期刊
IEEE JOURNAL OF BIOMEDICAL AND HEALTH INFORMATICS
卷 26, 期 11, 页码 5418-5427出版社
IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC
DOI: 10.1109/JBHI.2022.3199206
关键词
Electroencephalography; Feature extraction; Transformers; Recording; Brain modeling; Detection algorithms; Sensitivity; Seizure detection; iEEG; transformer; self-attention; deep learning; epilepsy
类别
资金
- National Natural Science Foundation of China [61906132, 81925020]
- Key Project and Team Program of Tianjin City [XC202020]
- Natural Science Foundation of Tianjin City [20JCYBJC00920]
- National Key Research and Development Program of China [2021YFF1200600]
This study proposes an algorithm for automatic seizure detection using continuous, long-term iEEG data. By introducing the ability of transformer networks to calculate attention between input signal channels, the authors designed an end-to-end model that includes convolution and transformer layers. Through evaluation on two datasets, it was experimentally demonstrated that the transformer layer can improve the performance of the seizure detection algorithm, and the model provides better explainability of the seizure onset process.
Automatic seizure detection algorithms are necessary for patients with refractory epilepsy. Many excellent algorithms have achieved good results in seizure detection. Still, most of them are based on discontinuous intracranial electroencephalogram (iEEG) and ignore the impact of different channels on detection. This study aimed to evaluate the proposed algorithm using continuous, long-term iEEG to show its applicability in clinical routine. In this study, we introduced the ability of the transformer network to calculate the attention between the channels of input signals into seizure detection. We proposed an end-to-end model that included convolution and transformer layers. The model did not need feature engineering or format transformation of the original multi-channel time series. Through evaluation on two datasets, we demonstrated experimentally that the transformer layer could improve the performance of the seizure detection algorithm. For the SWEC-ETHZ iEEG dataset, we achieved 97.5% event-based sensitivity, 0.06/h FDR, and 13.7 s latency. For the TJU-HH iEEG dataset, we achieved 98.1% event-based sensitivity, 0.22/h FDR, and 9.9 s latency. In addition, statistics showed that the model allocated more attention to the channels close to the seizure onset zone within 20 s after the seizure onset, which improved the explainability of the model. This paper provides a new method to improve the performance and explainability of automatic seizure detection.
作者
我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。
推荐
暂无数据