent individual values. c indicates main effect of condition. Full article ">

29 pages, 2248 KiB

Open AccessArticle

Method for Automatic Estimation of Instantaneous Frequency and Group Delay in Time–Frequency Distributions with Application in EEG Seizure Signals Analysis

by Vedran Jurdana, Miroslav Vrankic, Nikola Lopac and Guruprasad Madhale Jadav

Sensors 2023, 23(10), 4680; https://doi.org/10.3390/s23104680 - 11 May 2023

Cited by 3 | Viewed by 1914

Abstract

Instantaneous frequency (IF) is commonly used in the analysis of electroencephalogram (EEG) signals to detect oscillatory-type seizures. However, IF cannot be used to analyze seizures that appear as spikes. In this paper, we present a novel method for the automatic estimation of IF [...] Read more.

Instantaneous frequency (IF) is commonly used in the analysis of electroencephalogram (EEG) signals to detect oscillatory-type seizures. However, IF cannot be used to analyze seizures that appear as spikes. In this paper, we present a novel method for the automatic estimation of IF and group delay (GD) in order to detect seizures with both spike and oscillatory characteristics. Unlike previous methods that use IF alone, the proposed method utilizes information obtained from localized Rényi entropies (LREs) to generate a binary map that automatically identifies regions requiring a different estimation strategy. The method combines IF estimation algorithms for multicomponent signals with time and frequency support information to improve signal ridge estimation in the time–frequency distribution (TFD). Our experimental results indicate the superiority of the proposed combined IF and GD estimation approach over the IF estimation alone, without requiring any prior knowledge about the input signal. The LRE-based mean squared error and mean absolute error metrics showed improvements of up to 95.70% and 86.79%, respectively, for synthetic signals and up to 46.45% and 36.61% for real-life EEG seizure signals. Full article

(This article belongs to the Special Issue EEG Signal Processing Techniques and Applications—2nd Edition)

► Show Figures

Figure 1

20 pages, 1200 KiB

Open AccessArticle

Motor Imagery Multi-Tasks Classification for BCIs Using the NVIDIA Jetson TX2 Board and the EEGNet Network

by Tat’y Mwata-Velu, Edson Niyonsaba-Sebigunda, Juan Gabriel Avina-Cervantes, Jose Ruiz-Pinales, Narcisse Velu-A-Gulenga and Adán Antonio Alonso-Ramírez

Sensors 2023, 23(8), 4164; https://doi.org/10.3390/s23084164 - 21 Apr 2023

Cited by 2 | Viewed by 2570

Abstract

Nowadays, Brain–Computer Interfaces (BCIs) still captivate large interest because of multiple advantages offered in numerous domains, explicitly assisting people with motor disabilities in communicating with the surrounding environment. However, challenges of portability, instantaneous processing time, and accurate data processing remain for numerous BCI [...] Read more.

Nowadays, Brain–Computer Interfaces (BCIs) still captivate large interest because of multiple advantages offered in numerous domains, explicitly assisting people with motor disabilities in communicating with the surrounding environment. However, challenges of portability, instantaneous processing time, and accurate data processing remain for numerous BCI system setups. This work implements an embedded multi-tasks classifier based on motor imagery using the EEGNet network integrated into the NVIDIA Jetson TX2 card. Therefore, two strategies are developed to select the most discriminant channels. The former uses the accuracy based-classifier criterion, while the latter evaluates electrode mutual information to form discriminant channel subsets. Next, the EEGNet network is implemented to classify discriminant channel signals. Additionally, a cyclic learning algorithm is implemented at the software level to accelerate the model learning convergence and fully profit from the NJT2 hardware resources. Finally, motor imagery Electroencephalogram (EEG) signals provided by HaLT’s public benchmark were used, in addition to the k-fold cross-validation method. Average accuracies of 83.7% and 81.3% were achieved by classifying EEG signals per subject and motor imagery task, respectively. Each task was processed with an average latency of 48.7 ms. This framework offers an alternative for online EEG-BCI systems’ requirements, dealing with short processing times and reliable classification accuracy. Full article

(This article belongs to the Special Issue EEG Signal Processing Techniques and Applications—2nd Edition)

► Show Figures

Figure 1

Figure 1
The proposed method overall flowchart. EEG signals of six MI tasks are provided by [<a href="#B40-sensors-23-04164" class="html-bibr">40</a>]. The red rectangle centered on the circle refers to “Passive” and moves according to the subject’s MI task. The first step consists of selecting discriminant channels from the 19 provided. Next, two comparative methods are used: the ARbC method and the CMIbA. Therefore, the EEGNet network classifies the feature signals into six classes to give the output. Full article ">Figure 2
Channels’ spatial location on the skull in making the referred dataset. According to the 10–20 system, uppercase letters define the brain cortex where an electrode is placed. F for Frontal, T for temporal, P for parietal, and O for occipital cortex. The lowercase ”z” is utilized to locate electrodes on the skull’s longitudinal axis. A1 and A2 mean left and right reference voltage electrodes, respectively. Full article ">Figure 3
Overview of the EEG acquisition and processing in the experimental paradigm. The red rectangle on the eGUI moves over the specific limb icon as a visual stimulus to engage the respective mental task of imagined movement. MI–EEG signals from six mental states were recorded by EEG-1200 equipment and processed using Neurofax recording software [<a href="#B40-sensors-23-04164" class="html-bibr">40</a>]. In addition, ASCII data were converted into Matlab files for further processing. Full article ">Figure 4
The encapsulated EEGNet structure. EEG signals were organized by subject, channel, and sample length. This data matrix was expanded to four dimensions fulfilling the EEGNet input matrix dimension. In Part (a), temporal features are extracted by Conv2D, and in Part (b), spatial filters are applied to enhance feature maps. Then, feature maps are combined in Separable Conv2D (Part (c)), providing the output class probability (Part (d)). Full article ">Figure 5
t-SNE distribution illustrations of selected channels’ signals for all subjects using the ARbC method and CMIbA before the main processing step. All figures were plotted in 2-D embedded space using the Euclidean metric, setting the nearest neighbors’ number at 10, the number of iterations for the optimization at 1000, and the gradient norm at 0.0001. (a) ARbC: six-channel combination: distribution of {Fp1,F8,Fp2,F7,P3,Cz} channel signals, (b) CMIbA: six-channel combination: distribution of {P4,T6,T3,P3,F4,O2} channel signals, (c) ARbC: eight-channel combination: distribution of {Fp1,F8,Fp2,F7,P3,Cz,O1,P4} channel signals, (d) CMIbA: eight-channel combination: distribution of {P4,T6,T3,P3,F4,O2,Fp2,Fz} channel signals. Full article ">Figure 6
Illustration of MI–EEG features before and after the classification using the EEGNet network. In this example, subject J’s data are provided by the Fp1 channel. The window was set to 170 samples, corresponding to task duration. The normalized magnitude is given in <math display="inline"><semantics> <mi mathvariant="sans-serif">μ</mi> </semantics></math>V, while SPS and <math display="inline"><semantics> <mi>ρ</mi> </semantics></math> mean the number of samples per second and feature magnitude, respectively. (a) MI–EEG signals before classification. (b) MI–EEG features after classification. Full article ">

Journal Menu

Journal Browser

EEG Signal Processing Techniques and Applications—2nd Edition

Share This Special Issue

Special Issue Editors

Special Issue Information

Keywords

Benefits of Publishing in a Special Issue

Related Special Issue

Published Papers (16 papers)

Research

Further Information

Guidelines

MDPI Initiatives

Follow MDPI