Heart rate regularity analysis obtained from pulse oximetric recordings in the diagnosis of obstructive sleep apnea

Abstract

Approximate entropy (ApEn) is a technique that can be used to quantify the irregularity or variability of time series. We prospectively evaluated the validity of ApEn of heart rate data obtained from pulse oximetric recordings as a diagnostic test for obstructive sleep apnea (OSA) in patients clinically suspected of suffering this disease. A sample of 187 referred outpatients (147 men and 40 women), with a mean age of 57.9+/-12.8 years and a body mass index of 29.5+/-5.5 kg/m(2), clinically suspected of having OSA were studied using nocturnal pulse oximetric recording performed simultaneously with complete polysomnography. A diagnosis of OSA was confirmed in 111 (59.3%). Patients with OSA presented significantly higher ApEn levels than those without OSA (1.334+/-0.189 vs 1.167+/-0.182). Chronic obstructive pulmonary disease (COPD) was diagnosed for 42 patients. Among these patients, 22 (52.4%) were diagnosed with OSA. COPD patients with OSA showed significantly higher ApEn levels than COPD patients without OSA (1.337+/-0.193 vs 1.184+/-0.173; p=0.01). ApEn correlated significantly with apnea-hypopnea index (r=0.38; p=0.000). There was no significant correlation between ApEn and either age or body mass index. No significant changes were observed in ApEn throughout the night in OSA patients. Using receiver operating characteristic curve analysis, we obtained a diagnostic sensitivity of 71.2%, specificity of 78.9%, positive predictive value of 81.3%, and negative predictive value of 66% at a threshold of 1.272. We conclude that ApEn analysis of heart rate data obtained from pulse oximetric recordings could be a useful tool in the study of OSA.

Fractal Based Analysis of the Influence of Odorants on Heart Activity

“…Based on their founding, head-up tilt resulted in a significant reduction in sample entropy of R-R intervals and cross-sample entropy, while mental arithmetic stress resulted in a significant reduction in coupling directed from R-R to QT. See also282930.…”

Influence of Chronic Obstructive Pulmonary Disease and Moderate-To-Severe Sleep Apnoea in Overnight Cardiac Autonomic Modulation: Time, Frequency and Non-Linear Analyses

“…Similarly, the entropy of respiratory sounds has been found to be higher in non-healthy subjects [ 36 ] and the entropy of mechanomyogram signals is also higher as the severity of COPD increases [ 52 ]. In the presence of OSAS, different studies reported higher non-linear measures values of overnight oximetry [ 33 , 34 ], airflow [ 35 ], and heart rate [ 21 , 26 , 53 ] derived from whole-night PSG. In these studies, it has been found that direct presence of apnoeic events modifies the normal cardiorespiratory dynamics towards higher disorderliness or disorganization, i.e., higher entropy.…”

“…It reflects changes with periodicities between 30 s and 70 s (0.014–0.033 Hz). The power spectrum in this frequency band has been found to be related to the repetition of apnoeic events during the night [ 26 , 28 ]. It is usually normalised to the total signal power (OSASFn).…”

“…Moreover, SampEn has been previously used to analyse cardiorespiratory-related signals in the context of OSAS [ 20 , 21 , 22 ] and COPD [ 12 , 23 , 24 , 25 ]. Nevertheless, few studies have used non-linear measures to assess changes in HRV due to the simultaneous presence of COPD and OSAS [ 9 , 26 ]. In the present study, we hypothesized that SampEn could provide additional and relevant information on the changes in overnight HRV dynamics of COPD patients due to the presence of concomitant moderate-to-severe OSAS.…”

Accuracy of ECG-based screening for sleep-disordered breathing: a survey of all male workers in a transport company

“…The ACAT algorithm also showed a good performance [AUC, 0.979 (95 % CI, 0.912–0.998); 90 % sensitivity and 100 % specificity] for this database [19]. Although only a few studies have examined the performance of the algorithms in clinical settings, they have reported modest diagnostic accuracies [18, 20, 29]. In a study of 150 patients referred to a university hospital for clinically suspected SDB, Roche et al [18] reported that an algorithm that used the relative power of the very low frequency component detected patients with AHI ≥15 with an AUC of 0.70, a sensitivity of 64 %, and a specificity of 69 %.…”