Improving two-dimensional linear discriminant analysis with L1 norm for optimizing EEG signal

Information Sciences, Год журнала: 2024, Номер unknown, С. 121585 - 121585

Опубликована: Окт. 1, 2024

Язык: Английский

---

Multilevel Laser-Induced Pain Measurement with Wasserstein Generative Adversarial Network — Gradient Penalty Model

International Journal of Neural Systems, Год журнала: 2023, Номер 34(01)

Опубликована: Окт. 20, 2023

Pain is an experience of unpleasant sensations and emotions associated with actual or potential tissue damage. In the global context, billions people are affected by pain disorders. There particular challenges in measurement assessment pain, commonly used measuring tools include traditional subjective scoring methods biomarker-based measures. The main for analysis electroencephalography (EEG), electrocardiography functional magnetic resonance. EEG-based quantitative measurements immense value clinical management can provide objective assessments intensity. now primarily limited to identification presence absence less research on multilevel pain. High power laser stimulation experimental paradigm five level classification based EEG data augmentation presented. First, features extracted using modified S-transform, time-frequency information retained. Based recognition effect, 20-40[Formula: see text]Hz frequency band optimized. Afterwards Wasserstein generative adversarial network gradient penalty feature augmentation. It be inferred from good performance parietal region brain that sensory function lobe effectively activated during occurrence By comparing latest algorithms, proposed method has significant advantages five-level dataset. This provides new ways thinking related recognition, which essential study neural mechanisms regulatory

Язык: Английский

---

STaRNet: A spatio-temporal and Riemannian network for high-performance motor imagery decoding

Neural Networks, Год журнала: 2024, Номер 178, С. 106471 - 106471

Опубликована: Июнь 26, 2024

Язык: Английский

---

Enhancing Motor Imagery Classification with Residual Graph Convolutional Networks and Multi-Feature Fusion

International Journal of Neural Systems, Год журнала: 2024, Номер unknown

Опубликована: Сен. 20, 2024

Stroke, an abrupt cerebrovascular ailment resulting in brain tissue damage, has prompted the adoption of motor imagery (MI)-based brain–computer interface (BCI) systems stroke rehabilitation. However, analyzing electroencephalogram (EEG) signals from patients poses challenges. To address issues low accuracy and efficiency EEG classification, particularly involving MI, study proposes a residual graph convolutional network (M-ResGCN) framework based on modified S-transform (MST), introduces self-attention mechanism into (ResGCN). This uses MST to extract time-frequency domain features, derives spatial features by calculating absolute Pearson correlation coefficient (aPcc) between channels, devises method construct adjacency matrix using aPcc measure strength connection channels. Experimental results 16 healthy subjects demonstrate significant improvements classification quality robustness across tests subjects. The highest reached 94.91% Kappa 0.8918. average F1 scores 10 times 10-fold cross-validation are 94.38% 94.36%, respectively. By validating feasibility applicability networks constructed signal analysis feature encoding, it was established that effectively reflects overall activity. proposed presents novel approach exploring channel relationships MI-EEG improving performance. It holds promise for real-time applications MI-based BCI systems.

Язык: Английский

---

Time-resolved EEG signal analysis for motor imagery activity recognition

Biomedical Signal Processing and Control, Год журнала: 2023, Номер 86, С. 105179 - 105179

Опубликована: Июнь 28, 2023

Язык: Английский

---

Medical object detector jointly driven by knowledge and data

Neural Networks, Год журнала: 2023, Номер 172, С. 106084 - 106084

Опубликована: Дек. 29, 2023

Язык: Английский

---

Classification of the Attempted Arm and Hand Movements of Patients with Spinal Cord Injury Using Deep Learning Approach

medRxiv (Cold Spring Harbor Laboratory), Год журнала: 2023, Номер unknown

Опубликована: Июль 8, 2023

Abstract The primary objective of this research is to improve the average classification performance for specific movements in patients with cervical spinal cord injury (SCI). study utilizes a low-frequency multi-class electroencephalography (EEG) dataset obtained from Institute Neural Engineering at Graz University Technology. combines convolutional neural network (CNN) and long-short-term memory (LSTM) architectures uncover strong correlations between temporal spatial aspects EEG signals associated attempted arm hand movements. To achieve this, three different methods are used select relevant features, proposed model’s robustness against variations data validated using 10-fold cross-validation (CV). Furthermore, explores potential subject-specific adaptation an online paradigm, extending proof-of-concept classifying movement attempts. In summary, aims make valuable contributions field neuro-technology by developing EEG-controlled assistive devices generalized brain-computer interface (BCI) deep learning (DL) framework. focus on capturing high-level spatiotemporal features latent dependencies enhance usability EEG-based technologies.

Язык: Английский

---

Aided diagnosis of cervical spondylotic myelopathy using deep learning methods based on electroencephalography

Medical Engineering & Physics, Год журнала: 2023, Номер 121, С. 104069 - 104069

Опубликована: Ноя. 1, 2023

Язык: Английский

---

Deep learning classification of EEG-based BCI monitoring of the attempted arm and hand movements

Biomedical Engineering / Biomedizinische Technik, Год журнала: 2024, Номер unknown

Опубликована: Июнь 3, 2024

Abstract Objectives The primary objective of this research is to improve the average classification performance for specific movements in patients with cervical spinal cord injury (SCI). Methods study utilizes a low-frequency multi-class electroencephalography (EEG) dataset from Graz University Technology. combines convolutional neural network (CNN) and long-short-term memory (LSTM) architectures uncover correlations between temporal spatial aspects EEG signals associated attempted arm hand movements. To achieve this, three different methods are used select relevant features, proposed model’s robustness against variations data validated using 10-fold cross-validation (CV). also investigates subject-specific adaptation an online paradigm, extending movement proof-of-concept. Results combined CNN-LSTM model, enhanced by feature selection methods, demonstrates mean accuracy 75.75 % low standard deviation (+/− 0.74 %) cross-validation, confirming its reliability. Conclusions In summary, aims make valuable contributions field neuro-technology developing EEG-controlled assistive devices generalized brain-computer interface (BCI) deep learning (DL) framework. focus on capturing high-level spatiotemporal features latent dependencies enhance usability EEG-based technologies.

Язык: Английский

---

Functional connectivity of EEG motor rhythms after spinal cord injury

Cognitive Neurodynamics, Год журнала: 2024, Номер 18(5), С. 3015 - 3029

Опубликована: Июнь 10, 2024

Язык: Английский

---