Graph Neural Networks in EEG Spike Detection

SSRN Electronic Journal(2022)

引用 0|浏览7
暂无评分
摘要
•Two two-stage algorithms are developed to detect spikes in epileptic scalp EEG.•A convolutional neural network (CNN) is used to extract node embeddings.•Graph neural networks (GNNs) are applied on functional connectivity (FC) maps.•Attention mechanism is used to bypass the feature engineering step in the last point.•Higher average precision (AP) is achieved compared to standard architectures.
更多
查看译文
关键词
Scalp EEG, Interictal epileptiform discharge, Graph neural networks, Attention, Functional connectivity, Weighted phase lag index
AI 理解论文
溯源树
样例
生成溯源树,研究论文发展脉络
Chat Paper
正在生成论文摘要