Detection of obstructive sleep apnea by analysis of phase angle using the forced oscillation signal

Pharyngeal collapse in patients with obstructive sleep apnea syndrome (OSAS ) is linked to decreased upper airway muscle activity. We hypothesised that decreased muscle activity causes decreased stiffness of the upper airway w all and assumed that a decrease in wall stiffness would result in a change point (CP) of the morphology of phase angle time series phi(t) obtained by forced oscillation technique (FOT). We developed an algorithm to detect CPs in phi(t) and investigated phi(t) data measured in parallel to all-night p olysomnography in seven patients. A total of 2724 CPs were detected by algo rithm. The CPs were marked on the polysomnograms and compared with polysomn ogram scoring performed jointly by three sleep expert physicians. A total o f 791 (67%) of the 1172 handscored respiratory events showed a CP in phi(t) within a time interval of 8 sec before and 6 sec after the marked onset. A total of 672 (57%) respiratory events were detected at an earlier time by algorithm, and 119 (10%) were algorithmically detected later. The average d etection time by the CP algorithm was 4.0 +/- 1.1 sec earlier than the manu al scoring. We conjecture that a CP in phi(t) indicates a change in upper a irway collapsibility and that phi(t) might be a potentially useful paramete r for detection of impending upper airway obstruction. (C) 2000 Elsevier Sc ience B.V. All rights reserved.