3D Camera and Pulse Oximeter for Respiratory Events Detection

Objective: The purpose of this study was to derive a respiratory movement signal from 3D time-of-flight camera and investigate if it can be used in combination with SpO 2 detect events comparable polysomnography (PSG) based detection. xmlns:xlink="http://www.w3.org/1999/xlink">Methods: We derived developed new algorithm that detects reduced desaturation score events. method tested on 61 patients’ synchronized video PSG recordings. predicted apnea-hypopnea index (AHI), calculated total sleep time, severity were compared manual annotations (manualPSG). Predicted AHI evaluation, measured by intraclass correlation (ICC), classification performed. Furthermore, the results evaluated 30-second epoch analysis, labelled either as event or normal breathing, wherein accuracy, sensitivity, specificity Cohen's kappa calculated. xmlns:xlink="http://www.w3.org/1999/xlink">Results: scored an ICC xmlns:xlink="http://www.w3.org/1999/xlink">r = 0.94 (0.90 – 0.96 at 95% confidence interval, xmlns:xlink="http://www.w3.org/1999/xlink">p < 0.001) manualPSG. Severity 80% no misclassification more than one level. Based 0.72, accuracy 0.88, sensitivity 0.80, 0.91. xmlns:xlink="http://www.w3.org/1999/xlink">Conclusion: Our detection using had excellent reliability substantial agreement PSG-based scoring. xmlns:xlink="http://www.w3.org/1999/xlink">Significance: This showed potential reliably without airflow belt sensors, sensors uncomfortable patients susceptible artefacts.