Identification of apnea events using a chest-worn monitor compared to laboratory-based polysomnography in patients suspected of obstructive sleep apnea

Obstructive sleep apnea (OSA) is an under-diagnosed risk factor for several adverse health outcomes. The gold standard diagnostic test for OSA is laboratory-based polysomnography (PSG). Portable sleep monitoring has been studied as an alternative for patients lacking access to PSG. This study aimed to assess the validity of the Zephyr BioHarness 3 (BH3), a chestworn activity monitor that records movement, electrocardiography, and respiratory parameters, to identify apnea events in patients suspected of OSA. Patients (N = 18) underwent single-night laboratory-based PSG while wearing the BH3. PSG data were scored in 30-second epochs by PSG technicians. PSG and BH3 data were sampled and analyzed using three sets of features with a radial basis function support vector machine and three-layer neural networks: (1) apnea events were identified second by second using 5-second windows of raw BH3 data (sensitivity = 48.0 ± 8.7%, specificity = 75.6 ± 3.0%, accuracy = 74.4 ± 2.7%); (2) apnea events were identified second by second using mean, median, and variance values of 5-second windows of BH3 data (sensitivity = 54.7 ± 17.3%, specificity = 66.5 ± 12.1%, accuracy = 66.0 ± 10.9%); and (3) apnea events were identified second by second using phase-space transformation of BH3 data (sensitivity = 68.4 ± 9.0%, specificity = 81.5 ± 2.7%, accuracy = 80.9 ± 2.5% for τ = 60; sensitivity = 64.0 ± 7.9%, specificity = 81.8 ± 2.5%, accuracy = 81.0 ± 2.3% for τ = 70). The BH3 may be useful for patients suspected of OSA without timely access to PSG.