A Portable Automated Assessment Tool for Sleep Apnea Using a Combined Holter-Oximeter

Heneghan, Conor; Chua, Chern-Pin; Garvey, John; Chazal, Philip de; Shouldice, R.; Boyle, Patricia; McNicholas, Walter T.

doi:10.5665/sleep/31.10.1432

Cited by 20 publications

(10 citation statements)

References 35 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…However, most include statistical pattern recognition based on characteristics extracted from single-lead electrocardiogram (ECG) signals [17][18][19][20][21][22][23][24][25][26][27][28][29] and blood oxygen saturation (SpO 2 ) , which are measured by a pulse oximeter. In some cases, the single-lead ECG signal is combined with the SpO 2 signal [56][57][58][59][60][61].…”

Section: Motivation and Problem Descriptionmentioning

confidence: 99%

“…There are two groups of studies in which the single-lead ECG signal is combined with the SpO 2 signal [56][57][58][59][60][61]. In the first, research focuses on AHI estimation and, thus, on OSA diagnosis (per-recording classification) [56,86].…”

Section: Cheng Et Al [77]mentioning

confidence: 99%

“…They also proposed a correlation classifier to correlate the ECG signal and the SpO 2 data, resulting in an estimated AHI. In the second group, the studies carry out a real-time analysis aiming to detect apneic episodes on a minute-by-minute basis (per-segment classification) [57,58]. Heneghan et al [57] put forward an algorithm that provided an estimation of apnea occurrences epoch-by-epoch and estimated an overall per-recording AHI.…”

Section: Cheng Et Al [77]mentioning

confidence: 99%

“…In the second group, the studies carry out a real-time analysis aiming to detect apneic episodes on a minute-by-minute basis (per-segment classification) [57,58]. Heneghan et al [57] put forward an algorithm that provided an estimation of apnea occurrences epoch-by-epoch and estimated an overall per-recording AHI. De Chazal et al [58] proposed an algorithm to identify epochs as normal or sleep-disordered breathing and to classify the pathological ones into one of six classes: obstructive, mixed, and central apnea, as well as obstructive, mixed, and central hypopnea.…”

Section: Cheng Et Al [77]mentioning

confidence: 99%

“…The different colors represent the control subjects (green), the desaturators (red), and the non-desaturators (blue). There is no defined criterion for the AHI value to discriminate between healthy and pathological subjects [44], so there are three horizontal lines in each graph representing AHI values commonly used in the OSA diagnosis-5, 10, or 15 [41,57]-thus calculating the results according to the AHI value. We perform the per-recording classification analysis graphically, as it is a good way to assess how our system's regression line fits the function y = x.…”

Section: Per-recording Classificationmentioning

confidence: 99%

See 4 more Smart Citations

Combining Heart Rate Variability and Oximetry to Improve Apneic Event Screening in Non-Desaturating Patients

Martin-Gonzalez

Ravelo-García

Navarro-Mesa

et al. 2023

Sensors

View full text Add to dashboard Cite

In this paper, we thoroughly analyze the detection of sleep apnea events in the context of Obstructive Sleep Apnea (OSA), which is considered a public health problem because of its high prevalence and serious health implications. We especially evaluate patients who do not always show desaturations during apneic episodes (non-desaturating patients). For this purpose, we use a database (HuGCDN2014-OXI) that includes desaturating and non-desaturating patients, and we use the widely used Physionet Apnea Dataset for a meaningful comparison with prior work. Our system combines features extracted from the Heart-Rate Variability (HRV) and SpO2, and it explores their potential to characterize desaturating and non-desaturating events. The HRV-based features include spectral, cepstral, and nonlinear information (Detrended Fluctuation Analysis (DFA) and Recurrence Quantification Analysis (RQA)). SpO2-based features include temporal (variance) and spectral information. The features feed a Linear Discriminant Analysis (LDA) classifier. The goal is to evaluate the effect of using these features either individually or in combination, especially in non-desaturating patients. The main results for the detection of apneic events are: (a) Physionet success rate of 96.19%, sensitivity of 95.74% and specificity of 95.25% (Area Under Curve (AUC): 0.99); (b) HuGCDN2014-OXI of 87.32%, 83.81% and 88.55% (AUC: 0.934), respectively. The best results for the global diagnosis of OSA patients (HuGCDN2014-OXI) are: success rate of 95.74%, sensitivity of 100%, and specificity of 89.47%. We conclude that combining both features is the most accurate option, especially when there are non-desaturating patterns among the recordings under study.

show abstract