Seizure detection from multi-channel EEG using entropy-based dynamic graph embedding

作者：

Highlights：

• We proposed a novel dynamic graph embedding model for seizure detection.

• Graph entropies are exploited to measure the similarity between graphs.

• The proposed dynamic graph embedding model is based on graph entropy.

• Proposed model outperformed the baselines on CHB-MIT Scalp EEG database.

摘要

•We proposed a novel dynamic graph embedding model for seizure detection.•Graph entropies are exploited to measure the similarity between graphs.•The proposed dynamic graph embedding model is based on graph entropy.•Proposed model outperformed the baselines on CHB-MIT Scalp EEG database.

论文关键词：Seizure detection,Dynamic graph embedding,Graph entropy

论文评审过程：Received 13 April 2021, Revised 26 October 2021, Accepted 27 October 2021, Available online 3 November 2021, Version of Record 4 November 2021.

论文官网地址：https://doi.org/10.1016/j.artmed.2021.102201