Staging by the Book: Automatic Sleep Stage Classification Using Scoring Rules
A new method for automated sleep staging has been proposed, focusing on transparency and adherence to clinical scoring rules. This rule-based approach operationalizes the American Academy of Sleep Medicine's scoring logic and provides natural language justifications. While it shows lower agreement with reference standards compared to deep learning models, it offers deterministic decisions that can aid in auditing and debugging.
- ▪The proposed method achieves 60.5% agreement with a majority-vote reference across 50 polysomnography recordings.
- ▪Higher agreement was noted during development, with 77.1% accuracy on a specific dataset.
- ▪The method's recall rates were highest for sleep stage N2 at 83.5% and moderate for stage R at 68.7%.
Opening excerpt (first ~120 words) tap to expand
Electrical Engineering and Systems Science > Signal Processing arXiv:2605.22859 (eess) [Submitted on 19 May 2026] Title:Staging by the Book: Automatic Sleep Stage Classification Using Scoring Rules Authors:Emil Hardarson, Konstantin Popov, Sigridur Sigurdardottir, Anna Sigridur Islind, Erna Sif Arnardóttir, María Óskarsdóttir View a PDF of the paper titled Staging by the Book: Automatic Sleep Stage Classification Using Scoring Rules, by Emil Hardarson and 5 other authors View PDF HTML (experimental) Abstract:Automated sleep staging is commonly approached as a supervised machine learning problem, with deep learning methods dominating recent research.
…
Excerpt limited to ~120 words for fair-use compliance. The full article is at arXiv cs.AI.
Discussion
0 commentsMore from arXiv cs.AI
