Sleep-stage dependence and co-existence of cardio-respiratory coordination and phase synchronization.

Interactions between the cardiac and respiratory systems play a pivotal role in physiological functioning. Nonetheless, the intricacies of cardio-respiratory couplings, such as cardio-respiratory phase synchronization (CRPS) and cardio-respiratory coordination (CRC), remain elusive, and an automated algorithm for CRC detection is lacking. This paper introduces an automated CRC detection algorithm, which allowed us to conduct a comprehensive comparison of CRPS and CRC during sleep for the first time using an extensive database. We found that CRPS is more sensitive to sleep-stage transitions, and intriguingly, there is a negative correlation between the degree of CRPS and CRC when fluctuations in breathing frequency are high. This comparative analysis holds promise in assisting researchers in gaining deeper insights into the mechanics of and distinctions between these two physiological phenomena. Additionally, the automated algorithms we devised have the potential to offer valuable insights into the clinical applications of CRC and CRPS.

Automatic sleep-stage classification of heart rate and actigraphy data using deep and transfer learning approaches.

Manual sleep-stage scoring based on full-night polysomnography data recorded in a sleep lab has been the gold standard of clinical sleep medicine. This costly and time-consuming approach is unfit for long-term studies as well as assessment of sleep on a population level. With the vast amount of physiological data becoming available from wrist-worn devices, deep learning techniques provide an opportunity for fast and reliable automatic sleep-stage classification tasks. However, training a deep neural network requires large annotated sleep databases, which are not available for long-term epidemiological studies. In this paper, we introduce an end-to-end temporal convolutional neural network able to automatically score sleep stages from raw heartbeat RR interval (RRI) and wrist actigraphy data. Moreover, a transfer learning approach enables the training of the network on a large public database (Sleep Heart Health Study, SHHS) and its subsequent application to a much smaller database recorded by a wristband device. The transfer learning significantly shortens training time and improves sleep-scoring accuracy from 68.9% to 73.8% and inter-rater reliability (Cohen's kappa) from 0.51 to 0.59. We also found that for the SHHS database, automatic sleep-scoring accuracy using deep learning shows a logarithmic relationship with the training size. Although deep learning approaches for automatic sleep scoring are not yet comparable to the inter-rater reliability among sleep technicians, performance is expected to significantly improve in the near future when more large public databases become available. We anticipate those deep learning techniques, when combined with our transfer learning approach, will leverage automatic sleep scoring of physiological data from wearable devices and enable the investigation of sleep in large cohort studies.

Long- and short-term fluctuations compared for several organ systems across sleep stages.

Some details of cardiovascular and cardio-respiratory regulation and their changes during different sleep stages remain still unknown. In this paper we compared the fluctuations of heart rate, pulse rate, respiration frequency, and pulse transit times as well as EEG alpha-band power on time scales from 6 to 200 s during different sleep stages in order to better understand regulatory pathways. The five considered time series were derived from ECG, photoplethysmogram, nasal air flow, and central electrode EEG measurements from full-night polysomnography recordings of 246 subjects with suspected sleep disorders. We applied detrended fluctuation analysis, distinguishing between short-term (6-16 s) and long-term (50-200 s) correlations, i.e., scaling behavior characterized by the fluctuation exponents α 1 and α 2 related with parasympathetic and sympathetic control, respectively. While heart rate (and pulse rate) are characterized by sex and age-dependent short-term correlations, their long-term correlations exhibit the well-known sleep stage dependence: weak long-term correlations during non-REM sleep and pronounced long-term correlations during REM sleep and wakefulness. In contrast, pulse transit times, which are believed to be mainly affected by blood pressure and arterial stiffness, do not show differences between short-term and long-term exponents. This is in constrast to previous results for blood pressure time series, where α 1 was much larger than α 2, and therefore questions a very close relation between pulse transit times and blood pressure values. Nevertheless, very similar sleep-stage dependent differences are observed for the long-term fluctuation exponent α 2 in all considered signals including EEG alpha-band power. In conclusion, we found that the observed fluctuation exponents are very robust and hardly modified by body mass index, alcohol consumption, smoking, or sleep disorders. The long-term fluctuations of all observed systems seem to be modulated by patterns following sleep stages generated in the brain and thus regulated in a similar manner, while short-term regulations differ between the organ systems. Deviations from the reported dependence in any of the signals should be indicative of problems in the function of the particular organ system or its control mechanisms.

The Reconstruction of Causal Networks in Physiology.

We systematically compare strengths and weaknesses of two methods that can be used to quantify causal links between time series: Granger-causality and Bivariate Phase Rectified Signal Averaging (BPRSA). While a statistical test method for Granger-causality has already been established, we show that BPRSA causality can also be probed with existing statistical tests. Our results indicate that more data or stronger interactions are required for the BPRSA method than for the Granger-causality method to detect an existing link. Furthermore, the Granger-causality method can distinguish direct causal links from indirect links as well as links that arise from a common source, while BPRSA cannot. However, in contrast to Granger-causality, BPRSA is suited for the analysis of non-stationary data. We demonstrate the practicability of the Granger-causality method by applying it to polysomnography data from sleep laboratories. An algorithm is presented, which addresses the stationarity condition of Granger-causality by splitting non-stationary data into shorter segments until they pass a stationarity test. We reconstruct causal networks of heart rate, breathing rate, and EEG amplitude from young healthy subjects, elderly healthy subjects, and subjects with obstructive sleep apnea, a condition that leads to disruption of normal respiration during sleep. These networks exhibit differences not only between different sleep stages, but also between young and elderly healthy subjects on the one hand and subjects with sleep apnea on the other hand. Among these differences are 1) weaker interactions in all groups between heart rate, breathing rate and EEG amplitude during deep sleep, compared to light and REM sleep, 2) a stronger causal link from heart rate to breathing rate but disturbances in respiratory sinus arrhythmia (breathing to heart rate coupling) in subjects with sleep apnea, 3) a stronger causal link from EEG amplitude to breathing rate during REM sleep in subjects with sleep apnea. The Granger-causality method, although initially developed for econometric purposes, can provide a quantitative, testable measure for causality in physiological networks.

Connectivity of EEG synchronization networks increases for Parkinson's disease patients with freezing of gait.

Freezing of gait (FoG), a paroxysmal gait disturbance commonly experienced by patients with Parkinson's disease (PD), is characterized by sudden episodes of inability to generate effective forward stepping. Recent studies have shown an increase in beta frequency of local-field potentials in the basal-ganglia during FoG, however, comprehensive research on the synchronization between different brain locations and frequency bands in PD patients is scarce. Here, by developing tools based on network science and non-linear dynamics, we analyze synchronization networks of electroencephalography (EEG) brain waves of three PD patient groups with different FoG severity. We find higher EEG amplitude synchronization (stronger network links) between different brain locations as PD and FoG severity increase. These results are consistent across frequency bands (theta, alpha, beta, gamma) and independent of the specific motor task (walking, still standing, hand tapping) suggesting that an increase in severity of PD and FoG is associated with stronger EEG networks over a broad range of brain frequencies. This observation of a direct relationship of PD/FoG severity with overall EEG synchronization together with our proposed EEG synchronization network approach may be used for evaluating FoG propensity and help to gain further insight into PD and the pathophysiology leading to FoG.

Self-organized multistability in the forest fire model.

The forest fire model in statistical physics represents a paradigm for systems close to but not completely at criticality. For large tree growth probabilities p we identify periodic attractors, where the tree density ρ oscillates between discrete values. For lower p this self-organized multistability persists with incrementing numbers of states. Even at low p the system remains quasiperiodic with a frequency ≈p on the way to chaos. In addition, the power-spectrum shows 1/f^{2} scaling (Brownian noise) at the low frequencies f, which turns into white noise for very long simulation times.

Capability of processed EEG parameters to monitor conscious sedation in endoscopy is similar to general anaesthesia.

BACKGROUND: Reliable and safe sedation is a prerequisite for endoscopic interventions. The current standard is rather safe, yet, an objective device to measure sedation depth is missing. To date, anaesthesia monitors based on processed electroencephalogram (EEG) have not been utilised in conscious sedation. OBJECTIVE: To investigate EEG parameters to differentiate consciousness in endoscopic propofol sedation. METHODS: In total, 171 patients aged 21-83 years (ASA I-III) undergoing gastrointestinal and bronchial endoscopy were enrolled. Standard monitoring and a frontotemporal two-channel EEG were recorded. The state of consciousness was identified by repeated requests to squeeze the investigator's hand. RESULTS: In total, 1132 state-of-consciousness (SOC) transitions were recorded in procedures ranging from 5 to 69 min. Thirty-four EEG parameters from the frequency domain, time-frequency domain and complexity measures were calculated. Area under the curve ranged from 0.51 to 0.82 with complexity and optimised frequency domain parameters yielding the best results. CONCLUSION: Prediction of the SOC with processed EEG parameters is feasible, and the results for sedation in endoscopic procedures are similar to those reported from general anaesthesia. These results are insufficient for a clinical application, but prediction capability may be increased with optimisation and modelling.

Reconstruction of the respiratory signal through ECG and wrist accelerometer data.

Respiratory rate and changes in respiratory activity provide important markers of health and fitness. Assessing the breathing signal without direct respiratory sensors can be very helpful in large cohort studies and for screening purposes. In this paper, we demonstrate that long-term nocturnal acceleration measurements from the wrist yield significantly better respiration proxies than four standard approaches of ECG (electrocardiogram) derived respiration. We validate our approach by comparison with flow-derived respiration as standard reference signal, studying the full-night data of 223 subjects in a clinical sleep laboratory. Specifically, we find that phase synchronization indices between respiration proxies and the flow signal are large for five suggested acceleration-derived proxies with [Formula: see text] for males and [Formula: see text] for females (means ± standard deviations), while ECG-derived proxies yield only [Formula: see text] for males and [Formula: see text] for females. Similarly, respiratory rates can be determined more precisely by wrist-worn acceleration devices compared with a derivation from the ECG. As limitation we must mention that acceleration-derived respiration proxies are only available during episodes of non-physical activity (especially during sleep).

Unique sleep-stage transitions determined by obstructive sleep apnea severity, age and gender.

In obstructive sleep apnea, patients' sleep is fragmented leading to excessive daytime sleepiness and co-morbidities like arterial hypertension. However, traditional metrics are not always directly correlated with daytime sleepiness, and the association between traditional sleep quality metrics like sleep duration and arterial hypertension is still ambiguous. In a development cohort, we analysed hypnograms from mild (n = 209), moderate (n = 222) and severe (n = 272) obstructive sleep apnea patients as well as healthy controls (n = 105) from the European Sleep Apnea Database. We assessed sleep by the analysis of two-step transitions depending on obstructive sleep apnea severity and anthropometric factors. Two-step transition patterns were examined for an association to arterial hypertension or daytime sleepiness. We also tested cumulative distributions of wake as well as sleep-states for power-laws (exponent α) and exponential distributions (decay time τ) in dependency on obstructive sleep apnea severity and potential confounders. Independent of obstructive sleep apnea severity and potential confounders, wake-state durations followed a power-law distribution, while sleep-state durations were characterized by an exponential distribution. Sleep-stage transitions are influenced by obstructive sleep apnea severity, age and gender. N2 → N3 → wake transitions were associated with high diastolic blood pressure. We observed higher frequencies of alternating (symmetric) patterns (e.g. N2 → N1 → N2, N2 → wake → N2) in sleepy patients both in the development cohort and in a validation cohort (n = 425). In conclusion, effects of obstructive sleep apnea severity and potential confounders on sleep architecture are small, but transition patterns still link sleep fragmentation directly to obstructive sleep apnea-related clinical outcomes like arterial hypertension and daytime sleepiness.

A Biased Diffusion Approach to Sleep Dynamics Reveals Neuronal Characteristics.

We propose a biased diffusion model of accumulated subthreshold voltage fluctuations in wake-promoting neurons to account for stochasticity in sleep dynamics and to explain the occurrence of brief arousals during sleep. Utilizing this model, we derive four neurophysiological parameters related to neuronal noise level, excitability threshold, deep-sleep threshold, and sleep inertia. We provide the first analytic expressions for these parameters, and we show that there is good agreement between empirical findings from sleep recordings and our model simulation results. Our work suggests that these four parameters can be of clinical importance because we find them to be significantly altered in elderly subjects and in children with autism.

Excessive phase synchronization in cortical activation during locomotion in persons with Parkinson's disease.

INTRODUCTION: Parkinson's disease (PD) is characterized by gait disturbances, which become severe during the advanced stages of the disease. Though gait impairments in Parkinson's disease have been extensively described in terms of spatiotemporal gait parameters, little is known regarding associated patterns of cortical activity. The objective of the present study is to test if interhemispheric synchronization differs between participants with PD and healthy elderly controls (NPD). We analyzed electroencephalography (EEG) signals recorded during bilateral movements, i.e., locomotion and hand tapping. METHODS: Fifteen participants with PD ('OFF' their anti-parkinsonian medications) and eight NPD were assessed during quiet standing, straight-line walking, turning, and hand tapping tasks. Using a 32-electrode EEG array, we quantified the synchronization in periodic cortical activation between the brain hemispheres (interhemispheric phase synchronization; inter-PS). Theta, alpha, beta, and gamma bands were evaluated. RESULTS: In all bands, inter-PS was significantly higher for the PD group as compared with the NPD group during standing and walking (p < 0.001) and during bimanual tasks (p = 0.026). CONCLUSIONS: Persons with PD exhibit increased inter-PS as compared with NPD participants. These findings support previous evidence from animal studies, that bilateral cortical hypersynchronization emerges from the asymmetric neural degeneration that is at the base of the disease. Future studies should elucidate the long-term temporal development of this hypersynchronization and its clinical relevance (e.g., can it 'serve' as prodromal marker?).

Coupling Between Leg Muscle Activation and EEG During Normal Walking, Intentional Stops, and Freezing of Gait in Parkinson's Disease.

In this paper, we apply novel techniques for characterizing leg muscle activation patterns via electromyograms (EMGs) and for relating them to changes in electroencephalogram (EEG) activity during gait experiments. Specifically, we investigate changes of leg-muscle EMG amplitudes and EMG frequencies during walking, intentional stops, and unintended freezing-of-gait (FOG) episodes. FOG is a frequent paroxysmal gait disturbance occurring in many patients suffering from Parkinson's disease (PD). We find that EMG amplitudes and frequencies do not change significantly during FOG episodes with respect to walking, while drastic changes occur during intentional stops. Phase synchronization between EMG signals is most pronounced during walking in controls and reduced in PD patients. By analyzing cross-correlations between changes in EMG patterns and brain-wave amplitudes (from EEGs), we find an increase in EEG-EMG coupling at the beginning of stop and FOG episodes. Our results may help to better understand the enigmatic pathophysiology of FOG, to differentiate between FOG events and other gait disturbances, and ultimately to improve diagnostic procedures for patients suffering from PD.

Detection and analysis of pulse waves during sleep via wrist-worn actigraphy.

The high temporal and intensity resolution of modern accelerometers gives the opportunity of detecting even tiny body movements via motion-based sensors. In this paper, we demonstrate and evaluate an approach to identify pulse waves and heartbeats from acceleration data of the human wrist during sleep. Specifically, we have recorded simultaneously full-night polysomnography and 3d wrist actigraphy data of 363 subjects during one night in a clinical sleep laboratory. The acceleration data was segmented and cleaned, excluding body movements and separating episodes with different sleep positions. Then, we applied a bandpass filter and a Hilbert transform to uncover the pulse wave signal, which worked well for an average duration of 1.7 h per subject. We found that 81 percent of the detected pulse wave intervals could be correctly associated with the R peak intervals from independently recorded ECGs and obtained a median Pearson cross-correlation of 0.94. While the low-frequency components of both signals were practically identical, the high-frequency component of the pulse wave interval time series was increased, indicating a respiratory modulation of pulse transit times, probably as an additional contribution to respiratory sinus arrhythmia. Our approach could be used to obtain long-term nocturnal heartbeat interval time series and pulse wave signals from wrist-worn accelerometers without the need of recording ECG or photoplethysmography. This is particularly useful for an ambulatory monitoring of high-risk cardiac patients as well as for assessing cardiac dynamics in large cohort studies solely with accelerometer devices that are already used for activity tracking and sleep pattern analysis.

Performance-based approach for movement artifact removal from electroencephalographic data recorded during locomotion.

The appreciation for the need to record electroencephalographic (EEG) signals from humans while walking has been steadily growing in recent years, particularly in relation to understanding gait disturbances. Movement artefacts (MA) in EEG signals originate from mechanical forces applied to the scalp electrodes, inducing small electrode movements relative to the scalp which, in turn, cause the recorded voltage to change irrespectively of cortical activity. These mechanical forces, and thus MA, may have various sources (e.g., ground reaction forces, head movements, etc.) that are inherent to daily activities, notably walking. In this paper we introduce a systematic, integrated methodology for removing MA from EEG signals recorded during treadmill (TM) and over-ground (OG) walking, as well as quantify the prevalence of MA in different locomotion settings. In our experiments, participants performed walking trials at various speeds both OG and on a TM while wearing a 32-channel EEG cap and a 3-axis accelerometer, placed on the forehead. Data preprocessing included separating the EEG signals into statistically independent additive components using independent component analysis (ICA). We observed an increase in electro-physiological signals (e.g., neck EMG activations for stabilizing the head during heel-strikes) as the walking speed increased. These artefact independent-components (ICs), while not originating from electrode movement, still exhibit a similar spectral pattern to the MA ICs-a peak at the stepping frequency. MA was identified and quantified in each component using a novel method that utilizes the participant's stepping frequency, derived from a forehead-mounted accelerometer. We then benchmarked the EEG data by applying newly established metrics to quantify the success of our method in cleaning the data. The results indicate that our approach can be successfully applied to EEG data recorded during TM and OG walking, and is offered as a unified methodology for MA removal from EEG collected during gait trials.

Nocturnal Dynamics of Sleep-Wake Transitions in Patients With Narcolepsy.

Introduction: We investigate how characteristics of sleep-wake dynamics in humans are modified by narcolepsy, a clinical condition that is supposed to destabilize sleep-wake regulation. Subjects with and without cataplexy are considered separately. Differences in sleep scoring habits as a possible confounder have been examined. Aims and Methods: Four groups of subjects are considered: narcolepsy patients from China with (n = 88) and without (n = 15) cataplexy, healthy controls from China (n = 110) and from Europe (n = 187, 2 nights each). After sleep-stage scoring and calculation of sleep characteristic parameters, the distributions of wake-episode durations and sleep-episode durations are determined for each group and fitted by power laws (exponent α) and by exponentials (decay time τ). Results: We find that wake duration distributions are consistent with power laws for healthy subjects (China: α = 0.88, Europe: α = 1.02). Wake durations in all groups of narcolepsy patients, however, follow the exponential law (τ = 6.2-8.1 min). All sleep duration distributions are best fitted by exponentials on long time scales (τ = 34-82 min). Conclusions: We conclude that narcolepsy mainly alters the control of wake-episode durations but not sleep-episode durations, irrespective of cataplexy. Observed distributions of shortest wake and sleep durations suggest that differences in scoring habits regarding the scoring of short-term sleep stages may notably influence the fitting parameters but do not affect the main conclusion.

Quantifying cardio-respiratory phase synchronization-a comparison of five methods using ECGs of post-infarction patients.

OBJECTIVE: Phase synchronization between two weakly coupled oscillators occurs in many natural systems. Since it is difficult to unambiguously detect such synchronization in experimental data, several methods have been proposed for this purpose. Five popular approaches are systematically optimized and compared here. APPROACH: We study and apply the automated synchrogram method, the reduced synchrogram method, two variants of a gradient method, and the Fourier mode method, analyzing 24h data records from 1455 post-infarction patients, the same data with artificial inaccuracies, and corresponding surrogate data generated by Fourier phase randomization. MAIN RESULTS: We find that the automated synchrogram method is the most robust of all studied approaches when applied to records with missing data or artifacts, whereas the gradient methods should be preferred for noisy data and low-accuracy R-peak positions. We also show that a strong circadian rhythm occurs with much more frequent phase synchronization episodes observed during night time than during day time by all five methods. SIGNIFICANCE: In specific applications, the identified characteristic differences as well as strengths and weaknesses of each method in detecting episodes of cardio-respiratory phase synchronization will be useful for selecting an appropriate method with respect to the type of systematic and dynamical noise in the data.

Age and gender dependency of physiological networks in sleep.

Recently, time delay stability analysis of biosignals has been successfully applied as a multivariate time series analysis method to assess the human physiological network in young adults. The degree of connectivity between different network nodes is described by the so-called link strength. Based on polysomnographic recordings (PSGs), it could be shown that the network changes with the sleep stage. Here, we apply the method to a large set of healthy controls spanning six decades of age. As it is well known, that the overall sleep architecture is dependent both on age and on gender, we particularly address the question, if these changes are also found in the network dynamics. We find moderate dependencies of the network on gender. Significantly higher link strengths up to 13% are found in women for some links in different frequency bands of central and occipital regions in REM and light sleep (N2). Higher link strengths are found in men consistently in cardio-cerebral links in N2, but not significant. Age dependency is more pronounced. In particular a significant overall weakening of the network with age is found for wakefulness and non-REM sleep stages. The largest overall decrease is observed in N2 with 0.017 per decade. For individual links decrease rates up to 0.08 per decade are found, in particular for intra-brain links in non-REM sleep. Many of them show a significant decrease with age. Non-linear regression employing an artificial neural network can predict the age with a mean absolute error (MAE) of about five years, suggesting that an age-resolution of about a decade would be appropriate in normative data for physiological networks.

Modulations of Heart Rate, ECG, and Cardio-Respiratory Coupling Observed in Polysomnography.

The cardiac component of cardio-respiratory polysomnography is covered by ECG and heart rate recordings. However, their evaluation is often underrepresented in summarizing reports. As complements to EEG, EOG, and EMG, these signals provide diagnostic information for autonomic nervous activity during sleep. This review presents major methodological developments in sleep research regarding heart rate, ECG, and cardio-respiratory couplings in a chronological (historical) sequence. It presents physiological and pathophysiological insights related to sleep medicine obtained by new technical developments. Recorded nocturnal ECG facilitates conventional heart rate variability (HRV) analysis, studies of cyclical variations of heart rate, and analysis of ECG waveform. In healthy adults, the autonomous nervous system is regulated in totally different ways during wakefulness, slow-wave sleep, and REM sleep. Analysis of beat-to-beat heart-rate variations with statistical methods enables us to estimate sleep stages based on the differences in autonomic nervous system regulation. Furthermore, up to some degree, it is possible to track transitions from wakefulness to sleep by analysis of heart-rate variations. ECG and heart rate analysis allow assessment of selected sleep disorders as well. Sleep disordered breathing can be detected reliably by studying cyclical variation of heart rate combined with respiration-modulated changes in ECG morphology (amplitude of R wave and T wave).

Sleep Assessment in Large Cohort Studies with High-Resolution Accelerometers.

Accelerometry can be a practical replacement for polysomnography in large observational studies. This review discusses the need for sleep characterization in large observational studies, exemplified by the practices of the ongoing German National Cohort study. After brief descriptions of the physical principles and state-of-the-art accelerometer devices and an overview of public data analysis algorithms for sleep-wake differentiation, we demonstrate that the spectral properties of acceleration data provide additional features that can be exploited. This leads to a periodogram-based sleep detection algorithm. Finally, we address issues of data handling and quality assurance in large cohort studies.

The Detection of Emerging Trends Using Wikipedia Traffic Data and Context Networks.

Can online media predict new and emerging trends, since there is a relationship between trends in society and their representation in online systems? While several recent studies have used Google Trends as the leading online information source to answer corresponding research questions, we focus on the online encyclopedia Wikipedia often used for deeper topical reading. Wikipedia grants open access to all traffic data and provides lots of additional (semantic) information in a context network besides single keywords. Specifically, we suggest and study context-normalized and time-dependent measures for a topic's importance based on page-view time series of Wikipedia articles in different languages and articles related to them by internal links. As an example, we present a study of the recently emerging Big Data market with a focus on the Hadoop ecosystem, and compare the capabilities of Wikipedia versus Google in predicting its popularity and life cycles. To support further applications, we have developed an open web platform to share results of Wikipedia analytics, providing context-rich and language-independent relevance measures for emerging trends.