A review of automated sleep stage based on EEG signals

Journal of Applied Biomedicine, Год журнала: 2024, Номер 44(3), С. 651 - 673

Опубликована: Июнь 29, 2024

Язык: Английский

---

A novel deep learning model based on transformer and cross modality attention for classification of sleep stages

Journal of Biomedical Informatics, Год журнала: 2024, Номер 157, С. 104689 - 104689

Опубликована: Июль 18, 2024

Язык: Английский

---

EDA-Graph: Graph Signal Processing of Electrodermal Activity for Emotional States Detection

IEEE Journal of Biomedical and Health Informatics, Год журнала: 2024, Номер 28(8), С. 4599 - 4612

Опубликована: Май 27, 2024

The continuous detection of emotional states has many applications in mental health, marketing, human-computer interaction, and assistive robotics. Electrodermal activity (EDA), a signal modulated by sympathetic nervous system activity, provides insight into states. However, EDA possesses intricate nonstationary nonlinear characteristics, making the extraction emotion-relevant information challenging. We propose novel graph processing (GSP) approach to model signals as graphical networks, termed EDA-graph. GSP leverages theory concepts capture complex relationships time-series data. To test usefulness EDA-graphs detect emotions, we processed recordings from CASE emotion dataset using quantizing linking values based on Euclidean distance between nearest neighbors. From these EDA-graphs, computed features analysis, including total load centrality (TLC), harmonic (THC), number cliques (GNC), diameter, radius, compared those with obtained traditional techniques. EDA-graph encompassing TLC, THC, GNC, radius demonstrated significant differences (p < 0.05) five (Neutral, Amused, Bored, Relaxed, Scared). Using machine learning models for classifying evaluated leave-one-subject-out cross-validation, achieved five-class F1 score up 0.68.

Язык: Английский

---

Innovative feature engineering methods for graph data science

Elsevier eBooks, Год журнала: 2025, Номер unknown, С. 119 - 134

Опубликована: Янв. 1, 2025

Язык: Английский

---

CellGAT: A GAT-Based Method for Constructing a Cell Communication Network Integrating Multiomics Information

Biomolecules, Год журнала: 2025, Номер 15(3), С. 342 - 342

Опубликована: Фев. 27, 2025

The growth, development, and differentiation of multicellular organisms are primarily driven by intercellular communication, which coordinates the activities diverse cell types. This cell-to-cell signaling is typically mediated various types protein–protein interactions, including ligand–receptor; receptor–receptor, extracellular matrix–receptor interactions. Currently, computational methods for inferring ligand–receptor communication depend on gene expression data pairs spatial information cells. Some approaches integrate protein complexes; transcription factors; or pathway to construct networks. However, few consider critical role interactions (PPIs) in networks, especially when predicting between different absence type information. These often rely that lack PPI evidence, potentially compromising accuracy their predictions. To address this issue, we propose CellGAT, a framework infers integrating pairs, information, complex data, experimentally validated CellGAT not only builds priori models but also uses node embedding algorithms graph attention networks build based scRNA-seq (single-cell RNA sequencing) datasets includes built-in clustering algorithm. Through comparisons with methods, accurately predicts cell–cell (CCC) analyzes its impact downstream pathways; neighboring cells; drug interventions.

Язык: Английский

---

Fusion of multi-scale feature extraction and adaptive multi-channel graph neural network for 12-lead ECG classification

Computer Methods and Programs in Biomedicine, Год журнала: 2025, Номер unknown, С. 108725 - 108725

Опубликована: Март 1, 2025

Язык: Английский

---

Graph-informed convolutional autoencoder to classify brain responses during sleep

Frontiers in Neuroscience, Год журнала: 2025, Номер 19

Опубликована: Апрель 28, 2025

Automated machine-learning algorithms that analyze biomedical signals have been used to identify sleep patterns and health issues. However, their performance is often suboptimal, especially when dealing with imbalanced datasets. In this paper, we present a robust state (SlS) classification algorithm utilizing electroencephalogram (EEG) signals. To aim, pre-processed EEG recordings from 33 healthy subjects. Then, functional connectivity features recurrence quantification analysis were extracted sub-bands. The graphical representation was calculated phase locking value, coherence, phase-amplitude coupling. Statistical select p-values of less than 0.05. These compared between four states: wakefulness, non-rapid eye movement (NREM) sleep, rapid (REM) during presenting auditory stimuli, REM without stimuli. Eighteen types different stimuli including instrumental natural sounds presented participants REM. selected significant train novel deep-learning classifiers. We designed graph-informed convolutional autoencoder called GICA extract high-level the features. Furthermore, an attention layer based on rate EEGs incorporated into classifier enhance dynamic ability model. proposed model assessed by comparing it baseline systems in literature. accuracy SlS-GICA 99.92% feature set. This achievement could be considered real-time automatic applications develop new therapeutic strategies for sleep-related disorders.

Язык: Английский

---

Advancing Sleep Stages Classification through a Dual-Graphormer Approach

Expert Systems with Applications, Год журнала: 2025, Номер unknown, С. 128220 - 128220

Опубликована: Май 1, 2025

Язык: Английский

---

Temporal–spatial skeleton sequence recognition with self-supervised representation learning

The Journal of Supercomputing, Год журнала: 2025, Номер 81(8)

Опубликована: Июнь 5, 2025

Язык: Английский

---

Temporal Self-Attentional and Adaptive Graph Convolutional Mixed Model for Sleep Staging

IEEE Sensors Journal, Год журнала: 2024, Номер 24(8), С. 12840 - 12852

Опубликована: Март 5, 2024

Evaluating sleep quality through reliable staging is of paramount importance. Although many studies reached fair performances in stage classification, effectively leveraging the spatial–temporal characteristics derived from multichannel brain recordings remains challenging. We develop a novel temporal self-attentional and adaptive graph convolutional mixed model (TS-AGCMM), comprising feature extraction module (FEM), dynamic time warping (DTW)-based attention module, context (TCM), (AGCM) this study. First, FEM enables capturing representative information raw data. Then, DTW-based utilizes programming algorithm to enhance spatial expression ability extracted features. The TCM includes multihead mechanisms that capture dependencies. In particular, we employ an named normalization-based (NAM), which contributing factors weights suppress less salient information. Meanwhile, AGCM can obtain optimal functional connections between polysomnography (PSG) channels, benefit learning property adjacency matrix. Finally, fuse features by concat operation prediction results. utilize Montreal archive (MASS) ISRUC-S3 assess TS-AGCMM. TS-AGCMM exhibits performance comparable other currently available approaches as per our results, achieving accuracy 89.1% 81.2%, macroaveraging F1-score 84.7% 79.5%, well Cohen's kappa coefficient 83.9% 75.8% on two databases, respectively.

Язык: Английский

---