Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

AI Generalisation Gap In Comorbid Sleep Disorder Staging

Saswata Bose, Suvadeep Maiti, Shivam Kumar Sharma, Mythirayee S, Tapabrata Chakraborti, Srijitesh Rajendran, Raju S. Bapi

Abstract

Accurate sleep staging is essential for diagnosing OSA and hypopnea in stroke patients. Although PSG is reliable, it is costly, labor-intensive, and manually scored. While deep learning enables automated EEG-based sleep staging in healthy subjects, our analysis shows poor generalization to clinical populations with disrupted sleep. Using Grad-CAM interpretations, we systematically demonstrate this limitation. We introduce iSLEEPS, a newly clinically annotated ischemic stroke dataset (to be publicly released), and evaluate a SE-ResNet plus bidirectional LSTM model for single-channel EEG sleep staging. As expected, cross-domain performance between healthy and diseased subjects is poor. Attention visualizations, supported by clinical expert feedback, show the model focuses on physiologically uninformative EEG regions in patient data. Statistical and computational analyses further confirm significant sleep architecture differences between healthy and ischemic stroke cohorts, highlighting the need for subject-aware or disease-specific models with clinical validation before deployment. A summary of the paper and the code is available at https://himalayansaswatabose.github.io/iSLEEPS_Explainability.github.io/

AI Generalisation Gap In Comorbid Sleep Disorder Staging

Abstract

Accurate sleep staging is essential for diagnosing OSA and hypopnea in stroke patients. Although PSG is reliable, it is costly, labor-intensive, and manually scored. While deep learning enables automated EEG-based sleep staging in healthy subjects, our analysis shows poor generalization to clinical populations with disrupted sleep. Using Grad-CAM interpretations, we systematically demonstrate this limitation. We introduce iSLEEPS, a newly clinically annotated ischemic stroke dataset (to be publicly released), and evaluate a SE-ResNet plus bidirectional LSTM model for single-channel EEG sleep staging. As expected, cross-domain performance between healthy and diseased subjects is poor. Attention visualizations, supported by clinical expert feedback, show the model focuses on physiologically uninformative EEG regions in patient data. Statistical and computational analyses further confirm significant sleep architecture differences between healthy and ischemic stroke cohorts, highlighting the need for subject-aware or disease-specific models with clinical validation before deployment. A summary of the paper and the code is available at https://himalayansaswatabose.github.io/iSLEEPS_Explainability.github.io/
Paper Structure (14 sections, 5 figures, 2 tables)

This paper contains 14 sections, 5 figures, 2 tables.

Table of Contents

  1. Introduction
  2. Methodology
  3. Model Architecture
  4. Explainability Methods
  5. Experiments
  6. Results and Discussion
  7. Model Performance
  8. Ablation studies
  9. Model Explainability and Clinician Feedback
  10. Clincian-in-the-loop Analysis of Sleep Transition Graphs
  11. Conclusion
  12. Ethical Compliance
  13. Acknowledgements
  14. Conflict of Interest

Figures (5)

  • Figure 1: Deep Model Architecture, with the $n$th epoch being input to the model, having a window size of $w$ and a stride length of $s$.
  • Figure 2: Medically Annotated Epochs from iSLEEPS showing well-focussed and ill-focussed sleep-relevant artefacts based on the model trained on SHHS and tested on iSLEEPS. The title also includes the probability of the predicted class provided by the model in brackets.
  • Figure 3: GradCAM visualization of raw EEG epoch of sleep stage N2 (correctly predicted by the model) and boxes in green indicating K-complex and spindles, extracted from the model trained and tested on iSLEEPS.
  • Figure 4: Transition graph of the dataset (excluding insignificant transitions and self transitions) elucidating the probabilities of transition from one state to the other for the corresponding dataset. The nodes represent corresponding sleep stages
  • Figure 5: Distribution of the Average Run Length in the Datasets. The Violin Chart clearly highlights that the iSLEEPS dataset has a much wider range for the feature while SleepEDF-78 is more constrained within a smaller range of values.