Apnea-ECG Database Research Articles

Apnea-Hypopnea Index Prediction Using Electrocardiogram Acquired During the Sleep-Onset Period.

The most widely used methods for predicting obstructive sleep apnea are based on clinical or anatomico-functional features. To improve exactitude in obstructive sleep apnea screening, this study aimed to devise a new predictor of apnea-hypopnea index. We hypothesized that less irregular respiration cycles would be observed in the patients with more severe obstructive sleep apnea during the sleep-onset period. From each of the 156 and 70 single-lead electrocardiograms collected from the internal polysomnographic database and from the Physionet Apnea-ECG database, respectively, the 150-s sleep-onset period was determined and the respiration cycles during this period were detected. Using the coefficient of variation of the respiration cycles, obtained from the internal dataset, as a predictor, the apnea-hypopnea index predictive model was developed through regression analyses and k-fold cross-validations. The apnea-hypopnea index predictability of the regression model was tested with the Physionet Apnea-ECG database. The regression model trained and validated from the 143 and 13 data, respectively, produced an absolute error (mean ± SD) of 3.65 ±2.98 events/h and a Pearson's correlation coefficient of 0.97 (P < 0.01) between the apnea-hypopnea index predictive values and the reference values for the 70 test data. The new predictor of apnea-hypopnea index has the potential to be utilized in making more reasoned clinical decisions on the need for formal diagnosis and treatment of obstructive sleep apnea. Our study is the first study that presented the strategy for providing a reliable apnea-hypopnea index without overnight recording.

Development of an apnea detection algorithm based on temporal analysis of thoracic respiratory effort signal

This work describes the design of an algorithm for detecting apnea episodes, based on analysis of thorax respiratory effort signal. Inspiration and expiration time, and range amplitude of respiratory cycle were evaluated. For range analysis the standard deviation statistical tool was used over respiratory signal temporal windows. The validity of its performance was carried out in 8 records of Apnea-ECG database that has annotations of apnea episodes. The results are: sensitivity (Se) 73%, specificity (Sp) 83%. These values can be improving eliminating artifact of signal records.

Spectral Heart Rate Variability analysis using the heart timing signal for the screening of the Sleep Apnea–Hypopnea Syndrome

Some approaches have been published in the past using Heart Rate Variability (HRV) spectral features for the screening of Sleep Apnea–Hypopnea Syndrome (SAHS) patients. However there is a big variability among these methods regarding the selection of the source signal and the specific spectral components relevant to the analysis. In this study we investigate the use of the Heart Timing (HT) as the source signal in comparison to the classical approaches of Heart Rate (HR) and Heart Period (HP). This signal has the theoretical advantage of being optimal under the Integral Pulse Frequency Modulation (IPFM) model assumption. Only spectral bands defined as standard for the study of HRV are considered, and for each method the so-called LF/HF and VLFn features are derived. A comparative statistical analysis between the different resulting methods is performed, and subject classification is investigated by means of ROC analysis and a Naïve-Bayes classifier. The standard Apnea-ECG database is used for validation purposes. Our results show statistical differences between SAHS patients and controls for all the derived features. In the subject classification task the best performance in the testing set was obtained using the LF/HF ratio derived from the HR signal (Area under ROC curve=0.88). Only slight differences are obtained due to the effect of changing the source signal. The impact of using the HT signal in this domain is therefore limited, and has not shown relevant differences with respect to the use of the classical approaches of HR or HP.

An Obstructive Sleep Apnea Detection Approach Using a Discriminative Hidden Markov Model From ECG Signals.

Obstructive sleep apnea (OSA) syndrome is a common sleep disorder suffered by an increasing number of people worldwide. As an alternative to polysomnography (PSG) for OSA diagnosis, the automatic OSA detection methods used in the current practice mainly concentrate on feature extraction and classifier selection based on collected physiological signals. However, one common limitation in these methods is that the temporal dependence of signals are usually ignored, which may result in critical information loss for OSA diagnosis. In this study, we propose a novel OSA detection approach based on ECG signals by considering temporal dependence within segmented signals. A discriminative hidden Markov model (HMM) and corresponding parameter estimation algorithms are provided. In addition, subject-specific transition probabilities within the model are employed to characterize the subject-to-subject differences of potential OSA patients. To validate our approach, 70 recordings obtained from the Physionet Apnea-ECG database were used. Accuracies of 97.1% for per-recording classification and 86.2% for per-segment OSA detection with satisfactory sensitivity and specificity were achieved. Compared with other existing methods that simply ignore the temporal dependence of signals, the proposed HMM-based detection approach delivers more satisfactory detection performance and could be extended to other disease diagnosis applications.

Obstructive sleep apnea detection using spectrum and bispectrum analysis of single-lead ECG signal

A novel method for the automatic diagnosis of obstructive sleep apnea (OSA) from an electrocardiogram (ECG) is presented. This method aims to detect OSA utilizing exclusively ECG recordings during sleep and present a minute-by-minute signal processing technique. In the proposed algorithm, a wide range of features based on heart rate variability (HRV) and ECG-derived respiratory (EDR) signals are considered. The novelty of this study arises from employing bispectral analysis to the HRV and EDR signals in order to illustrate quadratic phase-coupling that can be observed among signal components with different frequencies. From this perspective, in the proposed algorithm, a new feature set based on a higher order spectrum of HRV and EDR signals is introduced and it is utilized to extract information regarding their non-linearity and non-Gaussianity. This feature vector is then fed into the input of a least-square support vector machine classifier. To implement the proposed method, the apnea-ECG database, which contains 70 nocturnal ECG records gathered from volunteer men and women, is used in this work. Results obtained from cross-validating 35 data records show that the normal recordings could be separated from the apneic recordings with an accuracy of 95.57% and a sensitivity and specificity of 98.64% and 92.51%, respectively. In addition, 35 other records were used for a pure independent validation of the proposed method and the obtained accuracy, sensitivity and specificity was 94.12%, 93.46% and 94.79% respectively in OSA episode detection. The performance of our proposed technique is better than in other existing approaches. It can be used as a reliable tool for automatic OSA identification and as a result, it will improve medical services.

Sleep apnoea episodes recognition by a committee of ELM classifiers from ECG signal.

This paper describes a system for the recognition of sleep apnoea episodes from ECG signals using a committee of extreme learning machine (ELM) classifiers. RR-interval parameters (heart rate variability) have been used as the identifying features as they are directly affected by sleep apnoea. The MIT PhysioNet Apnea-ECG database was used. A committee of five ELM classifiers has been employed to classify one-minute epochs of ECG into normal or apnoeic epochs. Our results show that the classification performance from the committee of networks was superior to the results of a single ELM classifier for fan-outs from 1 to 100. Classification performance reached a plateau at a fan-out of 10. The maximum accuracy was 82.5% with a sensitivity of 81.9% and a specificity of 82.8%. The results were comparable to other published research with the same input data.

Robust detection of sleep apnea from Holter ECGs. Joint assessment of modulations in QRS amplitude and respiratory myogram interference.

This article is part of the Focus Theme of Methods of Information in Medicine on "Biosignal Interpretation: Advanced Methods for Studying Cardiovascular and Respiratory Systems". Detect presence of sleep-related breathing disorders (SRBD) in epochs of 1 min by signal analysis of Holter ECG recordings. In 121 patients, 140 synchronized polysomnograms (PSGs) and 8-channel Holter ECGs were recorded. The only excluded condition was persistent arrhythmias. Respiratory events as scored from the PSGs were mapped to a 1 min grid and served as reference for ECG-based detection. Moreover, 69/70 recordings of the Physionet Sleep Apnea ECG Database (PADB) were included. We performed receiver operating characteristics analysis of a single, novel time-domain feature, the joint local similarity index (jLSI). Based on cross-correlation, the jLSI quantifies the time-locked occurrence of characteristic low-frequency (LF) modulations in ECG respiratory myogram interference (RMI), QRS amplitude (QRSA) and heart rate. Joint oscillations in QRSA, RMI and the envelope of RMI identified positive epochs with a sensitivity of 0.855 (PADB: 0.873) and a specificity of 0.86 (PADB: 0.88). Inclusion of heart rate did not improve detection accuracy. Joint occurrence of LF-modulations in QRSA and RMI is a characteristic feature of SRBD that is robustly quantified by the jLSI and permits reliable and reproducible detection of sleep apnea in very heterogeneous settings.

Building a Cepstrum-HMM kernel for Apnea identification

Authors present an approach based on the transformation of the Cepstral domain on Hidden Markov Model, which is employed for the automatic diagnosis of the Obstructive Sleep Apnea syndrome. The approach includes an Electrocardiogram artefacts removal and R wave detection stage. In addition, the system is modeled by a transformation of the Cepstral domain sequence using Hidden Markov Models (HMM). Final decisions are taken with two different approaches: A Hidden Markov Model and Support Vector Machine classifiers, where the parameterization is based on the transformation of HMM by a kernel. Two public databases have been used for experiments. Firstly, Physionet Apnea-ECG Database for building algorithms, and finally, The St. Vincent's University Hospital/University College Dublin Sleep Apnea Database for testing out with a blind independent dataset. Achieved results were up to 99.23% for Physionet Apnea-ECG Database, and 98.64% for The St. Vincent's Database.

Sleep Apnea Detection Using Adaptive Neuro Fuzzy Inference System

This paper presents an efficient and easy implemented method for detecting minute based analysis of sleep apnea. The nasal, chest and abdominal based respiratory signals extracted from polysomnography recordings are obtained from PhysioNet apnea-ECG database. Wavelet transforms are applied on the 1-minute and 3-minute length recordings. According to the preliminary tests, the variances of 10th and 11th detail components can be used as discriminative features for apneas. The features obtained from total 8 recordings are used for training and testing of an adaptive neuro fuzzy inference system (ANFIS). Training and testing process have been repeated by using the randomly obtained five different sequences of whole data for generalization of the ANFIS. According to results, ANFIS based classification has sufficient accuracy for apnea detection considering of each type of respiratory. However, the best result is obtained by analyzing the 3-minute length nasal based respiratory signal. In this study, classification accuracies have been obtained greater than 95.2% for each of the five sequences of entire data.

A Novel Method for the Detection of Sleep Apnea Syndrome Based on Single-Lead ECG Signal

The diagnosis of sleep apnea syndrome (SAS) has important clinical significance for the prevention of hypertension, coronary heart disease, arrhythmias, stroke and other diseases. In this paper, a novel method for the detection of SAS based on single-lead Electrocardiogram (ECG) signal was proposed. Firstly, the R-peak points of ECG recordings were pre-detected to calculate RR interval series and ECG-derived respiratory signal (EDR). Then 40 time- and spectral-domain features were extracted and normalized. Finally, support vector machine (SVM) was employed to these features as a classifier to detect SAS events. The performance of the presented method was evaluated using the MIT-BIH Apnea-ECG database, results show that an accuracy of 95% in train sets and an accuracy of 88% in test sets are achievable.

Automatic Apnea Identification by Transformation of the Cepstral Domain

A new approach based on the transformation of the Cepstral domain is developed on this work. This approach reaches an automatic diagnosis for the syndrome of obstructive sleep apnea that includes a specific block for the removal of electrocardiogram (ECG) artifacts and the R wave detection. The system is modeled by a transformation of the Cepstral domain sequence using hidden Markov model (HMM). The final decision is done with two different approaches: one based on HMM as a classifier and a second one based on support vector machines classification and a parameterization based on the transformation of HMM by a kernel. The later approach reached results up to 99.23 %, using all test samples from Physionet Apnea-ECG Database.

ECG signal analysis for the assessment of sleep-disordered breathing and sleep pattern

The diagnosis of sleep-disordered breathing (SDB) usually relies on the analysis of complex polysomnographic measurements performed in specialized sleep centers. Automatic signal analysis is a promising approach to reduce the diagnostic effort. This paper addresses SDB and sleep assessment solely based on the analysis of a single-channel ECG recorded overnight by a set of signal analysis modules. The methodology of QRS detection, SDB analysis, calculation of ECG-derived respiration curves, and estimation of a sleep pattern is described in detail. SDB analysis detects specific cyclical variations of the heart rate by correlation analysis of a signal pattern and the heart rate curve. It was tested with 35 SDB-annotated ECGs from the Apnea-ECG Database, and achieved a diagnostic accuracy of 80.5%. To estimate sleep pattern, spectral parameters of the heart rate are used as stage classifiers. The reliability of the algorithm was tested with 18 ECGs extracted from visually scored polysomnographies of the SIESTA database; 57.7% of all 30s epochs were correctly assigned by the algorithm. Although promising, these results underline the need for further testing in larger patient groups with different underlying diseases.

Screening for Obstructive Sleep Apnea by Cyclic Variation of Heart Rate

Despite the adverse cardiovascular consequences of obstructive sleep apnea, the majority of patients remain undiagnosed. To explore an efficient ECG-based screening tool for obstructive sleep apnea, we examined the usefulness of automated detection of cyclic variation of heart rate (CVHR) in a large-scale controlled clinical setting. We developed an algorithm of autocorrelated wave detection with adaptive threshold (ACAT). The algorithm was optimized with 63 sleep studies in a training cohort, and its performance was confirmed with 70 sleep studies of the Physionet Apnea-ECG database. We then applied the algorithm to ECGs extracted from all-night polysomnograms in 862 consecutive subjects referred for diagnostic sleep study. The number of CVHR per hour (the CVHR index) closely correlated (r=0.84) with the apnea-hypopnea index, although the absolute agreement with the apnea-hypopnea index was modest (the upper and lower limits of agreement, 21 per hour and -19 per hour) with periodic leg movement causing most of the disagreement (P<0.001). The CVHR index showed a good performance in identifying the patients with an apnea-hypopnea index ≥15 per hour (area under the receiver-operating characteristic curve, 0.913; 83% sensitivity and 88% specificity, with the predetermined cutoff threshold of CVHR index ≥15 per hour). The classification performance was unaffected by older age (≥65 years) or cardiac autonomic dysfunction (SD of normal-to-normal R-R intervals over the entire length of recording <65 ms; area under the receiver-operating characteristic curve, 0.915 and 0.911, respectively). The automated detection of CVHR with the ACAT algorithm provides a powerful ECG-based screening tool for moderate-to-severe obstructive sleep apnea, even in older subjects and in those with cardiac autonomic dysfunction.

Real-Time Detection of Apneas on a PDA

Patients suspected of suffering sleep apnea and hypopnea syndrome (SAHS) have to undergo sleep studies such as expensive polysomnographies to be diagnosed. Healthcare professionals are constantly looking for ways to improve the ease of diagnosis and comfort for this kind of patients as well as reducing both the number of sleep studies they need to undergo and the waiting times. Relating to this scenario, some research proposals and commercial products are appearing, but all of them record the physiological data of patients to portable devices and, in the morning, these data are loaded into hospital computers where physicians analyze them by making use of specialized software. In this paper, we present an alternative proposal that promotes not only a transmission of physiological data but also a real-time analysis of these data locally at a mobile device. For that, we have built a classifier that provides an accuracy of 93% and a receiver operating characteristic-area under the curve (ROC-AUC) of 98.5% on SpO(2) signals available in the annotated Apnea-ECG Database. This local analysis allows the detection of anomalous situations as soon as they are generated. The classifier has been implemented taking into consideration the restricted resources of mobile devices.

Detection of Sleep Apnea Episodes from Multi-lead ECGs Considering Different Physiological Influences

This article deals with recognition of sleep apnea, using solely information available from multi-lead ECGs. Characteristic variations in heart rhythm and amplitude of the ECG are compared with respect to their diagnostic accuracy by means of an ROC analysis that is performed on a local similarity index. In 38 8-lead ECGs, each minute is classified with respect to occurrence of apnea events and the result is validated against expert annotations derived from synchronized polysomnographic recordings. Moreover, the results are compared to those obtained from the well known Physionet apnea-ECG database. Whereas the effect of amplitude modulation yields consistent results on both data sets (ROC-area 89.0% vs. 88.3%), a remarkable loss in performance is observed for the frequently applied heart rhythm (89.8% vs. 77.9%). Examples illustrating the reasons for this difference are given and discussed. With respect to aggregation of multi-lead information, two methods (PCA vs. averaging) are compared. The results indicate that averaging performs better (89.3%) than the adaptively estimated PCA (87.2) even when applied to a reduced set of leads. It is concluded that sleep apnea recognition from heart rhythm should always be complemented by analysis of the amplitude variations of the ECG.

Simplified cataract aspirator.

<h3>Abstract</h3> <h3>Background</h3> Sleep apnoea has a high disease burden but remains underdiagnosed, in part due to the expensive and resource intensive nature of polysomnography, its definitive investigation. Emerging literature suggests that it may be possible to detect sleep apnoea using single-lead ECG signals, such as those obtained from smartwatches. In this study, we use two forms of recurrent neural networks (RNNs) to detect sleep apnoea events from single-lead ECG signals. <h3>Methods</h3> We use single-lead ECG data from the PhysioNet Apnea-ECG database, which contains data from 70 patients. We train a bidirectional gated recurrent unit (GRU) model and a bidirectional long short-term memory (LSTM) model on labelled ECG signals from 35 patients and test the models on the remaining 35 patients in the dataset. <h3>Results</h3> Both models achieved 97.1% accuracy, sensitivity and specificity to detect whether the ECG recordings belonged to a patient diagnosed with sleep apnoea. This corresponds to 34/35 patients in the dataset. At detecting individual apnoea events, the GRU and LSTM models achieved 90.4% and 91.7% accuracies respectively. <h3>Discussion</h3> The models achieved high levels of accuracy, specificity and sensitivity. Bidirectional RNNs are strengthened by the ability of the models to be informed by both past and future states when analysing sequential data, such as ECGs. The models also require minimal human intervention as they automatically extract features from the data. If single-lead ECGs prove a suitable tool for sleep apnoea detection, this may enhance the diagnosis of sleep apnoea and potentially allow widespread screening for the condition. <h3>Conclusions</h3> We note that using models such as bidirectional RNNs has the potential to augment model performance. However, more research and validation is required in order to test whether these may be applicable to other datasets and in clinical practice.