Personalized Sleep Staging System by combining Symbolic Fusion and Feedback System Control

Chen Chen (1, 2) , Adrien Ugon (1) , Xun Zhang (2) , Amara Amara (2) , Patrick Garda (1) , Jean-Gabriel Ganascia (3) , Carole Philippe (4) , Andrea Pinna (1)

1 SYEL - Systèmes Electroniques
2 LISITE - Laboratoire d'Informatique, Signal et Image, Electronique et Télécommunication
3 ACASA - Agents Cognitifs et Apprentissage Symbolique Automatique
4 CHU Pitié-Salpêtrière [AP-HP]

In this paper, a personalized sleep staging system is proposed by combining symbolic intelligence and feedback system control technique. Symbolic Fusion is dedicated to mimic decision-making process of clinical sleep staging. It starts from extraction of digital parameters from raw polysomnography (PSG) signals and goes up-to high level symbolic interpretation of feature parameters. At last, decision is generated using rules inspired by international guidelines in sleep medicine and applied to feature parameters. Meanwhile, Feedback System Control (FSC) is designed to provide optimal thresholds for Symbolic Fusion in transforming digital parameters into feature parameters which can take individual variability into consideration. Proposed system can be also integrated with portable PSG system to realize personal sleep evaluation or home care application in the future.

Domaines: Informatique [cs]

Chen Chen, Adrien Ugon, Xun Zhang, Amara Amara, Patrick Garda, et al.. Personalized Sleep Staging System by combining Symbolic Fusion and Feedback System Control. 11ème Colloque du GDR SoC-SiP, Jun 2016, Nantes, France. ⟨hal-01397723⟩