0279 Intermittent Analysis of Obstructive Sleep Apnea Dynamics

Trung Le,Huynh Phat,Arveity Setty,Abhay Sharma
DOI: https://doi.org/10.1093/sleep/zsae067.0279
IF: 6.313
2024-04-20
SLEEP
Abstract:Abstract Introduction The intermittent nature of obstructive sleep apnea's (OSA) pathophysiological processes is characterized by unpredictable fluctuations among OSA endotypes and phenotypes, reflecting the complex dynamics underlying this sleep disorder. It is challenging to character such intermittent behaviors. Methods We developed an enhanced version of Koopman spectral analysis, tailored to be robust against noise, a common issue in physiological data, to investigate the nonlinear intermittent behaviors in OSA. Our analysis dissects the complex, nonlinear physiological patterns of Heart Rate Variability (HRV) features, aiming to understand better the cardiorespiratory dynamic nature of OSA and its severity, as indicated by the apnea-hypopnea index (AHI). Utilizing the Apnea-ECG Database from Physionet.org, which includes records from 70 OSA patients with OSA annotation, we validated the model in the ability to characterize the intermittency behavior of the HRV features and the accuracy of using these features to detect OSA. Results The results demonstrate that the noise-robust Koopman spectral analysis effectively identifies key dynamical features in the pathophysiology of OSA. It revealed several previously unrecognized patterns and correlations in the data, offering more profound insights into the intermittent dynamics of OSA. The method showed high accuracy in distinguishing critical transitions between different severity levels of OSA. Conclusion The study successfully validates the efficacy of the noise-robust Koopman spectral analysis in analyzing complex systems like OSA. This approach provides an effective tool for understanding the intricate dynamics of pathophysiological processes, offering the potential for both diagnostic advancements and a better understanding of the underlying mechanisms of OSA. Support (if any) Huynh, P. K., Setty, A. R., Le, T. B., & Le, T. Q. (2023). A noise-robust Koopman spectral analysis of an intermittent dynamics method for complex systems: a case study in pathophysiological processes of obstructive sleep apnea. IISE Transactions on Healthcare Systems Engineering, 13(2), 101-116.
neurosciences,clinical neurology
What problem does this paper attempt to address?