New framework improves sleep stage classification using multiple expert models

Researchers have developed a novel framework called Personalized Scorer Modeling (PSM) to improve the accuracy of sleep stage classification. This learning-based approach addresses the issue of inter-scorer variability by constructing more reliable reference labels from the collective behavior of multiple experts, rather than relying on a single reference hypnogram. PSM models the stage-specific behavior of each scorer using machine learning-derived confusion matrices, which are then aggregated to assign final epoch labels. Evaluations using random forest classifiers on the DOD-H and DOD-O datasets demonstrated significant improvements, achieving up to 86.07% accuracy. AI

IMPACT This framework could lead to more accurate diagnostic tools for sleep disorders by improving the reliability of automated sleep stage classification.

RANK_REASON The item is an academic paper detailing a new framework for a machine learning task. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.AI →

• AmirHossein Eshghi
• electroencephalography
• Electromyography
• Learning-Based Hypnogram
• multilayer perceptron
• Personalized Scorer Modeling
• random forest
• support vector machine
• U.S. Department of Defense

• paper
• other

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →