A Pilot Study of Detecting Individual Sleep Apnea Events Using Noncontact Radar Technology, Pulse Oximetry, and Machine Learning

<table class="table-group" id="tab3"><tr><td><table class="table"><tr><td class="thead-hr" colspan="5"><hr/></td></tr><tr class="thead"><td class="align_left"></td><td class="align_center">5 class <i>κ</i></td><td class="align_center">3 class <i>κ</i></td><td class="align_center">2 class <i>κ</i></td><td class="align_center">REI 95% LoA</td></tr><tr><td class="thead-hr" colspan="5"><hr/></td></tr><tr><td class="align_left">Only pulse oximetry</td><td class="align_center">0.28</td><td class="align_center">0.34</td><td class="align_center">0.71</td><td class="align_center">-13.96, 11.18</td></tr><tr><td class="align_left">Only radar</td><td class="align_center">0.26</td><td class="align_center">0.34</td><td class="align_center">0.78</td><td class="align_center">-14.48, 9.98</td></tr><tr><td class="align_left">Somnofy+SpO<sub>2</sub></td><td class="align_center">0.36</td><td class="align_center">0.43</td><td class="align_center">0.81</td><td class="align_center">-8.25, 7.47</td></tr><tr class="table-tr"><td colspan="5"><hr class="tbody-hr"/></td></tr></table></td></tr><tr class="table-fn"><td><div><i>κ</i> = Cohen’s kappa for the whole dataset on seconds resolution, 5 class = no event/hypopnea/obstructive apnea/central apnea/mixed apnea, 3 class = no event/hypopnea/apnea, 2 class = no event/event, and REI 95% LoA = Bland-Altman 95% limits of agreement for REI.<br/></div></td></tr></table>

<div>Agreement for only pulse oximetry, only radar, and the combination of radar and pulse oximetry (Somnofy+SpO<sub>2</sub>) against HSAT.</div>

Journal of Sensors

tab3

Table 3

Table 3: A Pilot Study of Detecting Individual Sleep Apnea Events Using Noncontact Radar Technology, Pulse Oximetry, and Machine Learning 