Contactless Camera-Based Sleep Staging: The HealthBed Study

Temporal dynamics of awakenings from slow-wave sleep in non-rapid eye movement parasomnia

Non-rapid eye movement parasomnia disorders, also called disorders of arousal, are characterized by abnormal nocturnal behaviours, such as confusional arousals or sleep walking. Their pathophysiology is not yet fully understood, and objective diagnostic criteria are lacking. It is known, however, that behavioural episodes occur mostly in the beginning of the night, after an increase in slow-wave activity during slow-wave sleep. A better understanding of the prospect of such episodes may lead to new insights in the underlying mechanisms and eventually facilitate objective diagnosis. We investigated temporal dynamics of transitions from slow-wave sleep of 52 patients and 79 controls. Within the patient group, behavioural and non-behavioural N3 awakenings were distinguished. Patients showed a higher probability to wake up after an N3 bout ended than controls, and this probability increased with N3 bout duration. Bouts longer than 15 min resulted in an awakening in 73% and 34% of the time in patients and controls, respectively. Behavioural episodes reduced over sleep cycles due to a reduction in N3 sleep and a reducing ratio between behavioural and non-behavioural awakenings. In the first two cycles, N3 bouts prior to non-behavioural awakenings were significantly shorter than N3 bouts advancing behavioural awakenings in patients, and N3 awakenings in controls. Our findings provide insights in the timing and prospect of both behavioural and non-behavioural awakenings from N3, which may result in prediction and potentially prevention of behavioural episodes. This work, moreover, leads to a more complete characterization of a prototypical hypnogram of parasomnias, which could facilitate diagnosis.

Studying sleep: towards the identification of hypnogram features that drive expert interpretation

STUDY OBJECTIVES

Hypnograms contain a wealth of information and play an important role in sleep medicine. However, interpretation of the hypnogram is a difficult task and requires domain knowledge and "clinical intuition." This study aimed to uncover which features of the hypnogram drive interpretation by physicians. In other words, make explicit which features physicians implicitly look for in hypnograms.

METHODS

Three sleep experts evaluated up to 612 hypnograms, indicating normal or abnormal sleep structure and suspicion of disorders. ElasticNet and convolutional neural network classification models were trained to predict the collected expert evaluations using hypnogram features and stages as input. The models were evaluated using several measures, including accuracy, Cohen's kappa, Matthew's correlation coefficient, and confusion matrices. Finally, model coefficients and visual analytics techniques were used to interpret the models to associate hypnogram features with expert evaluation.

RESULTS

Agreement between models and experts (Kappa between 0.47 and 0.52) is similar to agreement between experts (Kappa between 0.38 and 0.50). Sleep fragmentation, measured by transitions between sleep stages per hour, and sleep stage distribution were identified as important predictors for expert interpretation.

CONCLUSIONS

By comparing hypnograms not solely on an epoch-by-epoch basis, but also on these more specific features that are relevant for the evaluation of experts, performance assessment of (automatic) sleep-staging and surrogate sleep trackers may be improved. In particular, sleep fragmentation is a feature that deserves more attention as it is often not included in the PSG report, and existing (wearable) sleep trackers have shown relatively poor performance in this aspect.

Deep transfer learning for automated single-lead EEG sleep staging with channel and population mismatches

Introduction: Automated sleep staging using deep learning models typically requires training on hundreds of sleep recordings, and pre-training on public databases is therefore common practice. However, suboptimal sleep stage performance may occur from mismatches between source and target datasets, such as differences in population characteristics (e.g., an unrepresented sleep disorder) or sensors (e.g., alternative channel locations for wearable EEG). Methods: We investigated three strategies for training an automated single-channel EEG sleep stager: pre-training (i.e., training on the original source dataset), training-from-scratch (i.e., training on the new target dataset), and fine-tuning (i.e., training on the original source dataset, fine-tuning on the new target dataset). As source dataset, we used the F3-M2 channel of healthy subjects (N = 94). Performance of the different training strategies was evaluated using Cohen's Kappa (κ) in eight smaller target datasets consisting of healthy subjects (N = 60), patients with obstructive sleep apnea (OSA, N = 60), insomnia (N = 60), and REM sleep behavioral disorder (RBD, N = 22), combined with two EEG channels, F3-M2 and F3-F4. Results: No differences in performance between the training strategies was observed in the age-matched F3-M2 datasets, with an average performance across strategies of κ = .83 in healthy, κ = .77 in insomnia, and κ = .74 in OSA subjects. However, in the RBD set, where data availability was limited, fine-tuning was the preferred method (κ = .67), with an average increase in κ of .15 to pre-training and training-from-scratch. In the presence of channel mismatches, targeted training is required, either through training-from-scratch or fine-tuning, increasing performance with κ = .17 on average. Discussion: We found that, when channel and/or population mismatches cause suboptimal sleep staging performance, a fine-tuning approach can yield similar to superior performance compared to building a model from scratch, while requiring a smaller sample size. In contrast to insomnia and OSA, RBD data contains characteristics, either inherent to the pathology or age-related, which apparently demand targeted training.

The impact of healthy pregnancy on features of heart rate variability and pulse wave morphology derived from wrist-worn photoplethysmography

Due to the association between dysfunctional maternal autonomic regulation and pregnancy complications, tracking non-invasive features of autonomic regulation derived from wrist-worn photoplethysmography (PPG) measurements may allow for the early detection of deteriorations in maternal health. However, even though a plethora of these features-specifically, features describing heart rate variability (HRV) and the morphology of the PPG waveform (morphological features)-exist in the literature, it is unclear which of these may be valuable for tracking maternal health. As an initial step towards clarity, we compute comprehensive sets of HRV and morphological features from nighttime PPG measurements. From these, using logistic regression and stepwise forward feature elimination, we identify the features that best differentiate healthy pregnant women from non-pregnant women, since these likely capture physiological adaptations necessary for sustaining healthy pregnancy. Overall, morphological features were more valuable for discriminating between pregnant and non-pregnant women than HRV features (area under the receiver operating characteristics curve of 0.825 and 0.74, respectively), with the systolic pulse wave deterioration being the most valuable single feature, followed by mean heart rate (HR). Additionally, we stratified the analysis by sleep stages and found that using features calculated only from periods of deep sleep enhanced the differences between the two groups. In conclusion, we postulate that in addition to HRV features, morphological features may also be useful in tracking maternal health and suggest specific features to be included in future research concerning maternal health.

Maternal cardiorespiratory coupling: differences between pregnant and nonpregnant women are further amplified by sleep-stage stratification

Pregnancy complications are associated with abnormal maternal autonomic regulation. Subsequently, thoroughly understanding maternal autonomic regulation during healthy pregnancy may enable the earlier detection of complications, in turn allowing for the improved management thereof. Under healthy autonomic regulation, reciprocal interactions occur between the cardiac and respiratory systems, i.e., cardiorespiratory coupling (CRC). Here, we investigate, for the first time, the differences in CRC between healthy pregnant and nonpregnant women. We apply two algorithms, namely, synchrograms and bivariate phase-rectified signal averaging, to nighttime recordings of ECG and respiratory signals. We find that CRC is present in both groups. Significantly less (P < 0.01) cardiorespiratory synchronization occurs in pregnant women (11% vs. 15% in nonpregnant women). Moreover, there is a smaller response in the heart rate of pregnant women corresponding to respiratory inhalations and exhalations. In addition, we stratified these analyses by sleep stages. As each sleep stage is governed by different autonomic states, this stratification not only amplified some of the differences between groups but also brought out differences that remained hidden when analyzing the full-night recordings. Most notably, the known positive relationship between CRC and deep sleep is less prominent in pregnant women than in their nonpregnant counterparts. The decrease in CRC during healthy pregnancy may be attributable to decreased maternal parasympathetic activity, anatomical changes to the maternal respiratory system, and the increased physiological stress accompanying pregnancy. This work offers novel insight into the physiology of healthy pregnancy and forms part of the base knowledge needed to detect abnormalities in pregnancy.NEW & NOTEWORTHY We compare CRC, i.e., the reciprocal interaction between the cardiac and respiratory systems, between healthy pregnant and nonpregnant women for the first time. Although CRC is present in both groups, CRC is reduced during healthy pregnancy; there is less synchronization between maternal cardiac and respiratory activity and a smaller response in maternal heart rate to respiratory inhalations and exhalations. Stratifying this analysis by sleep stages reveals that differences are most prominent during deep sleep.

Speckle Vibrometry for Instantaneous Heart Rate Monitoring

Instantaneous heart rate (IHR) has been investigated for sleep applications, such as sleep apnea detection and sleep staging. To ensure the comfort of the patient during sleep, it is desirable for IHR to be measured in a contact-free fashion. In this work, we use speckle vibrometry (SV) to perform on-skin and on-textile IHR monitoring in a sleep setting. Minute motions on the laser-illuminated surface can be captured by a defocused camera, enabling the detection of cardiac motions even on textiles. We investigate supine, lateral, and prone sleeping positions. Based on Bland-Altman analysis between SV cardiac measurements and electrocardiogram (ECG), with respect to each position, we achieve the best limits of agreement with ECG values of [-8.65, 7.79] bpm, [-9.79, 9.25] bpm, and [-10.81, 10.23] bpm, respectively. The results indicate the potential of using speckle vibrometry as a contact-free monitoring method for instantaneous heart rate in a setting where the participant is allowed to rest in a spontaneous position while covered by textile layers.

Feasibility of Exploiting Physiological and Motion Features for Camera-based Sleep Staging: A Clinical Study

Camera-based sleep monitoring is an emergent research topic in sleep medicine. The feasibility of using both the physiological features and motion features measured by a video camera for sleep staging was not thoroughly investigated. In this paper, we built a camera-based non-contact sleep monitoring setup in the Institute of Respiratory Diseases, Shenzhen People's Hospital, and created a clinical sleep dataset (nocturnal video data of 11 adults) including the expert-corrected PSG references synchronized with the video. The camera-based measurements have shown high correlations with the PSG. It obtains an overall Mean Absolute Error (MAE) of 1.5 bpm for heart-rate (HR), 0.7 bpm for breathing-rate (BR), 13.9 ms for heart-rate variability (HRV), and an accuracy of 93.5% for leg motion detection. The statistical analysis indicates that the averaged HR and variations of BR are distinct features for annotating four sleep stages (awake, REM, light sleep, and deep sleep). HRV parameter (SDNN) can clearly differentiate rapid-eye-movement (REM) and non-REM, while the leg movement is a distinctive feature for separating awake and sleep. The clinical trial demonstrated the feasibility of using physiological and motion features measured by a camera for joint sleep staging, and provides insights for sleep-related feature selection.

A computationally efficient algorithm for wearable sleep staging in clinical populations

This study describes a computationally efficient algorithm for 4-class sleep staging based on cardiac activity and body movements. Using an accelerometer to calculate gross body movements and a reflective photoplethysmographic (PPG) sensor to determine interbeat intervals and a corresponding instantaneous heart rate signal, a neural network was trained to classify between wake, combined N1 and N2, N3 and REM sleep in epochs of 30 s. The classifier was validated on a hold-out set by comparing the output against manually scored sleep stages based on polysomnography (PSG). In addition, the execution time was compared with that of a previously developed heart rate variability (HRV) feature-based sleep staging algorithm. With a median epoch-per-epoch κ of 0.638 and accuracy of 77.8% the algorithm achieved an equivalent performance when compared to the previously developed HRV-based approach, but with a 50-times faster execution time. This shows how a neural network, without leveraging any a priori knowledge of the domain, can automatically "discover" a suitable mapping between cardiac activity and body movements, and sleep stages, even in patients with different sleep pathologies. In addition to the high performance, the reduced complexity of the algorithm makes practical implementation feasible, opening up new avenues in sleep diagnostics.