Tients in our sample suffered from positional sleep apnea or, conversely, there was no connection

Tients in our sample suffered from positional sleep apnea or, conversely, there was no connection involving the occurrence of events and sleep position, we calculated the percentage of time spent at each sleep angle (i.e., for each and every sleep angle, , in increments of 1 , we computed the percentage of time spent in a 15 window centered at that angle: 7.five ) and the percentage of events occurring at that angle (i.e., for each and every sleep angle, , we computed the percentage of apneas and hypopneas occurring in sleep angles of 7.5 ) [34]. Then, we compared the percentage of time as well as the percentage of events occurring at every single sleep angle and subtracted the two curves to establish irrespective of whether additional events than expected occurred at every single position. two.3.4. Oral vs. Nasal Breathing We calculated, for every topic, the percentage of time that was spent breathing by way of the mouth during the night utilizing an algorithm created by our group to distinguish among nasal and oral breathing from the spectral characteristics of acoustic breathing signals [32]. Audio signals have been segmented into ten s L-Kynurenine Autophagy non-overlapping sliding windows, and every window was classified into nasal or oral breathing. To complete so, the quick Fourier transform (FFT) was calculated, in addition to a linear envelope extracted working with windows of 15 Hz. When many of the energy of nasal breathing is concentrated within the low-frequency band, the spectrum of oral breathing presents a prominent peak amongst 950 and two kHz [32]. Therefore, when the height on the maximum peak inside the 950000 Hz band was, at the least, 60 with the maximum worth of the envelope (inside the 500000 Hz band, to avoid the variable effect of basal noise concentrated at decrease frequencies), then the window was labeled as oral breathing and otherwise viewed as nasal breathing [32]. Once all windows have been labeled, we calculated the percentage of windows classified as oral breathing. two.four. Statistical Analysis The options described in Section two.three. had been extracted for all subjects. Data were also averaged for the SCI and manage groups. Means, normal deviations (SDs), and ranges are reported for each group. Mann hitney U tests had been applied to examine the two groups (SCI vs. control), considering the fact that normality assumption was not met as outlined by Kolmogorov mirnov tests. Inside the SCI group, the Spearman correlation coefficient was used to investigate the connection among the extracted features (particularly AHI and SpO2 parameters) and age, BMI, injury level, AIS, and time post-injury. An alpha degree of 0.05 was employed to establish significance for all statistical tests. three. Benefits The extracted functions are displayed in Table 2 for all SCI individuals, while a summary on the values for each group (SCI vs. manage) and the p-values from the statistical comparisons are presented in Table three. Under we describe in far more detail the results in terms of oxygen saturation, apneas, and hypopneas detected from acoustic signals, sleep position measured from accelerometer data, and prevalence of oral breathing.Sensors 2021, 21,8 ofTable 2. Outcome measures for all SCI sufferers. R428 Technical Information sufferers with mild sleep apnea (five AHI 15) are highlighted in yellow, individuals with moderate sleep apnea (15 AHI 30) in orange, and sufferers with extreme sleep apnea (AHI 30) in red.Patient ID SCI 1 SCI 2 SCI 3 SCI 4 SCI five SCI 6 SCI 7 SCI 8 SCI 9 SCI ten SCI 11 SCI 12 SCI 13 SCI 14 SCI 15 SCI 16 SCI 17 SCI 18 SCI 19 Awake SpO2 98 98 98 95 95 99 95 92 98 96 98 98 99 97 94 94 96 99 95 Median SpO2 95 98 95 95 94 92 95 90 96 93.