Obstructive Sleep Apnea Screening using Nonlinear Characteristics of Overnight Pulse Oximetry Recordings

摘要

Nonlinear signal processing methods were applied to oxygen saturation signals obtained from overnight pulse oximetry for 84 patients diagnosed for obstructive sleep apnea (OSA) and 58 healthy controls. The central tendency measure (CTM) was calculated and used to classify patients and controls. A modified CTM approach was developed in order to include the quality of apnea events into the evaluation. With a sensitivity of 100% and a specificity of 98.6% our study showed that nonlinear characteristics can provide valuable additional information for OSA diagnosis.