An AI-supported diagnostic tool for obstructive sleep apnea patients based on delta-alpha connectivity at the sensorimotor cortex [Abstract]

  • Background: The modulation of delta-alpha phase-amplitude cross-frequency coupling (PACFC) may influence information processing throughout the human cerebral cortex. We investigated whether this frequency band-specific modulation is impaired in patients with obstructive sleep apnea (OSA). Patients & Methods: In this study, the C3- and C4- electroencephalographic recordings of 170 participants (86 in main dataset: age 27-84 years, 44 subjects had moderate or severe OSA with respiratory disturbance index RDI>15/h; 84 in validation dataset: age 35 -75 years, 42 subjects with RDI>15/h) who underwent full-night polysomnography (PSG) were evaluated. We tested if the delta-alpha PACFC modulation index (MI) at the sensorimotor cortex differs between OSA patients with RDI>15/h and those with RDI≤15/h in distinct sleep stages. Further, by making use of a Support Vector Machine (SVM) algorithm, we tested if the sleep stage – specific MIs could predict RDI values of OSA patients. Results: InBackground: The modulation of delta-alpha phase-amplitude cross-frequency coupling (PACFC) may influence information processing throughout the human cerebral cortex. We investigated whether this frequency band-specific modulation is impaired in patients with obstructive sleep apnea (OSA). Patients & Methods: In this study, the C3- and C4- electroencephalographic recordings of 170 participants (86 in main dataset: age 27-84 years, 44 subjects had moderate or severe OSA with respiratory disturbance index RDI>15/h; 84 in validation dataset: age 35 -75 years, 42 subjects with RDI>15/h) who underwent full-night polysomnography (PSG) were evaluated. We tested if the delta-alpha PACFC modulation index (MI) at the sensorimotor cortex differs between OSA patients with RDI>15/h and those with RDI≤15/h in distinct sleep stages. Further, by making use of a Support Vector Machine (SVM) algorithm, we tested if the sleep stage – specific MIs could predict RDI values of OSA patients. Results: In both datasets, in OSA patients with RDI >15/h, the delta-alpha CFC-MI was significantly (p< 0.05) reduced at the sensorimotor cortex during REM and NREM1 stages, while increased during NREM2 compared to patients with RDI ≤15/h. In addition, the delta-alpha MI in REM sleep stage could provide with use of an SVM algorithm a quite reliable (82% accuracy) prediction of the RDI in OSA patients. Conclusions: This increase in disconnection at the cortical sensorimotor areas with increasing respiratory distress during sleep further supports the concept of a cortical sensorimotor dysfunction in OSA patients. Additionally, the delta – alpha MI during REM sleep may provide an objective neurophysiologic surrogate marker of respiratory distress in OSA patients.show moreshow less

Download full text files

Export metadata

Statistics

Number of document requests

Additional Services

Share in Twitter Search Google Scholar
Metadaten
Author:Haralampos Gouveris, Nabin Koirala, AbdulRauf Anwar, Hao Ding, Katharina Ludwig, Katharina Bahr, Tilman Huppertz, Christoph Matthias, Sergiu Groppa, Muthuraman MuthuramanORCiDGND
URN:urn:nbn:de:bvb:384-opus4-1096897
Frontdoor URLhttps://opus.bibliothek.uni-augsburg.de/opus4/109689
ISSN:1438-8685OPAC
Parent Title (English):Laryngo-Rhino-Otologie
Publisher:Georg Thieme
Place of publication:Stuttgart
Type:Article
Language:English
Year of first Publication:2022
Publishing Institution:Universität Augsburg
Release Date:2023/12/04
Volume:101
Issue:S02
First Page:S243
Last Page:S244
DOI:https://doi.org/10.1055/s-0042-1747042
Institutes:Fakultät für Angewandte Informatik
Fakultät für Angewandte Informatik / Institut für Informatik
Fakultät für Angewandte Informatik / Institut für Informatik / Professur für Informatik in der Medizintechnik
Dewey Decimal Classification:0 Informatik, Informationswissenschaft, allgemeine Werke / 00 Informatik, Wissen, Systeme / 004 Datenverarbeitung; Informatik
Licence (German):CC-BY-NC-ND 4.0: Creative Commons: Namensnennung - Nicht kommerziell - Keine Bearbeitung (mit Print on Demand)