Bridging AI and Clinical Practice: Integrating Automated Sleep Scoring Algorithm with Uncertainty-Guided Physician Review.

Michal BechnyGiuliana Monachino Luigi FiorilloJulia van der MeerMarkus H SchmidtClaudio L A BassettiAthina TzovaraFrancesca D Faraci

Published in: Nature and science of sleep (2024)

Inter-scorer variability limits the accuracy of the scoring algorithms to ~80%. By integrating an uncertainty estimation with U-Sleep, we enhance the review of predicted hypnograms, to align with the scoring taste of a responsible physician. Validated across ID and OOD data and various sleep-disorders, our approach offers a strategy to boost automated scoring tools' usability in clinical settings.

Keyphrases

machine learning
deep learning
sleep quality
physical activity
clinical practice
primary care
artificial intelligence
emergency department
high throughput
big data
single cell
social media
neural network