Detecting sleep apnea using non-linear measures of heart rate variability

Figures

• 

  Fig. 1. Patient selection pipeline for the analysis

• 

  Fig. 2. Scale-dependent sDFA exponent in severe apnea and healthy male participants. sDFA scaling exponent (\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\alpha $$\end{document}) as a function of scale for severe apnea and healthy male participants without (a) and with (b) cardiovascular disease (CVD). For both groups (no CVD: \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$n=101$$\end{document}; CVD: \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$n=47$$\end{document}), \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek}…

• 

  Fig. 3. Scale-dependent AUC of sDFA across apnea severity and CVD status. Area under the receiver operating characteristic curve (AUC) of the sDFA scaling exponent (\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\alpha $$\end{document}) as a function of scale for male participants without (a) and with (b) CVD, shown separately for all apnea severity groups. In both groups, two principal AUC peaks are observed: an initial local maximum of approximately 0.72 at intermediate scales (\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$s \approx 30$$\end{document}–45), followed by a decline toward chance level (AUC \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt}…

• 

  Fig. 4. Discriminative performance of sDFA and HRV metrics across apnea severity and CVD. AUC scores for sDFA and conventional HRV (cHRV) metrics across apnea severity levels and CVD status in male participants. Across all subgroups, sDFA achieves the highest discriminative performance, with AUC values ranging from 0.56 to 0.81 and a maximum AUC of 0.81 observed in the severe apnea group. In contrast, no conventional HRV metric consistently emerges as the second-best performer, with performance varying across severity and CVD subgroups. Notably, root mean square of successive differences (RMSSD) shows relatively strong performance in participants with severe apnea and CVD, reaching an AUC of approximately 0.69