Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

MSSC-BiMamba: Multimodal Sleep Stage Classification and Early Diagnosis of Sleep Disorders with Bidirectional Mamba

Chao Zhang, Weirong Cui, Jingjing Guo

TL;DR

MSSC-BiMamba introduces an efficient, multimodal sleep staging framework by fusing Efficient Channel Attention with Bidirectional Mamba to capture long-range temporal patterns in PSG data. The model targets sleep stage classification and sleep-health discrimination, achieving strong performance on ISRUC-S3/S1 and Sleep-EDF datasets while reducing parameter count relative to Transformer-based approaches. Key contributions include a novel ECA-enabled channel weighting for time-series channels, a bidirectional Mamba module for forward and backward sequence modeling, and comprehensive evaluation across classification, cross-dataset validation, and health status discrimination. The approach promises practical clinical utility by enabling accurate, scalable, and accessible sleep monitoring and early disorder detection in real-world settings.

Abstract

Monitoring sleep states is essential for evaluating sleep quality and diagnosing sleep disorders. Traditional manual staging is time-consuming and prone to subjective bias, often resulting in inconsistent outcomes. Here, we developed an automated model for sleep staging and disorder classification to enhance diagnostic accuracy and efficiency. Considering the characteristics of polysomnography (PSG) multi-lead sleep monitoring, we designed a multimodal sleep state classification model, MSSC-BiMamba, that combines an Efficient Channel Attention (ECA) mechanism with a Bidirectional State Space Model (BSSM). The ECA module allows for weighting data from different sensor channels, thereby amplifying the influence of diverse sensor inputs. Additionally, the implementation of bidirectional Mamba (BiMamba) enables the model to effectively capture the multidimensional features and long-range dependencies of PSG data. The developed model demonstrated impressive performance on sleep stage classification tasks on both the ISRUC-S3 and ISRUC-S1 datasets, respectively containing data with healthy and unhealthy sleep patterns. Also, the model exhibited a high accuracy for sleep health prediction when evaluated on a combined dataset consisting of ISRUC and Sleep-EDF. Our model, which can effectively handle diverse sleep conditions, is the first to apply BiMamba to sleep staging with multimodal PSG data, showing substantial gains in computational and memory efficiency over traditional Transformer-style models. This method enhances sleep health management by making monitoring more accessible and extending advanced healthcare through innovative technology.

MSSC-BiMamba: Multimodal Sleep Stage Classification and Early Diagnosis of Sleep Disorders with Bidirectional Mamba

TL;DR

MSSC-BiMamba introduces an efficient, multimodal sleep staging framework by fusing Efficient Channel Attention with Bidirectional Mamba to capture long-range temporal patterns in PSG data. The model targets sleep stage classification and sleep-health discrimination, achieving strong performance on ISRUC-S3/S1 and Sleep-EDF datasets while reducing parameter count relative to Transformer-based approaches. Key contributions include a novel ECA-enabled channel weighting for time-series channels, a bidirectional Mamba module for forward and backward sequence modeling, and comprehensive evaluation across classification, cross-dataset validation, and health status discrimination. The approach promises practical clinical utility by enabling accurate, scalable, and accessible sleep monitoring and early disorder detection in real-world settings.

Abstract

Monitoring sleep states is essential for evaluating sleep quality and diagnosing sleep disorders. Traditional manual staging is time-consuming and prone to subjective bias, often resulting in inconsistent outcomes. Here, we developed an automated model for sleep staging and disorder classification to enhance diagnostic accuracy and efficiency. Considering the characteristics of polysomnography (PSG) multi-lead sleep monitoring, we designed a multimodal sleep state classification model, MSSC-BiMamba, that combines an Efficient Channel Attention (ECA) mechanism with a Bidirectional State Space Model (BSSM). The ECA module allows for weighting data from different sensor channels, thereby amplifying the influence of diverse sensor inputs. Additionally, the implementation of bidirectional Mamba (BiMamba) enables the model to effectively capture the multidimensional features and long-range dependencies of PSG data. The developed model demonstrated impressive performance on sleep stage classification tasks on both the ISRUC-S3 and ISRUC-S1 datasets, respectively containing data with healthy and unhealthy sleep patterns. Also, the model exhibited a high accuracy for sleep health prediction when evaluated on a combined dataset consisting of ISRUC and Sleep-EDF. Our model, which can effectively handle diverse sleep conditions, is the first to apply BiMamba to sleep staging with multimodal PSG data, showing substantial gains in computational and memory efficiency over traditional Transformer-style models. This method enhances sleep health management by making monitoring more accessible and extending advanced healthcare through innovative technology.
Paper Structure (16 sections, 11 equations, 8 figures, 7 tables)

This paper contains 16 sections, 11 equations, 8 figures, 7 tables.

Table of Contents

  1. Introduction
  2. Related work
  3. Classification of sleep stages
  4. State space model and Mamba
  5. Methods
  6. Overview of the framework
  7. The Efficient Channel Attention module
  8. Bidirectional Mamba
  9. Evaluation metrics
  10. Datasets
  11. Experimental setup
  12. Experiments and numerical analysis
  13. Classification experiments on the ISRUC-S3 and ISRUC-S1 datasets
  14. Cross-validation experiments
  15. Sleep health discrimination
  16. ...and 1 more sections

Figures (8)

  • Figure 1: The architecture of the proposed MSSC-BiMamba model and its two tasks: 1) sleep stage classification using PSG data; 2) leveraging sleep stage information to identify sleep disorders.
  • Figure 2: The structure of the ECA module.
  • Figure 3: The structure of the BSSM Block.
  • Figure 4: The confusion matrices for experiments on the ISRUC-S3 dataset: A, CNN+1bimamba; B, CNN+ECA+1bimamba; C, CNN+ECA+2bimamba; D, CNN+ECA+10bimamba.
  • Figure 5: Visualization comparison of sleep stage classifications between expert-labeled outcomes and predictions by our S3 model for participant 8 in the ISRUC-S3 dataset.
  • ...and 3 more figures