A Clinically Applicable Interactive Micro and Macro-Sleep Staging Algorithm for Elderly and Patients with...

Area of Science:

• Neuroscience
• Sleep Medicine
• Biomedical Engineering

Background:

• Elderly individuals and patients with neurodegenerative diseases (NDD) frequently experience sleep disturbances and altered sleep architecture.
• Accurate and efficient sleep staging is crucial for diagnosing and managing these conditions.
• Current manual sleep staging is time-consuming and requires specialized expertise.

Purpose of the Study:

• To develop and validate a novel, automated algorithm for sleep staging in elderly healthy controls and patients with Parkinson's disease (PD), isolated REM sleep behavior disorder (iRBD), and PD with RBD (PD+RBD).
• To assess the algorithm's performance in both standard 30-second epochs (macro-staging) and 5-second mini-epochs (micro-staging).
• To introduce an interactive feature that identifies uncertain sleep epochs for expert manual review, optimizing accuracy and efficiency.

Main Methods:

• The algorithm utilizes a minimal channel set: one electroencephalographic (EEG) and two electrooculographic (EOG) channels.
• It performs both macro-staging (30-s epochs) and micro-staging (5-s epochs) for wakefulness, REM sleep, and non-REM sleep.
• Performance was evaluated on cohorts of healthy controls (HC), PD, iRBD, and PD+RBD patients, comparing overall accuracy with accuracy on 'certain' epochs.

Main Results:

• The algorithm achieved high accuracies for macro-sleep staging across all groups, with overall accuracies ranging from 0.76 to 0.87.
• Accuracies increased significantly when considering only 'certain' epochs, reaching up to 0.91 in HC.
• The proportion of uncertain epochs requiring manual review was relatively low, ranging from approximately 10% in HC to 19% in PD+RBD patients.

Conclusions:

• The proposed interactive sleep staging algorithm demonstrates high accuracy and efficiency in classifying sleep stages for elderly individuals and NDD patients.
• Its ability to perform both macro and micro-staging, coupled with an interactive review feature, offers a valuable tool for clinical settings.
• This automated approach can significantly reduce the manual workload associated with sleep staging in vulnerable patient populations.