Litcius/Paper detail

Interpretable Sleep Stage Classification Based on Layer-Wise Relevance Propagation

Dongdong Zhou, Qi Xu, Jiacheng Zhang, Lei Wu, Hongming Xu, Lauri Kettunen, Zheng Chang, Qiang Zhang, Fengyu Cong

2024IEEE Transactions on Instrumentation and Measurement11 citationsDOIOpen Access PDF

Abstract

Numerous deep learning-based methodologies have been proposed to facilitate automatic sleep stage classification tasks. Nevertheless, the black-box nature of these approaches is one of the skeptical factors hindering clinical application. Towards model interpretability, this study presents a novel interpretable sleep stage classification scheme based on layer-wise relevance propagation (LRP). We first adopt the short-time Fourier transform (STFT) to convert the raw electroencephalogram (EEG) signals to the time-frequency images, which could visually demonstrate EEG patterns of each sleep stage. Moreover, we introduce an efficient convolutional neural network (CNN) based model, namely MSSENet, that assembles with the Multi-Scale CNN module and residual Squeeze-and-Excitation block for the image input. The LRP method is eventually applied to evaluate the contribution of each frequency pixel in the input time-frequency image to the model prediction. Experimental findings show that the MSSENet could outperforms or achieves comparable performance to other state-of-the-art approaches on three polysomnography (PSG) datasets. Furthermore, through utilizing the heat mapping, the LRP-based explainability results validate the high relevance of specific EEG patterns to the prediction of the corresponding sleep stage, which is consistent with the sleep scoring guidelines.

Topics & Concepts

Relevance (law)Stage (stratigraphy)Sleep (system call)Computer scienceArtificial intelligenceLayer (electronics)Pattern recognition (psychology)Speech recognitionMaterials scienceGeologyOperating systemComposite materialLawPaleontologyPolitical scienceEEG and Brain-Computer InterfacesNon-Invasive Vital Sign MonitoringObstructive Sleep Apnea Research