Automated Sleep Staging via Parallel Frequency-Cut Attention

TL;DR

This paper introduces a new attention-based framework for automatic sleep staging from EEG signals, leveraging time-frequency analysis to improve accuracy and interpretability, validated on a large dataset with state-of-the-art results.

Contribution

It presents a novel parallel frequency-cut attention model that captures EEG time-frequency features for sleep staging, achieving superior performance and interpretability.

Findings

State-of-the-art F1 scores for sleep stages wake, N2, and N3.

High inter-rater reliability of 0.80 kappa.

Effective visualization of sleep staging decisions.

Abstract

This paper proposes a novel framework for automatically capturing the time-frequency nature of electroencephalogram (EEG) signals of human sleep based on the authoritative sleep medicine guidance. The framework consists of two parts: the first part extracts informative features by partitioning the input EEG spectrograms into a sequence of time-frequency patches. The second part is constituted by an attention-based architecture to efficiently search for the correlation between partitioned time-frequency patches and defining factors of sleep stages in parallel. The proposed pipeline is validated on the Sleep Heart Health Study dataset with new state-of-the-art results for the stages wake, N2, and N3, obtaining respective F1 scores of 0.93, 0.88, and 0.87, with only EEG signals used. The proposed method also has a high inter-rater reliability of 0.80 kappa. We also visualize the correspondence between sleep staging decisions and features extracted by the proposed method, providing strong interpretability for our model.