An explainable deep-learning model to stage sleep states in children and propose novel EEG-related patterns in sleep apnea

Automatic deep-learning models used for sleep scoring in children with obstructive sleep apnea (OSA) are perceived as black boxes, limiting their implementation in clinical settings. Accordingly, we aimed to develop an accurate and interpretable deep-learning model for sleep staging in children using single-channel electroencephalogram (EEG) recordings. We used EEG signals from the Childhood Adenotonsillectomy Trial (CHAT) dataset (n = 1637) and a clinical sleep database (n = 980). Three distinct deep-learning architectures were explored to automatically classify sleep stages from a single-channel EEG data. Gradient-weighted Class Activation Mapping (Grad-CAM), an explainable artificial intelligence (XAI) algorithm, was then applied to provide an interpretation of the singular EEG patterns contributing to each predicted sleep stage. Among the tested architectures, a standard convolutional neural network (CNN) demonstrated the highest performance for automated sleep stage detection in the CHAT test set (accuracy = 86.9% and five-class kappa = 0.827). Furthermore, the CNN-based estimation of total sleep time exhibited strong agreement in the clinical dataset (intra-class correlation coefficient = 0.772). Our XAI approach using Grad-CAM effectively highlighted the EEG features associated with each sleep stage, emphasizing their influence on the CNN's decision-making process in both datasets. Grad-CAM heatmaps also allowed to identify and analyze epochs within a recording with a highly likelihood to be misclassified, revealing mixed features from different sleep stages within these epochs. Finally, Grad-CAM heatmaps unveiled novel features contributing to sleep scoring using a single EEG channel. Consequently, integrating an explainable CNN-based deep-learning model in the clinical environment could enable automatic sleep staging in pediatric sleep apnea tests.

Derivation and validation of a blood biomarker score for 2-day mortality prediction from prehospital care: a multicenter, cohort, EMS-based study

Identifying potentially life-threatening diseases is a key challenge for emergency medical services. This study aims at examining the role of different prehospital biomarkers from point-of-care testing to derive and validate a score to detect 2-day in-hospital mortality. We conducted a prospective, observational, prehospital, ongoing, and derivation-validation study in three Spanish provinces, in adults evacuated by ambulance and admitted to the emergency department. A total of 23 ambulance-based biomarkers were collected from each patient. A biomarker score based on logistic regression was fitted to predict 2-day mortality from an optimum subset of variables from prehospital blood analysis, obtained through an automated feature selection stage. 2806 cases were analyzed, with a median age of 68 (interquartile range 51-81), 42.3% of women, and a 2-day mortality rate of 5.5% (154 non-survivors). The blood biomarker score was constituted by the partial pressure of carbon dioxide, lactate, and creatinine. The score fitted with logistic regression using these biomarkers reached a high performance to predict 2-day mortality, with an AUC of 0.933 (95% CI 0.841-0.973). The following risk levels for 2-day mortality were identified from the score: low risk (score < 1), where only 8.2% of non-survivors were assigned to; medium risk (1 ≤ score < 4); and high risk (score ≥ 4), where the 2-day mortality rate was 57.6%. The novel blood biomarker score provides an excellent association with 2-day in-hospital mortality, as well as real-time feedback on the metabolic-respiratory patient status. Thus, this score can help in the decision-making process at critical moments in life-threatening situations.

A deep learning model based on the combination of convolutional and recurrent neural networks to enhance pulse oximetry ability to classify sleep stages in children with sleep apnea

Characterization of sleep stages is essential in the diagnosis of sleep-related disorders but relies on manual scoring of overnight polysomnography (PSG) recordings, which is onerous and labor-intensive. Accordingly, we aimed to develop an accurate deep-learning model for sleep staging in children suffering from pediatric obstructive sleep apnea (OSA) using pulse oximetry signals. For this purpose, pulse rate (PR) and blood oxygen saturation (SpO2) from 429 childhood OSA patients were analyzed. A CNN-RNN architecture fed with PR and SpO2 signals was developed to automatically classify wake (W), non-Rapid Eye Movement (NREM), and REM sleep stages. This architecture was composed of: (i) a convolutional neural network (CNN), which learns stage-related features from raw PR and SpO2 data; and (ii) a recurrent neural network (RNN), which models the temporal distribution of the sleep stages. The proposed CNN-RNN model showed a high performance for the automated detection of W/NREM/REM sleep stages (86.0% accuracy and 0.743 Cohen's kappa). Furthermore, the total sleep time estimated for each children using the CNN-RNN model showed high agreement with the manually derived from PSG (intra-class correlation coefficient = 0.747). These results were superior to previous works using CNN-based deep-learning models for automatic sleep staging in pediatric OSA patients from pulse oximetry signals. Therefore, the combination of CNN and RNN allows to obtain additional information from raw PR and SpO2 data related to sleep stages, thus being useful to automatically score sleep stages in pulse oximetry tests for children evaluated for suspected OSA.Clinical Relevance-This research establishes the usefulness of a CNN-RNN architecture to automatically score sleep stages in pulse oximetry tests for pediatric OSA diagnosis.

Conventional Machine Learning Methods Applied to the Automatic Diagnosis of Sleep Apnea

The overnight polysomnography shows a range of drawbacks to diagnose obstructive sleep apnea (OSA) that have led to the search for artificial intelligence-based alternatives. Many classic machine learning methods have been already evaluated for this purpose. In this chapter, we show the main approaches found in the scientific literature along with the most used data to develop the models, useful and large easily available databases, and suitable methods to assess performances. In addition, a range of results from selected studies are presented as examples of these methods. Very high diagnostic performances are reported in these results regardless of the approaches taken. This leads us to conclude that conventional machine learning methods are useful techniques to develop new OSA diagnosis simplification proposals and to act as benchmark for other more recent methods such as deep learning.

Oximetry Indices in the Management of Sleep Apnea: From Overnight Minimum Saturation to the Novel Hypoxemia Measures

Obstructive sleep apnea (OSA) is a multidimensional disease often underdiagnosed due to the complexity and unavailability of its standard diagnostic method: the polysomnography. Among the alternative abbreviated tests searching for a compromise between simplicity and accurateness, oximetry is probably the most popular. The blood oxygen saturation (SpO₂) signal is characterized by a near-constant profile in healthy subjects breathing normally, while marked drops (desaturations) are linked to respiratory events. Parameterization of the desaturations has led to a great number of indices of severity assessment commonly used to assist in OSA diagnosis. In this chapter, the main methodologies used to characterize the overnight oximetry profile are reviewed, from visual inspection and simple statistics to complex measures involving signal processing and pattern recognition techniques. We focus on the individual performance of each approach, but also on the complementarity among the great amount of indices existing in the state of the art, looking for the most relevant oximetric feature subset. Finally, a quick overview of SpO₂-based deep learning applications for OSA management is carried out, where the raw oximetry signal is analyzed without previous parameterization. Our research allows us to conclude that all the methodologies (conventional, time, frequency, nonlinear, and hypoxemia-based) demonstrate high ability to provide relevant oximetric indices, but only a reduced set provide non-redundant complementary information leading to a significant performance increase. Finally, although oximetry is a robust tool, greater standardization and prospective validation of the measures derived from complex signal processing techniques are still needed to homogenize interpretation and increase generalizability.

Deep-Learning Model Based on Convolutional Neural Networks to Classify Apnea–Hypopnea Events from the Oximetry Signal

Automated analysis of the blood oxygen saturation (SpO₂) signal from nocturnal oximetry has shown usefulness to simplify the diagnosis of obstructive sleep apnea (OSA), including the detection of respiratory events. However, the few preceding studies using SpO₂ recordings have focused on the automated detection of respiratory events versus normal respiration, without making any distinction between apneas and hypopneas. In this sense, the characteristics of oxygen desaturations differ between obstructive apnea and hypopnea episodes. In this chapter, we use the SpO₂ signal along with a convolutional neural network (CNN)-based deep-learning architecture for the automatic identification of apnea and hypopnea events. A total of 398 SpO₂ signals from adult OSA patients were used for this purpose. A CNN architecture was trained using 30-s epochs from the SpO₂ signal for the automatic classification of three classes: normal respiration, apnea, and hypopnea. Then, the apnea index (AI), the hypopnea index (HI), and the apnea-hypopnea index (AHI) were obtained by aggregating the outputs of the CNN for each subject (AI_CNN, HI_CNN, and AHI_CNN). This model showed a promising diagnostic performance in an independent test set, with 80.3% 3-class accuracy and 0.539 3-class Cohen's kappa for the classification of respiratory events. Furthermore, AI_CNN, HI_CNN, and AHI_CNN showed a high agreement with the values obtained from the standard PSG: 0.8023, 0.6774, and 0.8466 intra-class correlation coefficients (ICCs), respectively. This suggests that CNN can be used to analyze SpO₂ recordings for the automated diagnosis of OSA in at-home oximetry tests.

Automatic Sleep Staging in Children with Sleep Apnea using Photoplethysmography and Convolutional Neural Networks

Sleep staging is of paramount importance in children with suspicion of pediatric obstructive sleep apnea (OSA). Complexity, cost, and intrusiveness of overnight polysomnography (PSG), the gold standard, have led to the search for alternative tests. In this sense, the photoplethysmography signal (PPG) carries useful information about the autonomous nervous activity associated to sleep stages and can be easily acquired in pediatric sleep apnea home tests with a pulse oximeter. In this study, we use the PPG signal along with convolutional neural networks (CNN), a deep-learning technique, for the automatic identification of the three main levels of sleep: wake (W), rapid eye movement (REM), and non-REM sleep. A database of 366 PPG recordings from pediatric OSA patients is involved in the study. A CNN architecture was trained using 30-s epochs from the PPG signal for three-stage sleep classification. This model showed a promising diagnostic performance in an independent test set, with 78.2% accuracy and 0.57 Cohen's kappa for W/NREM/REM classification. Furthermore, the percentage of time in wake stage obtained for each subject showed no statistically significant differences with the manually scored from PSG. These results were superior to the only state-of-the-art study focused on the analysis of the PPG signal in the automated detection of sleep stages in children suffering from OSA. This suggests that CNN can be used along with PPG recordings for sleep stages scoring in pediatric home sleep apnea tests.

Heart Rate Variability as a Potential Biomarker of Pediatric Obstructive Sleep Apnea Resolution

STUDY OBJECTIVES

Pediatric obstructive sleep apnea (OSA) affects cardiac autonomic regulation, altering heart rate variability (HRV). Although changes in classical HRV parameters occur after OSA treatment, they have not been evaluated as reporters of OSA resolution. Specific frequency bands (named BW1, BW2 and BWRes) have been recently identified in OSA. We hypothesized that changes with treatment in these spectral bands can reliably identify changes in OSA severity and reflect OSA resolution.

METHODS

404 OSA children (5-9.9 years) from the prospective Childhood Adenotonsillectomy Trial (CHAT) were included; 206 underwent early adenotonsillectomy (eAT), while 198 underwent watchful waiting with supportive care (WWSC). HRV changes from baseline to follow-up were computed for classical and OSA-related frequency bands. Causal mediation analysis was conducted to evaluate how treatment influences HRV through mediators such as OSA resolution and changes in disease severity. Disease resolution was initially assessed by considering only obstructive events, and was followed by adding central apneas to the analyses.

RESULTS

Treatment, regardless of eAT or WWSC, affects HRV activity, mainly in the specific frequency band BW2 (0.028-0.074 Hz). Furthermore, only changes in BW2 were specifically attributable to all OSA resolution mediators. HRV activity in BW2 also showed statistically significant differences between resolved and non-resolved OSA.

CONCLUSIONS

OSA treatment affects HRV activity in terms of change in severity and disease resolution, especially in OSA-related BW2 frequency band. This band allowed to differentiate HRV activity between children with and without resolution, so we propose BW2 as potential biomarker of pediatric OSA resolution.

A Convolutional Neural Network Architecture to Enhance Oximetry Ability to Diagnose Pediatric Obstructive Sleep Apnea

This study aims at assessing the usefulness of deep learning to enhance the diagnostic ability of oximetry in the context of automated detection of pediatric obstructive sleep apnea (OSA). A total of 3196 blood oxygen saturation (SpO2) signals from children were used for this purpose. A convolutional neural network (CNN) architecture was trained using 20-min SpO2 segments from the training set (859 subjects) to estimate the number of apneic events. CNN hyperparameters were tuned using Bayesian optimization in the validation set (1402 subjects). This model was applied to three test sets composed of 312, 392, and 231 subjects from three independent databases, in which the apnea-hypopnea index (AHI) estimated for each subject (AHICNN) was obtained by aggregating the output of the CNN for each 20-min SpO2 segment. AHICNN outperformed the 3% oxygen desaturation index (ODI3), a clinical approach, as well as the AHI estimated by a conventional feature-engineering approach based on multi-layer perceptron (AHIMLP). Specifically, AHICNN reached higher four-class Cohen's kappa in the three test databases than ODI3 (0.515 vs 0.417, 0.422 vs 0.372, and 0.423 vs 0.369) and AHIMLP (0.515 vs 0.377, 0.422 vs 0.381, and 0.423 vs 0.306). In addition, our proposal outperformed state-of-the-art studies, particularly for the AHI severity cutoffs of 5 e/h and 10 e/h. This suggests that the information automatically learned from the SpO2 signal by deep-learning techniques helps to enhance the diagnostic ability of oximetry in the context of pediatric OSA.

Heart rate variability during wakefulness as a marker of obstructive sleep apnea severity

STUDY OBJECTIVES

Patients with obstructive sleep apnea (OSA) exhibit heterogeneous heart rate variability (HRV) during wakefulness and sleep. We investigated the influence of OSA severity on HRV parameters during wakefulness in a large international clinical sample.

METHODS

1247 subjects (426 without OSA and 821 patients with OSA) were enrolled from the Sleep Apnea Global Interdisciplinary Consortium. HRV parameters were calculated during a 5-minute wakefulness period with spontaneous breathing prior to the sleep study, using time-domain, frequency-domain and nonlinear methods. Differences in HRV were evaluated among groups using analysis of covariance, controlling for relevant covariates.

RESULTS

Patients with OSA showed significantly lower time-domain variations and less complexity of heartbeats compared to individuals without OSA. Those with severe OSA had remarkably reduced HRV compared to all other groups. Compared to non-OSA patients, those with severe OSA had lower HRV based on SDNN (adjusted mean: 37.4 vs. 46.2 ms; p < 0.0001), RMSSD (21.5 vs. 27.9 ms; p < 0.0001), ShanEn (1.83 vs. 2.01; p < 0.0001), and Forbword (36.7 vs. 33.0; p = 0.0001). While no differences were found in frequency-domain measures overall, among obese patients there was a shift to sympathetic dominance in severe OSA, with a higher LF/HF ratio compared to obese non-OSA patients (4.2 vs. 2.7; p = 0.009).

CONCLUSIONS

Time-domain and nonlinear HRV measures during wakefulness are associated with OSA severity, with severe patients having remarkably reduced and less complex HRV. Frequency-domain measures show a shift to sympathetic dominance only in obese OSA patients. Thus, HRV during wakefulness could provide additional information about cardiovascular physiology in OSA patients.

CLINICAL TRIAL INFORMATION

A Prospective Observational Cohort to Study the Genetics of Obstructive Sleep Apnea and Associated Co-Morbidities (German Clinical Trials Register - DKRS, DRKS00003966) https://www.drks.de/drks_web/navigate.do?navigationId=trial.HTML&TRIAL_ID=DRKS00003966.

Wavelet Analysis of Overnight Airflow to Detect Obstructive Sleep Apnea in Children

This study focused on the automatic analysis of the airflow signal (AF) to aid in the diagnosis of pediatric obstructive sleep apnea (OSA). Thus, our aims were: (i) to characterize the overnight AF characteristics using discrete wavelet transform (DWT) approach, (ii) to evaluate its diagnostic utility, and (iii) to assess its complementarity with the 3% oxygen desaturation index (ODI3). In order to reach these goals, we analyzed 946 overnight pediatric AF recordings in three stages: (i) DWT-derived feature extraction, (ii) feature selection, and (iii) pattern recognition. AF recordings from OSA patients showed both lower detail coefficients and decreased activity associated with the normal breathing band. Wavelet analysis also revealed that OSA disturbed the frequency and energy distribution of the AF signal, increasing its irregularity. Moreover, the information obtained from the wavelet analysis was complementary to ODI3. In this regard, the combination of both wavelet information and ODI3 achieved high diagnostic accuracy using the common OSA-positive cutoffs: 77.97%, 81.91%, and 90.99% (AdaBoost.M2), and 81.96%, 82.14%, and 90.69% (Bayesian multi-layer perceptron) for 1, 5, and 10 apneic events/hour, respectively. Hence, these findings suggest that DWT properly characterizes OSA-related severity as embedded in nocturnal AF, and could simplify the diagnosis of pediatric OSA.

EEG-Inception: A Novel Deep Convolutional Neural Network for Assistive ERP-Based Brain-Computer Interfaces

In recent years, deep-learning models gained attention for electroencephalography (EEG) classification tasks due to their excellent performance and ability to extract complex features from raw data. In particular, convolutional neural networks (CNN) showed adequate results in brain-computer interfaces (BCI) based on different control signals, including event-related potentials (ERP). In this study, we propose a novel CNN, called EEG-Inception, that improves the accuracy and calibration time of assistive ERP-based BCIs. To the best of our knowledge, EEG-Inception is the first model to integrate Inception modules for ERP detection, which combined efficiently with other structures in a light architecture, improved the performance of our approach. The model was validated in a population of 73 subjects, of which 31 present motor disabilities. Results show that EEG-Inception outperforms 5 previous approaches, yielding significant improvements for command decoding accuracy up to 16.0%, 10.7%, 7.2%, 5.7% and 5.1% in comparison to rLDA, xDAWN + Riemannian geometry, CNN-BLSTM, DeepConvNet and EEGNet, respectively. Moreover, EEG-Inception requires very few calibration trials to achieve state-of-the-art performances taking advantage of a novel training strategy that combines cross-subject transfer learning and fine-tuning to increase the feasibility of this approach for practical use in assistive applications.

The Different Facets of Heart Rate Variability in Obstructive Sleep Apnea

Obstructive sleep apnea (OSA), a heterogeneous and multifactorial sleep related breathing disorder with high prevalence, is a recognized risk factor for cardiovascular morbidity and mortality. Autonomic dysfunction leads to adverse cardiovascular outcomes in diverse pathways. Heart rate is a complex physiological process involving neurovisceral networks and relative regulatory mechanisms such as thermoregulation, renin-angiotensin-aldosterone mechanisms, and metabolic mechanisms. Heart rate variability (HRV) is considered as a reliable and non-invasive measure of autonomic modulation response and adaptation to endogenous and exogenous stimuli. HRV measures may add a new dimension to help understand the interplay between cardiac and nervous system involvement in OSA. The aim of this review is to introduce the various applications of HRV in different aspects of OSA to examine the impaired neuro-cardiac modulation. More specifically, the topics covered include: HRV time windows, sleep staging, arousal, sleepiness, hypoxia, mental illness, and mortality and morbidity. All of these aspects show pathways in the clinical implementation of HRV to screen, diagnose, classify, and predict patients as a reasonable and more convenient alternative to current measures.

Assessment of Nocturnal Autonomic Cardiac Imbalance in Positional Obstructive Sleep Apnea. A Multiscale Nonlinear Approach

Positional obstructive sleep apnea (POSA) is a major phenotype of sleep apnea. Supine-predominant positional patients are frequently characterized by milder symptoms and less comorbidity due to a lower age, body mass index, and overall apnea-hypopnea index. However, the bradycardia-tachycardia pattern during apneic events is known to be more severe in the supine position, which could affect the cardiac regulation of positional patients. This study aims at characterizing nocturnal heart rate modulation in the presence of POSA in order to assess potential differences between positional and non-positional patients. Patients showing clinical symptoms of suffering from a sleep-related breathing disorder performed unsupervised portable polysomnography (PSG) and simultaneous nocturnal pulse oximetry (NPO) at home. Positional patients were identified according to the Amsterdam POSA classification (APOC) criteria. Pulse rate variability (PRV) recordings from the NPO readings were used to assess overnight cardiac modulation. Conventional cardiac indexes in the time and frequency domains were computed. Additionally, multiscale entropy (MSE) was used to investigate the nonlinear dynamics of the PRV recordings in POSA and non-POSA patients. A total of 129 patients (median age 56.0, interquartile range (IQR) 44.8-63.0 years, median body mass index (BMI) 27.7, IQR 26.0-31.3 kg/m2) were classified as POSA (37 APOC I, 77 APOC II, and 15 APOC III), while 104 subjects (median age 57.5, IQR 49.0-67.0 years, median BMI 29.8, IQR 26.6-34.7 kg/m2) comprised the non-POSA group. Overnight PRV recordings from positional patients showed significantly higher disorderliness than non-positional subjects in the smallest biological scales of the MSE profile (τ = 1: 0.25, IQR 0.20-0.31 vs. 0.22, IQR 0.18-0.27, p < 0.01) (τ = 2: 0.41, IQR 0.34-0.48 vs. 0.37, IQR 0.29-0.42, p < 0.01). According to our findings, nocturnal heart rate regulation is severely affected in POSA patients, suggesting increased cardiac imbalance due to predominant positional apneas.

Bispectral analysis of overnight airflow to improve the pediatric sleep apnea diagnosis

Pediatric Obstructive Sleep Apnea (OSA) is a respiratory disease whose diagnosis is performed through overnight polysomnography (PSG). Since it is a complex, time-consuming, expensive, and labor-intensive test, simpler alternatives are being intensively sought. In this study, bispectral analysis of overnight airflow (AF) signal is proposed as a potential approach to replace PSG when indicated. Thus, our objective was to characterize AF through bispectrum, and assess its performance to diagnose pediatric OSA. This characterization was conducted using 13 bispectral features from 946 AF signals. The oxygen desaturation index ≥3% (ODI3), a common clinical measure of OSA severity, was also obtained to evaluate its complementarity to the AF bispectral analysis. The fast correlation-based filter (FCBF) and a multi-layer perceptron (MLP) were used for subsequent automatic feature selection and pattern recognition stages. FCBF selected 3 bispectral features and ODI3, which were used to train a MLP model with ability to estimate apnea-hypopnea index (AHI). The model reached 82.16%, 82.49%, and 90.15% accuracies for the common AHI cut-offs 1, 5, and 10 events/h, respectively. The different bispectral approaches used to characterize AF in children provided complementary information. Accordingly, bispectral analysis showed that the occurrence of apneic events decreases the non-gaussianity and non-linear interaction of the AF harmonic components, as well as the regularity of the respiratory patterns. Moreover, the bispectral information from AF also showed complementarity with ODI3. Our findings suggest that AF bispectral analysis may serve as a useful tool to simplify the diagnosis of pediatric OSA, particularly for children with moderate-to-severe OSA.

Automatic Assessment of Pediatric Sleep Apnea Severity Using Overnight Oximetry and Convolutional Neural Networks

In this study, we use the overnight blood oxygen saturation (SpO₂) signal along with convolutional neural networks (CNN) for the automatic estimation of pediatric sleep apnea-hypopnea syndrome (SAHS) severity. The few preceding studies have focused on the application of conventional feature extraction methods to obtain information from the SpO₂ signal, which may omit relevant data related to the illness. In contrast, deep learning techniques are able to automatically learn features from raw input signal. Thus, we propose to assess whether CNN, a deep learning algorithm, could automatically estimate the apnea-hypopnea index (AHÍ) from nocturnal oximetry to help establish pediatric SAHS presence and severity. A database of 746 SpO₂ recordings is involved in the study. CNN was trained using 20-min segments from the SpO₂ signal in the training set (400 subjects). Hyperparameters of the CNN architecture were tuned using a validation set (100 subjects). This model was applied to a test set (246 subjects), in which the final AHI of each patient was obtained as the average of the output of the CNN for all the segments of the corresponding SpO₂ signal. The AHI estimated by the CNN showed a promising diagnostic performance, with 74.8%, 90.7%, and 95.1% accuracies for the common AHI severity thresholds of 1, 5, and 10 events per hour (e/h), respectively. Furthermore, this model reached 28.6, 32.9, and 120.0 positive likelihood ratios for the above-mentioned AHI thresholds. This suggests that the information extracted from the oximetry signal by deep learning techniques may be useful to both establish pediatric SAHS and its severity.

Intraindividual Characterization of the Sleep Spindle Variability in Healthy Subjects

Spatial and frequency characterization of sleep spindles have been extensively addressed using M/EEG or fMRI recordings. However, its intraindividual variability across time has not been addressed. Here we propose to assess the intraindividual variability of sleep spindles in a time-resolved way by means of a trial-to-trial-variability (TTV) measure. For that purpose, the EEG of 26 healthy subjects were recorded overnight. After an exhaustive preprocessing pipeline to remove artifacts, spindles were automatically detected using a complex demodulation-based method. Then, the Wavelet Scalogram was estimated to validate it. Spindle TTV of each participant was also computed for all the conventional EEG frequency bands. Root mean square (RMS) of each TTV signal was calculated as a measure of the total variability of each spindle. Results showed significant differences in the variability between frequencies. Specifically, RMS in the beta-1 frequency band showed higher values as compared to all the other frequency bands (p<0.001). TTV curves showed a dichotomic trend, with lower frequencies showing an increase in the variability before the spindle onset, and higher frequencies showing such increase after the onset. The dependence of the spindle variability with the frequency could be explained by the influence of the multiple cortical generators involved.Clinical Relevance- Sleep spindles are similarly affected in different cognitive-related disorders, which supports the relevance of assessing abnormal sleep patterns as a possible cause for such cognitive deficits.

Convolutional Neural Networks to Detect Pediatric Apnea-Hypopnea Events from Oximetry

Pediatric sleep apnea-hypopnea syndrome (SAHS) is a highly prevalent breathing disorder that is related to many negative consequences for the children's health and quality of life when it remains untreated. The gold standard for pediatric SAHS diagnosis (overnight polysomnography) has several limitations, which has led to the search for alternative tests. In this sense, automated analysis of overnight oximetry has emerged as a simplified technique. Previous studies have focused on the extraction of ad-hoc features from the blood oxygen saturation (SpO₂) signal, which may miss useful information related to apnea and hypopnea (AH) events. In order to overcome this limitation of traditional approaches, we propose the use of convolutional neural networks (CNN), a deep learning technique, to automatically detect AH events from the SpO₂ raw data. CHAT-baseline dataset, composed of 453 SpO₂ recordings, was used for this purpose. A CNN model was trained using 60-s segments from the SpO₂ signal using a training set (50% of subjects). Optimum hyperparameters of the CNN architecture were obtained using a validation set (25% of subjects). This model was applied to a third test set (25% of subjects), reaching 93.6% accuracy to detect AH events. These results suggest that the application of CNN may be useful to detect changes produced in the oximetry signal by AH events in pediatric SAHS patients.

A machine learning-based test for adult sleep apnoea screening at home using oximetry and airflow

The most appropriate physiological signals to develop simplified as well as accurate screening tests for obstructive sleep apnoea (OSA) remain unknown. This study aimed at assessing whether joint analysis of at-home oximetry and airflow recordings by means of machine-learning algorithms leads to a significant diagnostic performance increase compared to single-channel approaches. Consecutive patients showing moderate-to-high clinical suspicion of OSA were involved. The apnoea-hypopnoea index (AHI) from unsupervised polysomnography was the gold standard. Oximetry and airflow from at-home polysomnography were parameterised by means of 38 time, frequency, and non-linear variables. Complementarity between both signals was exhaustively inspected via automated feature selection. Regression support vector machines were used to estimate the AHI from single-channel and dual-channel approaches. A total of 239 patients successfully completed at-home polysomnography. The optimum joint model reached 0.93 (95%CI 0.90–0.95) intra-class correlation coefficient between estimated and actual AHI. Overall performance of the dual-channel approach (kappa: 0.71; 4-class accuracy: 81.3%) significantly outperformed individual oximetry (kappa: 0.61; 4-class accuracy: 75.0%) and airflow (kappa: 0.42; 4-class accuracy: 61.5%). According to our findings, oximetry alone was able to reach notably high accuracy, particularly to confirm severe cases of the disease. Nevertheless, oximetry and airflow showed high complementarity leading to a remarkable performance increase compared to single-channel approaches. Consequently, their joint analysis via machine learning enables accurate abbreviated screening of OSA at home.

Usefulness of Spectral Analysis of Respiratory Rate Variability to Help in Pediatric Sleep Apnea-Hypopnea Syndrome Diagnosis

The sleep apnea-hypopnea syndrome (SAHS) is a chronic respiratory disorder of high prevalence among children (up to 4%). Nocturnal polysomnography (PSG) is the gold standard method to diagnose SAHS, which is a complex, expensive, and time-consuming test. Consequently, alternative simplified methods are demanded. We propose the analysis of the respiratory rate variability (RRV) signal, directly obtained from the airflow (AF) signals. The aim of our study is to evaluate the usefulness of the spectral information obtained from RRV in the diagnosis of pediatric SAHS. A database composed of 946 AF and blood oxygen saturation (SpO2) recordings from children between 0 and 13 years old was used. Our database was divided into four severity groups according to the apnea-hipopnea index (AHI): no-SAHS (AHI <; 1 events/h), mild (1 events/h ≤ AHI <; 5 events/h), moderate (5 events/h ≤ AHI <; 10 events/h), and severe SAHS (AHI ≥ 10 events/h). RRV and 3% oxygen desaturation index (ODI3) were obtained from AF and SpO2 recordings, respectively. A spectral band of interest was determined (0.09-0.20 Hz.) and a total of 12 spectral features were extracted. Nine of these features showed statistically significant differences (p-value <; 0.05) among the four severity groups. The spectral features from RRV along with ODI3 were used as inputs to binary logistic regression (LR) classifiers. The diagnostic performance of LR models were evaluated for the AHI cut-off points of 1, 5, and 10 e/h, achieving 66.5%, 84.0%, and 88.5% accuracy, respectively. These results outperformed those obtained by single ODI3. The joint use of the spectral information from RRV and ODI3 achieved a high diagnostic capability in the most severely-affected children, thus showing their complementarity. These results suggest that the information contained in RRV spectrum together with ODI3 is useful to help identify moderate-to-severe SAHS.

Usefulness of recurrence plots from airflow recordings to aid in paediatric sleep apnoea diagnosis

BACKGROUND AND OBJECTIVE

In-laboratory overnight polysomnography (PSG) is the gold standard method to diagnose the Sleep Apnoea-Hypopnoea Syndrome (SAHS). PSG is a complex, expensive, labour-intensive and time-consuming test. Consequently, simplified diagnostic methods are desirable. We propose the analysis of the airflow (AF) signal by means of recurrence plots (RP) features. The main goal of our study was to evaluate the utility of the information from RPs of the AF signals to detect paediatric SAHS at different levels of severity. In addition, we also evaluated the complementarity with the 3% oxygen desaturation index (ODI₃).

METHODS

946 AF and blood oxygen saturation (SpO₂) recordings from children ages 0-13 years were used. The population under study was randomly split into training (60%) and test (40%) sets. RP was computed and 9 RP features were extracted from each AF recording. ODI₃ was also calculated from each SpO₂ recording. A feature selection stage was conducted in the training group by means of the fast correlation-based filter (FCBF) methodology to obtain a relevant and non-redundant optimum feature subset. A multi-layer perceptron neural network with Bayesian approach (BY-MLP), trained with these optimum features, was used to estimate the apnoea-hypopnoea index (AHI).

RESULTS

8 of the RP features showed statistically significant differences (p-value <0.01) among the SAHS severity groups. FCBF selected the maximum length of the diagonal lines from RP, as well as the ODI₃. Using these optimum features, the BY-MLP model achieved 83.2%, 78.5%, and 91.0% accuracy in the test group for the AHI thresholds 1, 5, and 10 events/h, respectively. Moreover, this model reached a negative likelihood ratio of 0.1 for 1 event/h and a positive likelihood ratio of 13.7 for 10 events/h.

CONCLUSIONS

RP analysis enables extraction of useful SAHS-related information from overnight AF paediatric recordings. Moreover, it provides complementary information to the widely-used clinical variable ODI₃. Thus, RP applied to AF signals can be used along with ODI₃ to help in paediatric SAHS diagnosis, particularly to either confirm the absence of SAHS or the presence of severe SAHS.

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

Chronic obstructive pulmonary disease (COPD) is one of the most prevalent lung diseases worldwide. COPD patients show major dysfunction in cardiac autonomic modulation due to sustained hypoxaemia, which has been significantly related to higher risk of cardiovascular disease. Obstructive sleep apnoea syndrome (OSAS) is a frequent comorbidity in COPD patients. It has been found that patients suffering from both COPD and OSAS simultaneously, the so-called overlap syndrome, have notably higher morbidity and mortality. Heart rate variability (HRV) has demonstrated to be useful to assess changes in autonomic functioning in different clinical conditions. However, there is still little scientific evidence on the magnitude of changes in cardiovascular dynamics elicited by the combined effect of both respiratory diseases, particularly during sleep, when apnoeic events occur. In this regard, we hypothesised that a non-linear analysis is able to provide further insight into long-term dynamics of overnight cardiovascular modulation. Accordingly, this study is aimed at assessing the usefulness of sample entropy (SampEn) to distinguish changes in overnight pulse rate variability (PRV) recordings among three patient groups while sleeping: COPD, moderate-to-severe OSAS, and overlap syndrome. In order to achieve this goal, a population composed of 297 patients were studied: 22 with COPD alone, 213 showing moderate-to-severe OSAS, and 62 with COPD and moderate-to-severe OSAS simultaneously (COPD+OSAS). Cardiovascular dynamics were analysed using pulse rate (PR) recordings from unattended pulse oximetry carried out at patients’ home. Conventional time- and frequency- domain analyses were performed to characterise sympathetic and parasympathetic activation of the nervous system, while SampEn was applied to quantify long-term changes in irregularity. Our analyses revealed that overnight PRV recordings from COPD+OSAS patients were significantly more irregular (higher SampEn) than those from patients with COPD alone (0.267 [0.210–0.407] vs. 0.212 [0.151–0.267]; p < 0.05) due to recurrent apnoeic events during the night. Similarly, COPD + OSAS patients also showed significantly higher irregularity in PRV during the night than subjects with OSAS alone (0.267 [0.210–0.407] vs. 0.241 [0.189–0.325]; p = 0.05), which suggests that the cumulative effect of both diseases increases disorganization of pulse rate while sleeping. On the other hand, no statistical significant differences were found between COPD and COPD + OSAS patients when traditional frequency bands (LF and HF) were analysed. We conclude that SampEn is able to properly quantify changes in overnight cardiovascular dynamics of patients with overlap syndrome, which could be useful to assess cardiovascular impairment in COPD patients due to the presence of concomitant OSAS.

Wavelet analysis of oximetry recordings to assist in the automated detection of moderate-to-severe pediatric sleep apnea-hypopnea syndrome

BACKGROUND

The gold standard for pediatric sleep apnea hypopnea syndrome (SAHS) is overnight polysomnography, which has several limitations. Thus, simplified diagnosis techniques become necessary.

OBJECTIVE

The aim of this study is twofold: (i) to analyze the blood oxygen saturation (SpO2) signal from nocturnal oximetry by means of features from the wavelet transform in order to characterize pediatric SAHS; (ii) to evaluate the usefulness of the extracted features to assist in the detection of pediatric SAHS.

METHODS

981 SpO2 signals from children ranging 2-13 years of age were used. Discrete wavelet transform (DWT) was employed due to its suitability to deal with non-stationary signals as well as the ability to analyze the SAHS-related low frequency components of the SpO2 signal with high resolution. In addition, 3% oxygen desaturation index (ODI3), statistical moments and power spectral density (PSD) features were computed. Fast correlation-based filter was applied to select a feature subset. This subset fed three classifiers (logistic regression, support vector machines (SVM), and multilayer perceptron) trained to determine the presence of moderate-to-severe pediatric SAHS (apnea-hypopnea index cutoff ≥ 5 events per hour).

RESULTS

The wavelet entropy and features computed in the D9 detail level of the DWT reached significant differences associated with the presence of SAHS. All the proposed classifiers fed with a selected feature subset composed of ODI3, statistical moments, PSD, and DWT features outperformed every single feature. SVM reached the highest performance. It achieved 84.0% accuracy (71.9% sensitivity, 91.1% specificity), outperforming state-of-the-art studies in the detection of moderate-to-severe SAHS using the SpO2 signal alone.

CONCLUSION

Wavelet analysis could be a reliable tool to analyze the oximetry signal in order to assist in the automated detection of moderate-to-severe pediatric SAHS. Hence, pediatric subjects suffering from moderate-to-severe SAHS could benefit from an accurate simplified screening test only using the SpO2 signal.

Improving the Diagnostic Ability of Oximetry Recordings in Pediatric Sleep Apnea-Hypopnea Syndrome by Means of Multi-Class AdaBoost

Pediatric sleep apnea-hypopnea syndrome (SAHS) is a highly prevalent respiratory disorder that may impose many negative effects on the health and development of children. Due to the drawbacks of overnight polysomnography (PSG), the gold standard diagnosis technique, automated analysis of nocturnal oximetry has emerged as a simplified alternative. In order to improve diagnosis ability of oximetry, we propose to evaluate the usefulness of AdaBoost, a classification boosting algorithm, in the context of pediatric SAHS. A database composed of 981 SpO₂ recordings from pediatric subjects was used. For this purpose, a signal processing approach divided into two main stages was conducted: (i) feature extraction, where 3% oxygen desaturation index (ODI3), spectral, and nonlinear features were computed from the oximetry signal, and (ii) AdaBoost classification, where an AdaBoost.M2 model was trained with these features in order to determine the severity of pediatric SAHS according to the apnea-hypopnea index (AHI): AHI<1 events per hour (e/h), 1≤AHI<5 e/h, and AHI≥5 e/h. Our AdaBoost.M2 model achieved a Cohen's kappa of 0.474 in an independent test set in the 3-class classification task. In addition, high accuracies were obtained when using the AHI cutoffs for diagnosis of mild (AHI=1 e/h) and moderate-to-severe (AHI=5 e/h) SAHS: 80.9% and 82.9%, respectively. These results achieved slightly higher diagnostic accuracies than ODI3 as well as state-of-the-art studies. Therefore, AdaBoost could help to enhance the diagnostic ability of the oximetry signal to assess pediatric SAHS severity.

Bispectral Analysis to Enhance Oximetry as a Simplified Alternative for Pediatric Sleep Apnea Diagnosis

This study aims at assessing the bispectral analysis of blood oxygen saturation (SpO²) from nocturnal oximetry to help in pediatric sleep apnea-hypopnea syndrome (SAHS) diagnosis. Recent studies have found excessive redundancy in the SAHS-related information usually extracted from SpO², while proposing only two features as a reduced set to be used. On the other hand, it has been suggested that SpO₂ bispectral analysis is able to provide complementary information to common anthropometric, spectral, and clinical variables. We address these novel findings to assess whether bispectrum provides new non-redundant information to help in SAHS diagnosis. Thus, we use 981 pediatric SpO₂ recordings to extract both the reduced set of features recently proposed as well as 9 bispectral features. Then, a feature selection method based on the fast correlationbased filter and bootstrapping is used to assess redundancy among all the features. Finally, the non-redundant ones are used to train a Bayesian multi-layer perceptron neural network (BYMLP) that estimate the apnea-hypopnea index (AHI), which is the diagnostic reference variable. Bispectral phase entropy was found complementary to the two previously recommended features and a BY-MLP model trained with the three of them reached high agreement with actual AHI (intra-class correlation coefficient = 0.889). Estimated AHI also showed high diagnostic ability, reaching 82.1%, 81.9%, and 90.3% accuracies and 0.814, 0.880, and 0.922 area under the receiver-operating characteristics curve for three common AHI thresholds: 1 e/h, 5 e/h, and 10 e/h, respectively. These results suggest that the information extracted from the bispectrum of SpO₂ can improve the diagnostic performance of the oximetry test.

Detrended fluctuation analysis of the oximetry signal to assist in paediatric sleep apnoea-hypopnoea syndrome diagnosis

OBJECTIVE

To evaluate whether detrended fluctuation analysis (DFA) provides information that improves the diagnostic ability of the oximetry signal in the diagnosis of paediatric sleep apnoea-hypopnoea syndrome (SAHS).

APPROACH

A database composed of 981 blood oxygen saturation (SpO₂) recordings in children was used to extract DFA-derived features in order to quantify the scaling behaviour and the fluctuations of the SpO₂ signal. The 3% oxygen desaturation index (ODI3) was also computed for each subject. Fast correlation-based filter (FCBF) was then applied to select an optimum subset of relevant and non-redundant features. This subset fed a multi-layer perceptron (MLP) neural network to estimate the apnoea-hypopnoea index (AHI).

MAIN RESULTS

ODI3 and four features from the DFA reached significant differences associated with the severity of SAHS. An optimum subset composed of the slope in the first scaling region of the DFA profile and the ODI3 was selected using FCBF applied to the training set (60% of samples). The MLP model trained with this feature subset showed good agreement with the actual AHI, reaching an intra-class correlation coefficient of 0.891 in the test set (40% of samples). Furthermore, the estimated AHI showed high diagnostic ability, reaching an accuracy of 82.7%, 81.9%, and 91.1% using three common AHI cut-offs of 1, 5, and 10 events per hour (e h^-1), respectively. These results outperformed the overall performance of ODI3.

SIGNIFICANCE

DFA may serve as a reliable tool to improve the diagnostic performance of oximetry recordings in the evaluation of paediatric patients with symptoms suggestive of SAHS.

Bispectral analysis of spontaneous EEG activity from patients with moderate dementia due to Alzheimer's disease

Dementia due to Alzheimer's disease (AD) is a common disorder with a great impact on the patients' quality of life. The aim of this pilot study was to characterize spontaneous electroencephalography (EEG) activity in dementia due to AD using bispectral analysis. Five minutes of EEG activity were recorded from 17 patients with moderate dementia due to AD and 19 age-matched controls. Bispectrum results revealed that AD patients are characterized by an increase of phase coupling at low frequencies in comparison with controls. Additionally, some bispectral features calculated from the bispectrum showed significant differences between both groups (p <; 0.05, Mann-Whitney U test with Bonferroni's correction). Finally, a stepwise logistic regression analysis with a leave-one-out cross-validation procedure was used for classification purposes. An accuracy of 86.11% (sensitivity = 88.24%; specificity =84.21%) was achieved. This study suggests the usefulness of bispectral analysis to provide further insights into the underlying brain dynamics associated with AD.

Utility of bispectrum in the screening of pediatric sleep apnea-hypopnea syndrome using oximetry recordings

BACKGROUND AND OBJECTIVE

The aim of this study was to assess the utility of bispectrum-based oximetry approaches as a complementary tool to traditional techniques in the screening of pediatric sleep apnea-hypopnea syndrome (SAHS).

METHODS

298 blood oxygen saturation (SpO₂) signals from children ranging 0-13 years of age were recorded during overnight polysomnography (PSG). These recordings were divided into three severity groups according to the PSG-derived apnea hypopnea index (AHI): AHI < 5 events per hour (e/h), 5 ≤ AHI < 10 e/h, AHI ≥ 10 e/h. For each pediatric subject, anthropometric variables, 3% oxygen desaturation index (ODI3) and spectral features from power spectral density (PSD) and bispectrum were obtained. Then, the fast correlation-based filter (FCBF) was applied to select a subset of relevant features that may be complementary, excluding those that are redundant. The selected features fed a multiclass multi-layer perceptron (MLP) neural network to build a model to estimate the SAHS severity degrees.

RESULTS

An optimum subset with features from all the proposed methodological approaches was obtained: variables from bispectrum, as well as PSD, ODI3, Age, and Sex. In the 3-class classification task, the MLP model trained with these features achieved an accuracy of 76.0% and a Cohen's kappa of 0.56 in an independent test set. Additionally, high accuracies were reached using the AHI cutoffs for diagnosis of moderate (AHI = 5 e/h) and severe (AHI = 10 e/h) SAHS: 81.3% and 85.3%, respectively. These results outperformed the diagnostic ability of a MLP model built without using bispectral features.

CONCLUSIONS

Our results suggest that bispectrum provides additional information to anthropometric variables, ODI3 and PSD regarding characterization of changes in the SpO₂ signal caused by respiratory events. Thus, oximetry bispectrum can be a useful tool to provide complementary information for screening of moderate-to-severe pediatric SAHS.

Nocturnal Oximetry-based Evaluation of Habitually Snoring Children

RATIONALE

The vast majority of children around the world undergoing adenotonsillectomy for obstructive sleep apnea-hypopnea syndrome (OSA) are not objectively diagnosed by nocturnal polysomnography because of access availability and cost issues. Automated analysis of nocturnal oximetry (nSp_O2), which is readily and globally available, could potentially provide a reliable and convenient diagnostic approach for pediatric OSA.

METHODS

Deidentified nSp_O2 recordings from a total of 4,191 children originating from 13 pediatric sleep laboratories around the world were prospectively evaluated after developing and validating an automated neural network algorithm using an initial set of single-channel nSp_O2 recordings from 589 patients referred for suspected OSA.

MEASUREMENTS AND MAIN RESULTS

The automatically estimated apnea-hypopnea index (AHI) showed high agreement with AHI from conventional polysomnography (intraclass correlation coefficient, 0.785) when tested in 3,602 additional subjects. Further assessment on the widely used AHI cutoff points of 1, 5, and 10 events/h revealed an incremental diagnostic ability (75.2, 81.7, and 90.2% accuracy; 0.788, 0.854, and 0.913 area under the receiver operating characteristic curve, respectively).

CONCLUSIONS

Neural network-based automated analyses of nSp_O2 recordings provide accurate identification of OSA severity among habitually snoring children with a high pretest probability of OSA. Thus, nocturnal oximetry may enable a simple and effective diagnostic alternative to nocturnal polysomnography, leading to more timely interventions and potentially improved outcomes.

Usefulness of discrete wavelet transform in the analysis of oximetry signals to assist in childhood sleep apnea-hypopnea syndrome diagnosis

Sleep apnea hypopnea syndrome (SAHS) is a highly prevalent respiratory disorder that may cause many negative consequences for the health and development of children. The gold standard for diagnosis is the overnight polysomnography (PSG), which is a high cost, complex, intrusive, and time-demanding technique. To improve the early detection of pediatric SAHS, we propose an automated analysis of the SpO₂ signal from nocturnal oximetry. A database composed of 298 SpO₂ recordings from children ranging from 0 to 13 years old was used for this purpose. Due to the abrupt changes caused by respiratory events in the SpO₂ signal, our goal was to evaluate the diagnostic ability of this by means of the discrete wavelet transform (DWT). To achieve this objective, we conducted a signal processing approach divided into two main stages: (i) feature extraction, where features from the DWT detail coefficients were computed, and (ii) feature classification, where a logistic regression (LR) model was used to classify children into SAHS negative or SAHS positive. Our results showed that respiratory events introduced more variability in two detail levels of the DWT from SpO₂: 0.024-0.049 Hz and 0.012-0.024 Hz. Moreover, the LR classifier achieved an 81.9% accuracy (79.1% sensitivity and 84.1% specificity) in an independent test set for a clinical cutoff point of 5 events/h, as derived from PSG. These results suggest that DWT analysis may be a useful tool to analyze SpO₂ recordings in the context of childhood SAHS.

A Bayesian neural network approach to compare the spectral information from nasal pressure and thermistor airflow in the automatic sleep apnea severity estimation

In the sleep apnea-hypopnea syndrome (SAHS) context, airflow signal plays a key role for the simplification of the diagnostic process. It is measured during the standard diagnostic test by the acquisition of two simultaneous sensors: a nasal prong pressure (NPP) and a thermistor (TH). The current study focuses on the comparison of their spectral content to help in the automatic SAHS-severity estimation. The spectral analysis of 315 NPP and corresponding TH recordings is firstly proposed to characterize the conventional band of interest for SAHS (0.025-0.050 Hz.). A magnitude squared coherence analysis is also conducted to quantify possible differences in the frequency components of airflow from both sensors. Then, a feature selection stage is implemented to assess the relevance and redundancy of the information extracted from the spectrum of NPP and TH airflow. Finally, a multiclass Bayesian multi-layer perceptron (BY-MLP) was used to perform an automatic estimation of SAHS severity (no-SAHS, mild, moderate, and severe), by the use of the selected spectral features from: airflow NPP alone, airflow TH alone, and both sensors jointly. The highest diagnostic performance was reached by BY-MLP only trained with NPP spectral features, reaching Cohen's κ = 0.498 in the overall four-class classification task. It also achieved 91.3%, 84.9%, and 83.3% of accuracy in the binary evaluation of the 3 apnea-hypopnea index cut-offs (5, 15, and 30 events/hour) that define the four SAHS degrees. Our results suggest that TH sensor might be not necessary for SAHS severity estimation if an automatic comprehensive characterization approach is adopted to simplify the diagnostic process.