[home] [Personal Program] [Help]

---

Session: Poster session I
Session starts: Thursday 24 January, 15:00

---

Margot Deviaene (KU Leuven)
Jesús Lázaro (University of Connecticut)
Dorien Huysmans (KU Leuven)
Dries Testelmans (UZ Leuven)
Bertien Buyse (UZ Leuven)
Sabine Van Huffel (KU Leuven)
Carolina Varon (KU Leuven)

Abstract:
Aim: Obstructive sleep apnea (OSA) often remains undiagnosed. Therefore, this study investigates the use of pulse photoplethysmography (PPG) for the detection of sleep apnea and its added value to oxygen saturation (SpO2). Methods: A dataset of 102 subjects, suspected of having OSA was recorded, the data was split in 1 minute segments to be classified as apneic or normal breathing. The PPG signals were preprocessed and 6 PPG series were extracted: the pulse rate, amplitude and width variabilities, slope transit time, maximal pulse upslope and the area under the PPG pulse [1]. Moreover, the instantaneous powers in the high and low frequency bands of the pulse rate were estimated using a point-process model [2]. For all these series, 5 features were computed over a 1 minute interval: the mean, the minimum and the maximum value, the standard deviation and the gradient. Feature selection resulted in the 6 most discriminative features for PPG-based detection of apneic intervals. These features were used as input for a least-squares support vector machine classifier. A second classifier was trained including SpO2 features extracted as described in [3]. Results: A classification accuracy of 68.7 % was achieved for PPG-based apnea detection. If only the most severe events with complete breathing cessations are considered, this value rises to 74.8 %. When SpO2 features were added to the classifier the accuracy increased to 83.4 %, which is only slightly higher than the 82.2 % obtained using only SpO2. These results suggest that the studied PPG features have potential for sleep apnea detection, and that, however, their added value to SpO2 is limited. References [1] Lázaro J. et al. Pulse photoplethysmography derived respiration for obstructive sleep apnea detection. Computing in cardiology 44 2017;1. [2] Barbieri R. et al. A point-process model of human heartbeat intervals: new definitions of heart rate and heart rate variability. Am J Physiol Heart Circ Physiol 2005;288.1;H424-H435. [3] Deviaene M. et al. Automatic screening of sleep apnea patients based on the SpO2 signal. IEEE JBHI 2018; Early Access; DOI:10.1109/JBHI.2018.2817368.