AUTHOR=Lu Liangfu , Zhang Feng , Wu Yubo , Ma Songnan , Zhang Xin , Ni Guangjian 

TITLE=A multi-frame network model for predicting seizure based on sEEG and iEEG data

JOURNAL=Frontiers in Computational Neuroscience

VOLUME=Volume 16 - 2022

YEAR=2022

URL=https://www.frontiersin.org/journals/computational-neuroscience/articles/10.3389/fncom.2022.1059565

DOI=10.3389/fncom.2022.1059565

ISSN=1662-5188

ABSTRACT=Analysis and prediction of seizures by processing the EEG signals could assist doctors in accurate diagnosis and improve the quality of patients’ life with epilepsy. Nowadays, seizure prediction models based on deep learning become one of the most popular topics in seizure studies, and a number of models have been presented. However, the prediction results are strongly related to the various complicated pre-processing strategies of models, and can not be directly applied to raw data in real time applications. Moreover, due to the inherent deficiencies in single frame models and non-stationary nature of EEG signal, the generalization ability of existing model frameworks is generally poor. Therefore, we propose an
end-to-end seizure prediction model in this paper, where we design a multi-frame network for automatic feature extraction and classification. Instance and sequence-based frames are proposed in our approach, which can help us simultaneously extract features of different modes for further classification. Moreover, complicated pre-processing steps are not included in our model, and the novel frames can be directly applied to the raw data. It should be noted that the
approaches proposed in the paper can be easily used as general model which has been validated and compared with existing model frames. The experimental results show that the multi-frame network proposed in this paper is superior to the existing model frame in accuracy, sensitivity, specificity, F1-score and AUC in the classification performance of EEG signals. Our results provide a new research idea
for this field. Researchers can further integrate the idea of multi-frame network into state-of-the-art single frame seizure prediction models, and then achieve better results.