Sleep occupies about one-third of human life and is crucial for health, but traditional sleep staging relies on experts manually performing polysomnography (PSG), a process that is time-consuming, labor-intensive, and susceptible to subjective differences between evaluators. With the development of deep learning technologies, particularly the application of convolutional neural networks and recurrent neural networks, significant progress has been made in automatic sleep staging. However, existing methods still face challenges in feature extraction and cross-modal data fusion. This paper introduces an innovative deep learning architecture, MultiSEss, aimed at solving key issues in automatic sleep stage classification. The MultiSEss architecture utilizes a multi-scale convolution module to capture signal features from different frequency bands and incorporates a Squeeze-and-Excitation attention mechanism to enhance the learning of channel feature weights. Furthermore, the architecture discards complex attention mechanisms or encoder–decoder structures in favor of a state–space sequence coupling module, which more accurately captures and integrates correlations between multi-modal data. Experiments show that MultiSEss achieved accuracy results of 83.84% and 82.30% in five-fold cross-subject testing on the Sleep-EDF-20 and Sleep-EDF-78 datasets. MultiSEss demonstrates its potential in improving sleep stage accuracy, which is significant for enhancing the diagnosis and treatment of sleep disorders.

With the rapid development of modern industry, people’s living pressures are gradually increasing, and an increasing number of individuals are affected by sleep disorders such as insomnia, hypersomnia, and sleep apnea syndrome. Many cardiovascular and psychiatric diseases are also closely related to sleep. Therefore, the early detection, accurate diagnosis, and treatment of sleep disorders an urgent research priority. Traditional manual sleep staging methods have many problems, such as being time-consuming and cumbersome, relying on expert experience, or being subjective. To address these issues, researchers have proposed multiple algorithmic strategies for sleep staging automation based on deep learning in recent years. This paper studies MASleepNet, a sleep staging neural network model that integrates multimodal deep features. This model takes multi-channel Polysomnography (PSG) signals (including EEG (Fpz-Cz, Pz-Oz), EOG, and EMG) as input and employs a multi-scale convolutional module to extract features at different time scales in parallel. It then adaptively weights and fuses the features from each modality using a channel-wise attention mechanism. The integrated temporal features are integrated into a Bidirectional Long Short-Term Memory (BiLSTM) sequence encoder, where an attention mechanism is introduced to identify key temporal segments. The final classification result is produced by the fully connected layer. The proposed model was experimentally evaluated on the Sleep-EDF dataset (consisting of two subsets, Sleep-EDF-78 and Sleep-EDF-20), achieving classification accuracies of 82.56% and 84.53% on the two subsets, respectively. These results demonstrate that deep models that integrate multimodal signals and an attention mechanism offer the possibility to enhance the efficiency of automatic sleep staging compared to cutting-edge methods.