Depression Assessment Method: An EEG Emotion Recognition Framework Based on Spatiotemporal Neural Network

Abstract

The main characteristic of depression is emotional dysfunction, manifested by increased levels of negative emotions and decreased levels of positive emotions. Therefore, accurate emotion recognition is an effective way to assess depression. Among the various signals used for emotion recognition, electroencephalogram (EEG) signal has attracted widespread attention due to its multiple advantages, such as rich spatiotemporal information in multi-channel EEG signals. First, we use filtering and Euclidean alignment for data preprocessing. In the feature extraction, we use short-time Fourier transform and Hilbert–Huang transform to extract time-frequency features, and convolutional neural networks to extract spatial features. Finally, bi-directional long short-term memory explored the timing relationship. Before performing the convolution operation, according to the unique topology of the EEG channel, the EEG features are converted into 3D tensors. This study has achieved good results on two emotion databases: SEED and Emotional BCI of 2020 WORLD ROBOT COMPETITION. We applied this method to the recognition of depression based on EEG and achieved a recognition rate of more than 70% under the five-fold cross-validation. In addition, the subject-independent protocol on SEED data has achieved a state-of-the-art recognition rate, which exceeds the existing research methods. We propose a novel EEG emotion recognition framework for depression detection, which provides a robust algorithm for real-time clinical depression detection based on EEG.

Cross-Subject Emotion Recognition Using Fused Entropy Features of EEG

“…Although the subjects watched their native language films during the experiment to elicit emotional changes more easily, and the stimulus types of the films are the same, there are differences between the Chinese and French subjects, which may lead to the lower accuracy in our study. In a study utilizing the same data as ours, a similar accuracy of about 70% was obtained in [ 37 ] for depression recognition.…”

“…The self-reported emotional states were then used to validate the emotion classification results of the study. The details about the database can be found in [ 11 , 37 , 40 ].…”

“…Mahmud et al trained an automated BiLSTM model to detect sleep apnea based on EEG and reached a high accuracy on different publicly available datasets [ 36 ]. Chang et al proposed a depression assessment framework based on the spatiotemporal network of EEG and BiLSTM and achieved more than 70% accuracy in the SEED dataset [ 37 ].…”

EEG-based major depressive disorder recognition by selecting discriminative features via stochastic search

“…Therefore, based on the analysis of data balance and test protocol, brain function networks [35] has the best recognition performance without feature selection. Lastly, when evaluated on the same dataset, spatiotemporal neural network [64], EEGNet [34], our method has better recognition rate than neural network, so stochastic search based feature selection method plays a significant role.…”

“…Therefore, the combination of the accuracy-related Besides, our method has many advantages through comprehensive comparative analysis with other methods. First of all, from the point view of the datasets, Spatio-temporal neural network [64], KNN [37], multivariate pattern analysis [69], feature-level fusion [67], improved empirical mode decomposition [14], EEGNet [34], case-based reasoning model [31], sequential backward feature selection [36], DCNN-LSTM [32] and our methods are all facing class imbalance (i.e. the difference in the number of subjects between MDD and HC is greater than 1), our method has the highest accuracy.…”