Sample-Based Data Augmentation Based on Electroencephalogram Intrinsic Characteristics

IEEE Journal of Biomedical and Health Informatics, Journal Year: 2022, Volume and Issue: 26(10), P. 4996 - 5003

Published: June 23, 2022

Deep learning for electroencephalogram-based classification is confronted with data scarcity, due to the time-consuming and expensive collection procedure. Data augmentation has been shown as an effective way improve efficiency. In addition, contrastive recently hold great promise in representations without human supervision, which potential recognition performance limited labeled data. However, heavy a key ingredient of learning. view number sample-based electroencephalogram processing, three methods, performance-measure-based time warp, frequency noise addition masking, are proposed based on characteristics signal. These methods parameter free, easy implement, can be applied individual samples. experiment, evaluated tasks, including situation awareness recognition, motor imagery brain-computer interface steady-state visually evoked potentials speller system. Results demonstrated that convolutional models trained yielded significantly improved over baselines. overall, this work provides more cope problem boost processing.

Language: Английский

---

Selective multi–view time–frequency decomposed spatial feature matrix for motor imagery EEG classification

Expert Systems with Applications, Journal Year: 2024, Volume and Issue: 247, P. 123239 - 123239

Published: Jan. 24, 2024

Language: Английский

---

Multi-branch spatial-temporal-spectral convolutional neural networks for multi-task motor imagery EEG classification

Biomedical Signal Processing and Control, Journal Year: 2024, Volume and Issue: 93, P. 106156 - 106156

Published: Feb. 28, 2024

Language: Английский

---

EEG based over-complete rational dilation wavelet transform coupled with autoregressive for motor imagery classification

Expert Systems with Applications, Journal Year: 2025, Volume and Issue: unknown, P. 126433 - 126433

Published: Jan. 1, 2025

Language: Английский

---

Multiscale Convolutional Transformer for EEG Classification of Mental Imagery in Different Modalities

IEEE Transactions on Neural Systems and Rehabilitation Engineering, Journal Year: 2022, Volume and Issue: 31, P. 646 - 656

Published: Dec. 15, 2022

A new kind of sequence-to-sequence model called a transformer has been applied to electroencephalogram (EEG) systems. However, the majority EEG-based models have attention mechanisms temporal domain, while connectivity between brain regions and relationship different frequencies neglected. In addition, many related studies on imagery-based brain-computer interface (BCI) limited classifying EEG signals within one type imagery. Therefore, it is important develop general learn various types neural representations. this study, we designed an experimental paradigm based motor imagery, visual speech imagery tasks interpret representations during mental in modalities. We conducted source localization investigate networks. propose multiscale convolutional for decoding which applies multi-head over spatial, spectral, domains. The proposed network shows promising performance with 0.62, 0.70, 0.72 accuracy private dataset, BCI competition IV 2a Arizona State University respectively, as compared conventional deep learning models. Hence, believe that will contribute significantly overcoming number classes low classification performances system.

Language: Английский

---

Recent Trends in EEG-Based Motor Imagery Signal Analysis and Recognition: A Comprehensive Review

IEEE Access, Journal Year: 2023, Volume and Issue: 11, P. 80518 - 80542

Published: Jan. 1, 2023

The electroencephalogram (EEG) motor imagery (MI) signals are the widespread paradigms in brain-computer interface (BCI). Its significant applications gaming, robotics, and medical fields drew our attention to perform a detailed analysis. However, problem is ill-posed as these highly nonlinear, unpredictable, noisy, hence making it exceedingly hard be analyzed adequately. This paper provides first-of-its-kind comprehensive review of conventional signal processing deep learning techniques for BCI MI comprises extensive works carried out domain recent past, highlighting current challenges problem. A new categorization existing approaches has been presented better clarification. An all-inclusive description corroborated by relevant area. Moreover, architectures various standard algorithms along with their merits demerits also explicated assist readers. tabular representations numerical results readily provided. work presents open research problems future directions.

Language: Английский

---

Application and Development of EEG Acquisition and Feedback Technology: A Review

Biosensors, Journal Year: 2023, Volume and Issue: 13(10), P. 930 - 930

Published: Oct. 17, 2023

This review focuses on electroencephalogram (EEG) acquisition and feedback technology its core elements, including the composition principles of devices, a wide range applications, commonly used EEG signal classification algorithms. First, we describe construction devices encompassing electrodes, processing, control systems, which collaborate to measure faint signals from scalp, convert them into interpretable data, accomplish practical applications using systems. Subsequently, examine diverse across various domains. In medical field, are employed for epilepsy diagnosis, brain injury monitoring, sleep disorder research. has revealed associations between functionality, cognition, emotions, providing essential insights psychologists neuroscientists. Brain-computer interface utilizes human-computer interaction, driving innovation in medical, engineering, rehabilitation Finally, introduce These tasks can identify different cognitive states, emotional disorders, brain-computer promote further development application technology. conclusion, deepen understanding while simultaneously promoting developments multiple domains, such as medicine, science, engineering.

Language: Английский

---

A robust multi-branch multi-attention-mechanism EEGNet for motor imagery BCI decoding

Journal of Neuroscience Methods, Journal Year: 2024, Volume and Issue: 405, P. 110108 - 110108

Published: March 6, 2024

Language: Английский

---

Modeling Nonlinear Dynamics in Human–Machine Interaction

IEEE Access, Journal Year: 2023, Volume and Issue: 11, P. 58664 - 58678

Published: Jan. 1, 2023

Language: Английский

---

Deep temporal networks for EEG-based motor imagery recognition

Scientific Reports, Journal Year: 2023, Volume and Issue: 13(1)

Published: Nov. 1, 2023

The electroencephalogram (EEG) based motor imagery (MI) signal classification, also known as motion recognition, is a highly popular area of research due to its applications in robotics, gaming, and medical fields. However, the problem ill-posed these signals are non-stationary noisy. Recently, lot efforts have been made improve MI classification using combination decomposition machine learning techniques but they fail perform adequately on large multi-class datasets. Previously, researchers implemented long short-term memory (LSTM), which capable time-series information, MI-EEG dataset for recognition. it can not model very long-term dependencies present recognition data. With advent transformer networks natural language processing (NLP), dependency issue has widely addressed. Motivated by success algorithms, this article, we propose transformer-based deep neural network architecture that performs raw BCI competition III IVa IV 2a validation results show proposed method achieves superior performance than existing state-of-the-art methods. produces accuracy 99.7% 84% binary class datasets, respectively. Further, compared with LSTM.

Language: Английский

---