Constructing an efficient and accurate epilepsy detection system is an urgent research task. In this paper, we developed an EEG-based multi-frequency multilayer brain network (MMBN) and an attentional mechanism based convolutional neural network (AM-CNN) model to study epilepsy detection. Specifically, based on the multi-frequency characteristics of the brain, we first use wavelet packet decomposition and reconstruction methods to divide the original EEG signals into eight frequency bands, and then construct MMBN through correlation analysis between brain regions, where each layer corresponds to a specific frequency band. The time, frequency and channel related information of EEG signals are mapped into the multilayer network topology. On this basis, a multi-branch AM-CNN model is designed, which completely matches the multilayer structure of the proposed brain network. The experimental results on public CHB-MIT datasets show that eight frequency bands divided in this work are all helpful for epilepsy detection, and the fusion of multi-frequency information can effectively decode the epileptic brain state, achieving accurate detection of epilepsy with an average accuracy of 99.75%, sensitivity of 99.43%, and specificity of 99.83%. All of these provide reliable technical solutions for EEG-based neurological disease detection, especially for epilepsy detection.

Constructing an efficient and accurate epilepsy detection system is an urgent research task. In this paper, we developed an EEG-based multi-frequency multilayer brain network (MMBN) and an attentional mechanism based convolutional neural network (AM-CNN) model to study epilepsy detection. Specifically, based on the multi-frequency characteristics and correlation analysis of the brain, we first construct MMBN, where each layer corresponds to a specific frequency band. The time, frequency and channel related information of EEG signals are mapped into the multilayer network topology. On this basis, a multi branch AM-CNN model is designed, which completely matches the multilayer structure of the proposed brain network. The experimental results on public CHB-MIT dataset show that the accuracy of brain state detection is positively correlated with the fineness of frequency band division. When the raw EEG signal is divided into eight frequency bands, this method can accurately detect epilepsy, with an average accuracy of 99.75%, sensitivity of 99.43%, and specificity of 99.83%. All of these provide a reliable technical solution for epilepsy detection.

Major depressive disorder (MDD) is a very serious mental illness that spreads all over the world and affects patients of all ages. Constructing an efficient and accurate MDD detection system is an urgent research task. In this paper, we develop an EEG-based multilayer brain network and an attention mechanism-based convolutional neural network (AM-CNN) model to study MDD. In detail, based on mutual information theory, we first construct a multilayer brain network, in which each layer corresponds to a specific frequency band. The experimental results show that such a design can effectively reveal the brain physiological changes of MDD patients, from the perspective of network topology analysis. On this basis, multi-branch AM-CNN model is then designed, which uses multilayer brain network as input and can well achieve feature extraction and detection of MDD. On the publicly available MDD dataset, the proposed method achieves an identification accuracy of 97.22%. Our approach and analysis provide novel insights into the physiological changes of MDD patients and a reliable technical solution for MDD detection.