Diagnosis of Schizophrenia in EEG Signals Using dDTF Effective Connectivity and New PreTrained CNN and Transformer Models

Lecture notes in computer science, Год журнала: 2024, Номер unknown, С. 150 - 160

Опубликована: Янв. 1, 2024

Язык: Английский

---

Vibration Signal Analysis for Intelligent Rotating Machinery Diagnosis and Prognosis: A Comprehensive Systematic Literature Review

Vibration, Год журнала: 2024, Номер 7(4), С. 1013 - 1062

Опубликована: Окт. 31, 2024

Many industrial processes, from manufacturing to food processing, incorporate rotating elements as principal components in their production chain. Failure of these often leads costly downtime and potential safety risks, further emphasizing the importance monitoring health state. Vibration signal analysis is now a common approach for this purpose, it provides useful information related dynamic behavior machines. This research aimed conduct comprehensive examination current methodologies employed stages vibration analysis, which encompass preprocessing, post-processing phases, ultimately leading application Artificial Intelligence-based diagnostics prognostics. An extensive search was conducted various databases, including ScienceDirect, IEEE, MDPI, Springer, Google Scholar, 2020 early 2024 following PRISMA guidelines. Articles that aligned with at least one targeted topics cited above provided unique methods explicit results qualified retention, while those were redundant or did not meet established inclusion criteria excluded. Subsequently, 270 articles selected an initial pool 338. The review highlighted several deficiencies preprocessing step experimental validation, implementation rates 15.41% 10.15%, respectively, prototype studies. Examination processing phase revealed time scale decomposition have become essential accurate signals, they facilitate extraction complex remains obscured original, undecomposed signals. Combining such time–frequency shown be ideal combination extraction. In context fault detection, support vector machines (SVMs), convolutional neural networks (CNNs), Long Short-Term Memory (LSTM) networks, k-nearest neighbors (KNN), random forests been identified five most frequently algorithms. Meanwhile, transformer-based models are emerging promising venue prediction RUL values, along data transformation. Given conclusions drawn, future researchers urged investigate interpretability integration diagnosis prognosis developed aim applying them real-time contexts. Furthermore, there need studies disclose details datasets operational conditions machinery, thereby improving reproducibility. Another area warrants investigation differentiation types present signals obtained bearings, defect overall system embedded within

Язык: Английский

---

Enhancing cross-subject EEG emotion recognition through multi-source manifold metric transfer learning

Computers in Biology and Medicine, Год журнала: 2024, Номер 174, С. 108445 - 108445

Опубликована: Апрель 9, 2024

Язык: Английский

---

Smart Healthcare: Exploring the Internet of Medical Things with Ambient Intelligence

Electronics, Год журнала: 2024, Номер 13(12), С. 2309 - 2309

Опубликована: Июнь 13, 2024

Ambient Intelligence (AMI) represents a significant advancement in information technology that is perceptive, adaptable, and finely attuned to human needs. It holds immense promise across diverse domains, with particular relevance healthcare. The integration of Artificial (AI) the Internet Medical Things (IoMT) create an AMI environment medical contexts further enriches this concept within This survey provides invaluable insights for both researchers practitioners healthcare sector by reviewing incorporation techniques IoMT. analysis encompasses essential infrastructure, including smart environments spectrum wearable non-wearable devices realize vision settings. Furthermore, comprehensive overview cutting-edge AI methodologies employed crafting IoMT systems tailored applications sheds light on existing research issues, aim guiding inspiring advancements dynamic field.

Язык: Английский

---

An Attention-Based Multi-Domain Bi-Hemisphere Discrepancy Feature Fusion Model for EEG Emotion Recognition

IEEE Journal of Biomedical and Health Informatics, Год журнала: 2024, Номер 28(10), С. 5890 - 5903

Опубликована: Июнь 24, 2024

Electroencephalogram (EEG)-based emotion recognition has become a research hotspot in the field of brain-computer interface. Previous methods have overlooked fusion multi-domain emotion-specific information to improve performance, and faced challenge insufficient interpretability. In this paper, we proposed novel EEG model that combined asymmetry brain hemisphere, spatial, spectral, temporal properties signals, aiming performance. Based on 10-20 standard system, global spatial projection matrix (GSPM) bi-hemisphere discrepancy (BDPM) are constructed. A dual-stream spatial-spectral-temporal convolution neural network is designed extract depth features from two paradigms. Finally, transformer-based module used learn dependence fused features, retain discriminative information. We conducted extensive experiments SEED, SEED-IV, DEAP public datasets, achieving excellent average results 98.33/2.46 %, 92.15/5.13 97.60/1.68 %(valence), 97.48/1.42 %(arousal) respectively. Visualization analysis supports interpretability model, ablation validate effectiveness fusion.

Язык: Английский

---

Cross-subject emotion recognition with contrastive learning based on EEG signal correlations

Biomedical Signal Processing and Control, Год журнала: 2025, Номер 104, С. 107511 - 107511

Опубликована: Янв. 10, 2025

Язык: Английский

---

Human-Centric Spatial Cognition Detecting System Based on Drivers’ Electroencephalogram Signals for Autonomous Driving

Sensors, Год журнала: 2025, Номер 25(2), С. 397 - 397

Опубликована: Янв. 10, 2025

Existing autonomous driving systems face challenges in accurately capturing drivers’ cognitive states, often resulting decisions misaligned with intentions. To address this limitation, study introduces a pioneering human-centric spatial cognition detecting system based on electroencephalogram (EEG) signals. Unlike conventional EEG-based that focus intention recognition or hazard perception, the proposed can further extract across two dimensions: relative distance and orientation. It consists of components: EEG signal preprocessing decoding, enabling to make more contextually aligned regarding targets drivers on. enhance detection accuracy cognition, we designed novel decoding method called Dual-Time-Feature Network (DTFNet). This approach integrates coarse-grained fine-grained temporal features signals different scales incorporates Squeeze-and-Excitation module evaluate importance electrodes. The DTFNet outperforms existing methods, achieving 65.67% 50.65% three-class tasks 84.46% 70.50% binary tasks. Furthermore, investigated dynamics observed perception occurs slightly later than their orientation, providing valuable insights into aspects processing.

Язык: Английский

---

A Resource-Efficient Multi-Entropy Fusion Method and Its Application for EEG-Based Emotion Recognition

Entropy, Год журнала: 2025, Номер 27(1), С. 96 - 96

Опубликована: Янв. 20, 2025

Emotion recognition is an advanced technology for understanding human behavior and psychological states, with extensive applications mental health monitoring, human–computer interaction, affective computing. Based on electroencephalography (EEG), the biomedical signals naturally generated by brain, this work proposes a resource-efficient multi-entropy fusion method classifying emotional states. First, Discrete Wavelet Transform (DWT) applied to extract five brain rhythms, i.e., delta, theta, alpha, beta, gamma, from EEG signals, followed acquisition of features, including Spectral Entropy (PSDE), Singular Spectrum (SSE), Sample (SE), Fuzzy (FE), Approximation (AE), Permutation (PE). Then, such entropies are fused into matrix represent complex dynamic characteristics EEG, denoted as Brain Rhythm Matrix (BREM). Next, Dynamic Time Warping (DTW), Mutual Information (MI), Spearman Correlation Coefficient (SCC), Jaccard Similarity (JSC) measure similarity between unknown testing BREM data positive/negative samples classification. Experiments were conducted using DEAP dataset, aiming find suitable scheme regarding measures, time windows, input numbers channel data. The results reveal that DTW yields best performance in measures 5 s window. In addition, single-channel mode outperforms single-region mode. proposed achieves 84.62% 82.48% accuracy arousal valence classification tasks, respectively, indicating its effectiveness reducing dimensionality computational complexity while maintaining over 80%. Such performances remarkable when considering limited resources concern, which opens possibilities innovative entropy can help design portable EEG-based emotion-aware devices daily usage.

Язык: Английский

---

Measuring Realistic Emotional Perception With EEG: A Comparison of Multimodal Videos and Naturalistic Scenes

Psychophysiology, Год журнала: 2025, Номер 62(1)

Опубликована: Янв. 1, 2025

ABSTRACT Emotional experiences involve dynamic multisensory perception, yet most EEG research uses unimodal stimuli such as naturalistic scene photographs. Recent suggests that realistic emotional videos reliably reduce the amplitude of a steady‐state visual evoked potential (ssVEP) elicited by flickering border. Here, we examine extent to which this video‐ssVEP measure compares with well‐established Late Positive Potential (LPP) is larger for relative neutral scenes. To address question, 45 participants viewed 90 matched pairs and Consistent prior work, reduced 7–8 Hz ssVEP was evident during videos. However, reduction in power not specific driving frequency 7.5 Hz, fact, Fourier transformation analyses limited were modulated video content. Still, at group level, video‐driven reductions LPP modulation scenes produced similarly large valence effects, both measures strongly correlated arousal ratings. previous research, scene‐LPP sensitive contents (erotica gore) somewhat inconsistent In contrast, oscillation did show content sensitivity better explained individual ratings per clip. sum, these results flickering‐border paradigm does index engagement stimuli, do evoke robust decreases 3–10 oscillatory distinct from scene‐evoked LPP. Matched responses compared within‐participant. Our findings align indicating around (7–8 Hz) serves reliable measure. further reveal effect attributable general decrease across range.

Язык: Английский

---

Classification of Speech Emotion State Based on Feature Map Fusion of TCN and Pretrained CNN Model from Korean Speech Emotion Data

IEEE Access, Год журнала: 2025, Номер 13, С. 19947 - 19963

Опубликована: Янв. 1, 2025

Язык: Английский

---