PearNet: A Pearson Correlation-based Graph Attention Network for Sleep Stage Recognition
This work addresses sleep stage recognition for medical diagnosis, but it is incremental as it builds on existing graph-based models by focusing on internal signal relationships.
The paper tackled the problem of sleep stage recognition by addressing internal relationships within electrode signals in a specific brain region, which existing graph-based models could not solve, and proposed PearNet, a Pearson correlation-based graph attention network that outperformed state-of-the-art baselines on Sleep-EDF-20 and Sleep-EDF-78 datasets.
Sleep stage recognition is crucial for assessing sleep and diagnosing chronic diseases. Deep learning models, such as Convolutional Neural Networks and Recurrent Neural Networks, are trained using grid data as input, making them not capable of learning relationships in non-Euclidean spaces. Graph-based deep models have been developed to address this issue when investigating the external relationship of electrode signals across different brain regions. However, the models cannot solve problems related to the internal relationships between segments of electrode signals within a specific brain region. In this study, we propose a Pearson correlation-based graph attention network, called PearNet, as a solution to this problem. Graph nodes are generated based on the spatial-temporal features extracted by a hierarchical feature extraction method, and then the graph structure is learned adaptively to build node connections. Based on our experiments on the Sleep-EDF-20 and Sleep-EDF-78 datasets, PearNet performs better than the state-of-the-art baselines.