Multistream Dilated Convolutional Feature Fusion Neural Network for SSVEP Classification DOI
Xiujun Li,

Yongzheng Zhang,

Yue Wu

et al.

Published: Nov. 17, 2023

Based on steady-state visual evoked potentials (SSVEP), a neuroelectric phenomenon where the brain's electrical signals respond to specific frequency stimuli, which holds significant application value. Due signal noise and individual differences, achieving accurate SSVEP classification remains highly challenging. To address these challenges, we propose multi-stream atrous convolutional feature fusion neural network (MACNN) model. The model adopts parallel structure with multiple streams of convolution for fusion. Each stream utilizes different dilation rates, sharing weights across various streams, incorporates module, allowing leverage information from maps. Finally, an attention mechanism is introduced adaptively emphasize critical channels, thereby enhancing discriminative power classification. Results data involving 35 subjects indicate that, 1 -second length, average accuracy transfer rate increase 79.94% 141.57 bits/min, respectively. Consequently, proposed method importance in research

Language: Английский

U-shaped convolutional transformer GAN with multi-resolution consistency loss for restoring brain functional time-series and dementia diagnosis DOI Creative Commons
Qiankun Zuo, Ruiheng Li, Binghua Shi

et al.

Frontiers in Computational Neuroscience, Journal Year: 2024, Volume and Issue: 18

Published: April 17, 2024

The blood oxygen level-dependent (BOLD) signal derived from functional neuroimaging is commonly used in brain network analysis and dementia diagnosis. Missing the BOLD may lead to bad performance misinterpretation of findings when analyzing neurological disease. Few studies have focused on restoration time-series data.

Language: Английский

Citations

4
A comprehensive review of deep learning power in steady-state visual evoked potentials DOI
Z.T. Al-Qaysi, A. S. Albahri, M. A. Ahmed

et al.

Neural Computing and Applications, Journal Year: 2024, Volume and Issue: 36(27), P. 16683 - 16706

Published: July 23, 2024

Language: Английский

Citations

3
A novel approach for ASD recognition based on graph attention networks DOI Creative Commons
Canhua Wang, Zhiyong Xiao, Yilu Xu

et al.

Frontiers in Computational Neuroscience, Journal Year: 2024, Volume and Issue: 18

Published: April 10, 2024

Early detection and diagnosis of Autism Spectrum Disorder (ASD) can significantly improve the quality life for affected individuals. Identifying ASD based on brain functional connectivity (FC) poses a challenge due to high heterogeneity subjects' fMRI data in different sites. Meanwhile, deep learning algorithms show efficacy identification but lack interpretability. In this paper, novel approach recognition is proposed graph attention networks. Specifically, we treat region interest (ROI) subjects as node, conduct wavelet decomposition BOLD signal each ROI, extract features, utilize them along with mean variance node optimized FC matrix adjacency matrix, respectively. We then employ self-attention mechanism capture long-range dependencies among features. To enhance interpretability, node-selection pooling layers are designed determine importance ROI prediction. The framework applied children (younger than 12 years old) from Brain Imaging Data Exchange datasets. Promising results demonstrate superior performance compared recent similar studies. obtained exhibit correspondence previous studies offer good

Language: Английский

Citations

2
A Comparative Review of Detection Methods in SSVEP-based Brain-Computer Interfaces DOI Creative Commons

Amin Besharat,

Nasser Samadzadehaghdam, Reyhaneh Afghan

et al.

IEEE Access, Journal Year: 2024, Volume and Issue: 12, P. 181232 - 181270

Published: Jan. 1, 2024

Language: Английский

Citations

2
Encoding Global Semantic and Localized Geographic Spatial-Temporal Relations for Traffic Accident Risk Prediction DOI
Fares Alhaek, Tianrui Li, Taha M. Rajeh

et al.

Information Sciences, Journal Year: 2024, Volume and Issue: unknown, P. 121767 - 121767

Published: Dec. 1, 2024

Language: Английский

Citations

1
Epileptic focus localization using transfer learning on multi-modal EEG DOI Creative Commons
Yong Yang, Feng Li, Jing Luo

et al.

Frontiers in Computational Neuroscience, Journal Year: 2023, Volume and Issue: 17

Published: Nov. 23, 2023

The standard treatments for epilepsy are drug therapy and surgical resection. However, around 1/3 of patients with intractable drug-resistant, requiring resection the epileptic focus. To address issue drug-resistant focus localization, we have proposed a transfer learning method on multi-modal EEG (iEEG sEEG). A 10-fold cross-validation approach was applied to validate performance pre-trained model Bern-Barcelona Bonn datasets, achieving accuracy rates 94.50 97.50%, respectively. experimental results demonstrated that outperforms competitive state-of-the-art baselines in terms accuracy, sensitivity, negative predictive value. Furthermore, fine-tuned our using dataset from Chongqing Medical University tested it leave-one-out method, obtaining an impressive average 90.15%. This shows significant feature differences between non-epileptic channels. By extracting data features neural networks, accurate classification channels can be achieved. Therefore, superior has is highly effective localizing aid physicians clinical localization diagnosis.

Language: Английский

Citations

3
Application of artificial intelligence in glaucoma. Part 1. Neural networks and deep learning in glaucoma screening and diagnosis DOI
Н. И. Курышева, Oxana Ye. Rodionova, Alexey L. Pomerantsev

et al.

Russian Annals of Ophthalmology, Journal Year: 2024, Volume and Issue: 140(3), P. 82 - 82

Published: Jan. 1, 2024

This article reviews literature on the use of artificial intelligence (AI) for screening, diagnosis, monitoring and treatment glaucoma. The first part review provides information how AI methods improve effectiveness glaucoma presents technologies using deep learning, including neural networks, analysis big data obtained by ocular imaging (fundus imaging, optical coherence tomography anterior posterior eye segments, digital gonioscopy, ultrasound biomicroscopy, etc.), a multimodal approach. results found in reviewed are contradictory, indicating that improvement models requires further research standardized networks timely detection based will reduce risk blindness associated with

Language: Английский

Citations

0
Multistream Dilated Convolutional Feature Fusion Neural Network for SSVEP Classification DOI
Xiujun Li,

Yongzheng Zhang,

Yue Wu

et al.

Published: Nov. 17, 2023

Based on steady-state visual evoked potentials (SSVEP), a neuroelectric phenomenon where the brain's electrical signals respond to specific frequency stimuli, which holds significant application value. Due signal noise and individual differences, achieving accurate SSVEP classification remains highly challenging. To address these challenges, we propose multi-stream atrous convolutional feature fusion neural network (MACNN) model. The model adopts parallel structure with multiple streams of convolution for fusion. Each stream utilizes different dilation rates, sharing weights across various streams, incorporates module, allowing leverage information from maps. Finally, an attention mechanism is introduced adaptively emphasize critical channels, thereby enhancing discriminative power classification. Results data involving 35 subjects indicate that, 1 -second length, average accuracy transfer rate increase 79.94% 141.57 bits/min, respectively. Consequently, proposed method importance in research

Language: Английский

Citations

0