From Scalp to Ear-EEG: A Generalizable Transfer Learning Model for Automatic Sleep Scoring in Older People

OBJECTIVE

Sleep monitoring has extensively utilized electroencephalogram (EEG) data collected from the scalp, yielding very large data repositories and well-trained analysis models. Yet, this wealth of data is lacking for emerging, less intrusive modalities, such as ear-EEG.

METHODS AND PROCEDURES

The current study seeks to harness the abundance of open-source scalp EEG datasets by applying models pre-trained on data, either directly or with minimal fine-tuning; this is achieved in the context of effective sleep analysis from ear-EEG data that was recorded using a single in-ear electrode, referenced to the ipsilateral mastoid, and developed in-house as described in our previous work. Unlike previous studies, our research uniquely focuses on an older cohort (17 subjects aged 65-83, mean age 71.8 years, some with health conditions), and employs LightGBM for transfer learning, diverging from previous deep learning approaches.

RESULTS

Results show that the initial accuracy of the pre-trained model on ear-EEG was 70.1%, but fine-tuning the model with ear-EEG data improved its classification accuracy to 73.7%. The fine-tuned model exhibited a statistically significant improvement (p < 0.05, dependent t-test) for 10 out of the 13 participants, as reflected by an enhanced average Cohen's kappa score (a statistical measure of inter-rater agreement for categorical items) of 0.639, indicating a stronger agreement between automated and expert classifications of sleep stages. Comparative SHAP value analysis revealed a shift in feature importance for the N3 sleep stage, underscoring the effectiveness of the fine-tuning process.

CONCLUSION

Our findings underscore the potential of fine-tuning pre-trained scalp EEG models on ear-EEG data to enhance classification accuracy, particularly within an older population and using feature-based methods for transfer learning. This approach presents a promising avenue for ear-EEG analysis in sleep studies, offering new insights into the applicability of transfer learning across different populations and computational techniques.

CLINICAL IMPACT

An enhanced ear-EEG method could be pivotal in remote monitoring settings, allowing for continuous, non-invasive sleep quality assessment in elderly patients with conditions like dementia or sleep apnea.

Physically Meaningful Surrogate Data for COPD

The rapidly increasing prevalence of debilitating breathing disorders, such as chronic obstructive pulmonary disease (COPD), calls for a meaningful integration of artificial intelligence (AI) into respiratory healthcare. Deep learning techniques are "data hungry" whilst patient-based data is invariably expensive and time consuming to record. To this end, we introduce a novel COPD-simulator, a physical apparatus with an easy to replicate design which enables rapid and effective generation of a wide range of COPD-like data from healthy subjects, for enhanced training of deep learning frameworks. To ensure the faithfulness of our domain-aware COPD surrogates, the generated waveforms are examined through both flow waveforms and photoplethysmography (PPG) waveforms (as a proxy for intrathoracic pressure) in terms of duty cycle, sample entropy, FEV1/FVC ratios and flow-volume loops. The proposed simulator operates on healthy subjects and is able to generate FEV1/FVC obstruction ratios ranging from greater than 0.8 to less than 0.2, mirroring values that can observed in the full spectrum of real-world COPD. As a final stage of verification, a simple convolutional neural network is trained on surrogate data alone, and is used to accurately detect COPD in real-world patients. When training solely on surrogate data, and testing on real-world data, a comparison of true positive rate against false positive rate yields an area under the curve of 0.75, compared with 0.63 when training solely on real-world data.

The Deep-Match Framework: R-Peak Detection in Ear-ECG

The Ear-ECG provides a continuous Lead I like electrocardiogram (ECG) by measuring the potential difference related to heart activity by electrodes which are embedded within earphones. However, the significant increase in wearability and comfort enabled by Ear-ECG is often accompanied by a degradation in signal quality - a common obstacle that is shared by the majority of wearable technologies. We aim to resolve this issue by introducing a Deep Matched Filter (Deep-MF) for the highly accurate detection of R-peaks in wearable ECG, thus enhancing the utility of Ear-ECG in real-world scenarios. The Deep-MF consists of an encoder stage (trained as part of an encoder-decoder module to reproduce ground truth ECG), and an R-peak classifier stage. Through its operation as a Matched Filter, the encoder section searches for matches with an ECG template pattern in the input signal, prior to filtering these matches with the subsequent convolutional layers and selecting peaks corresponding to the ground truth ECG. The so condensed latent representation of R-peak information is then fed into a simple R-peak classifier, of which the output provides precise R-peak locations. The proposed Deep Matched Filter is evaluated using leave-one-subject-out cross-validation over 36 subjects with an age range of 18-75, with the Deep-MF outperforming existing algorithms for R-peak detection in noisy ECG. The Deep-MF is benchmarked against a ground truth ECG, in the form of either chest-ECG or arm-ECG, via both R-peak recall and R-peak precision metrics. The Deep-MF achieves a median R-peak recall of 94.9% and a median precision of 91.2% across subjects when evaluated with leave-one-subject-out cross validation. Moreover, when evaluated across a range of thresholds, the Deep-MF achieves an area under the curve (AUC) value of 0.97. The interpretability of Deep-MF as a Matched Filter is further strengthened by the analysis of its response to partial initialisation with an ECG template. We demonstrate that the Deep Matched Filter algorithm not only retains the initialised ECG kernel structure during the training process, but also amplifies portions of the ECG which it deems most valuable - namely the P wave, and each aspect of the QRS complex. Overall, this Deep-Match framework serves as a valuable step forward for the real-world functionality of Ear-ECG and, through its interpretable operation, the acceptance of deep learning models in e-Health.

Hearables: feasibility of recording cardiac rhythms from single in-ear locations

The ear is well positioned to accommodate both brain and vital signs monitoring, via so-called hearable devices. Consequently, ear-based electroencephalography has recently garnered great interest. However, despite the considerable potential of hearable based cardiac monitoring, the biophysics and characteristic cardiac rhythm of ear-based electrocardiography (ECG) are not yet well understood. To this end, we map the cardiac potential on the ear through volume conductor modelling and measurements on multiple subjects. In addition, in order to demonstrate real-world feasibility of in-ear ECG, measurements are conducted throughout a long-time simulated driving task. As a means of evaluation, the correspondence between the cardiac rhythms obtained via the ear-based and standard Lead I measurements, with respect to the shape and timing of the cardiac rhythm, is verified through three measures of similarity: the Pearson correlation, and measures of amplitude and timing deviations. A high correspondence between the cardiac rhythms obtained via the ear-based and Lead I measurements is rigorously confirmed through agreement between simulation and measurement, while the real-world feasibility was conclusively demonstrated through efficacious cardiac rhythm monitoring during prolonged driving. This work opens new avenues for seamless, hearable-based cardiac monitoring that extends beyond heart rate detection to offer cardiac rhythm examination in the community.

The 2023 wearable photoplethysmography roadmap

Photoplethysmography is a key sensing technology which is used in wearable devices such as smartwatches and fitness trackers. Currently, photoplethysmography sensors are used to monitor physiological parameters including heart rate and heart rhythm, and to track activities like sleep and exercise. Yet, wearable photoplethysmography has potential to provide much more information on health and wellbeing, which could inform clinical decision making. This Roadmap outlines directions for research and development to realise the full potential of wearable photoplethysmography. Experts discuss key topics within the areas of sensor design, signal processing, clinical applications, and research directions. Their perspectives provide valuable guidance to researchers developing wearable photoplethysmography technology.

Design and Implementation of an Atrial Fibrillation Detection Algorithm on the ARM Cortex-M4 Microcontroller

At present, a medium-level microcontroller is capable of performing edge computing and can handle the computation of neural network kernel functions. This makes it possible to implement a complete end-to-end solution incorporating signal acquisition, digital signal processing, and machine learning for the classification of cardiac arrhythmias on a small wearable device. In this work, we describe the design and implementation of several classifiers for atrial fibrillation detection on a general-purpose ARM Cortex-M4 microcontroller. We used the CMSIS-DSP library, which supports Naïve Bayes and Support Vector Machine classifiers, with different kernel functions. We also developed Python scripts to automatically transfer the Python model (trained in Scikit-learn) to the C environment. To train and evaluate the models, we used part of the data from the PhysioNet/Computing in Cardiology Challenge 2020 and performed simple classification of atrial fibrillation based on heart-rate irregularity. The performance of the classifiers was tested on a general-purpose ARM Cortex-M4 microcontroller (STM32WB55RG). Our study reveals that among the tested classifiers, the SVM classifier with RBF kernel function achieves the highest accuracy of 96.9%, sensitivity of 98.4%, and specificity of 95.8%. The execution time of this classifier was 720 μs per recording. We also discuss the advantages of moving computing tasks to edge devices, including increased power efficiency of the system, improved patient data privacy and security, and reduced overall system operation costs. In addition, we highlight a problem with false-positive detection and unclear significance of device-detected atrial fibrillation.

Tensor decomposition and machine learning for the detection of arteriovenous fistula stenosis: An initial evaluation

Duplex ultrasound (DUS) is the most widely used method for surveillance of arteriovenous fistulae (AVF) created for dialysis. However, DUS is poor at predicting AVF outcomes and there is a need for novel methods that can more accurately evaluate multidirectional AVF flow. In this study we aimed to evaluate the feasibility of detecting AVF stenosis using a novel method combining tensor-decomposition of B-mode ultrasound cine loops (videos) of blood flow and machine learning classification. Classification of stenosis was based on the DUS assessment of blood flow volume, vessel diameter size, flow velocity, and spectral waveform features. Real-time B-mode cine loops of the arterial inflow, anastomosis, and venous outflow of the AVFs were analysed. Tensor decompositions were computed from both the 'full-frame' (whole-image) videos and 'cropped' videos (to include areas of blood flow only). The resulting output were labelled for the presence of stenosis, as per the DUS findings, and used as a set of features for classification using a Long Short-Term Memory (LSTM) neural network. A total of 61 out of 66 available videos were used for analysis. The whole-image classifier failed to beat random guessing, achieving a mean area under the receiver operating characteristics (AUROC) value of 0.49 (CI 0.48 to 0.50). In contrast, the 'cropped' video classifier performed better with a mean AUROC of 0.82 (CI 0.66 to 0.96), showing promising predictive power despite the small size of the dataset. The combined application of tensor decomposition and machine learning are promising for the detection of AVF stenosis and warrant further investigation.

Hearables: Automatic Sleep Scoring from Single-Channel Ear-EEG in Older Adults

Sleep disorders are a prevalent problem among older adults, yet obtaining an accurate and reliable assessment of sleep quality can be challenging. Traditional polysomnography (PSG) is the gold standard for sleep staging, but is obtrusive, expensive, and requires expert assistance. To this end, we propose a minimally invasive single-channel single ear-EEG automatic sleep staging method for older adults. The method employs features from the frequency, time, and structural complexity domains, which provide a robust classification of sleep stages from a standardised viscoelastic earpiece. Our method is verified on a dataset of older adults and achieves a kappa value of at least 0.61, indicating substantial agreement. This paves the way for a non-invasive, cost-effective, and portable alternative to traditional PSG for sleep staging.

Hearables: Heart Rate Variability from Ear Electrocardiogram and Ear Photoplethysmogram (Ear-ECG and Ear-PPG)

This work aims to classify physiological states using heart rate variability (HRV) features extracted from electrocardiograms recorded in the ears (ear-ECG). The physiological states considered in this work are: (a) normal breathing, (b) controlled slow breathing, and (c) mental exercises. Since both (b) and (c) cause higher variance in heartbeat intervals, breathing-related features (SpO2 and mean breathing interval) from the ear Photoplethysmogram (ear-PPG) are used to facilitate classification. This work: 1) proposes a scheme that, after initialization, automatically extracts R-peaks from low signal-to-noise ratio ear-ECG; 2) verifies the feasibility of extracting meaningful HRV features from ear-ECG; 3) quantitatively compares several ear-ECG sites; and 4) discusses the benefits of combining ear-ECG and ear-PPG features.

Feasibility of Transfer Learning from Finger PPG to In-Ear PPG

The success of deep learning methods has enabled many modern wearable health applications, but has also highlighted the critical caveat of their extremely data hungry nature. While the widely explored wrist and finger photoplethysmography (PPG) sites are less affected, given the large available databases, this issue is prohibitive to exploring the full potential of novel recording locations such as in-ear wearables. To this end, we assess the feasibility of transfer learning from finger PPG to in-ear PPG in the context of deep learning for respiratory monitoring. This is achieved by introducing an encoder-decoder framework which is set up to extract respiratory waveforms from PPG, whereby simultaneously recorded gold standard respiratory waveforms (capnography, impedance pneumography and air flow) are used as a training reference. Next, the data augmentation and training pipeline is examined for both training on finger PPG and the subsequent fine tuning on in-ear PPG. The results indicate that, through training on two large finger PPG data sets (95 subjects) and then retraining on our own small in-ear PPG data set (6 subjects), the model achieves lower and more consistent test error for the prediction of the respiratory waveforms, compared to training on the small in-ear data set alone. This conclusively demonstrates the feasibility of transfer learning from finger PPG to in-ear PPG, leading to better generalisation across a wide range of respiratory rates.

Reducing racial bias in SpO2 estimation: The effects of skin pigmentation

Accurate pulse-oximeter readings are critical for clinical decisions, especially when arterial blood-gas tests - the gold standard for determining oxygen saturation levels - are not available, such as when determining COVID-19 severity. Several studies demonstrate that pulse oxygen saturation estimated from photoplethysmography (PPG) introduces a racial bias due to the more profound scattering of light in subjects with darker skin due to the increased presence of melanin. This leads to an overestimation of blood oxygen saturation in those with darker skin that is increased for low blood oxygen levels and can result in a patient not receiving potentially life-saving supplemental oxygen. This racial bias has been comprehensively studied in conventional finger pulse oximetry but in other less commonly used measurement sites, such as in-ear pulse oximetry, it remains unexplored. Different measurement sites can have thinner epidermis compared with the finger and lower exposure to sunlight (such as is the case with the ear canal), and we hypothesise that this could reduce the bias introduced by skin tone on pulse oximetry. To this end, we compute SpO2 in different body locations, during rest and breath-holds, and compare with the index finger. The study involves a participant pool covering 6-pigmentation categories from Fitzpatrick's Skin Pigmentation scale. These preliminary results indicate that locations characterized by cartilaginous highly vascularized tissues may be less prone to the influence of melanin and pigmentation in the estimation of SpO2, paving the way for the development of non-discriminatory pulse oximetry devices.

Comprehensive Graph Gradual Pruning for Sparse Training in Graph Neural Networks

Graph neural networks (GNNs) tend to suffer from high computation costs due to the exponentially increasing scale of graph data and a large number of model parameters, which restricts their utility in practical applications. To this end, some recent works focus on sparsifying GNNs (including graph structures and model parameters) with the lottery ticket hypothesis (LTH) to reduce inference costs while maintaining performance levels. However, the LTH-based methods suffer from two major drawbacks: 1) they require exhaustive and iterative training of dense models, resulting in an extremely large training computation cost, and 2) they only trim graph structures and model parameters but ignore the node feature dimension, where vast redundancy exists. To overcome the above limitations, we propose a comprehensive graph gradual pruning framework termed CGP. This is achieved by designing a during-training graph pruning paradigm to dynamically prune GNNs within one training process. Unlike LTH-based methods, the proposed CGP approach requires no retraining, which significantly reduces the computation costs. Furthermore, we design a cosparsifying strategy to comprehensively trim all the three core elements of GNNs: graph structures, node features, and model parameters. Next, to refine the pruning operation, we introduce a regrowth process into our CGP framework, to reestablish the pruned but important connections. The proposed CGP is evaluated over a node classification task across six GNN architectures, including shallow models graph convolutional network (GCN) and graph attention network (GAT), shallow-but-deep-propagation models simple graph convolution (SGC) and approximate personalized propagation of neural predictions (APPNP), and deep models GCN via initial residual and identity mapping (GCNII) and residual GCN (ResGCN), on a total of 14 real-world graph datasets, including large-scale graph datasets from the challenging Open Graph Benchmark (OGB). Experiments reveal that the proposed strategy greatly improves both training and inference efficiency while matching or even exceeding the accuracy of the existing methods.

Graph-Regularized Tensor Regression: A Domain-Aware Framework for Interpretable Modeling of Multiway Data on Graphs

Modern data analytics applications are increasingly characterized by exceedingly large and multidimensional data sources. This represents a challenge for traditional machine learning models, as the number of model parameters needed to process such data grows exponentially with the data dimensions, an effect known as the curse of dimensionality. Recently, tensor decomposition (TD) techniques have shown promising results in reducing the computational costs associated with large-dimensional models while achieving comparable performance. However, such tensor models are often unable to incorporate the underlying domain knowledge when compressing high-dimensional models. To this end, we introduce a novel graph-regularized tensor regression (GRTR) framework, whereby domain knowledge about intramodal relations is incorporated into the model in the form of a graph Laplacian matrix. This is then used as a regularization tool to promote a physically meaningful structure within the model parameters. By virtue of tensor algebra, the proposed framework is shown to be fully interpretable, both coefficient-wise and dimension-wise. The GRTR model is validated in a multiway regression setting and compared against competing models and is shown to achieve improved performance at reduced computational costs. Detailed visualizations are provided to help readers gain an intuitive understanding of the employed tensor operations.

An In-Ear PPG-Based Blood Glucose Monitor: A Proof-of-Concept Study

Monitoring diabetes saves lives. To this end, we introduce a novel, unobtrusive, and readily deployable in-ear device for the continuous and non-invasive measurement of blood glucose levels (BGLs). The device is equipped with a low-cost commercially available pulse oximeter whose infrared wavelength (880 nm) is used for the acquisition of photoplethysmography (PPG). For rigor, we considered a full range of diabetic conditions (non-diabetic, pre-diabetic, type I diabetic, and type II diabetic). Recordings spanned nine different days, starting in the morning while fasting, up to a minimum of a two-hour period after eating a carbohydrate-rich breakfast. The BGLs from PPG were estimated using a suite of regression-based machine learning models, which were trained on characteristic features of PPG cycles pertaining to high and low BGLs. The analysis shows that, as desired, an average of 82% of the BGLs estimated from PPG lie in region A of the Clarke error grid (CEG) plot, with 100% of the estimated BGLs in the clinically acceptable CEG regions A and B. These results demonstrate the potential of the ear canal as a site for non-invasive blood glucose monitoring.

Ear-EEG sensitivity modeling for neural sources and ocular artifacts

The ear-EEG has emerged as a promising candidate for real-world wearable brain monitoring. While experimental studies have validated several applications of ear-EEG, the source-sensor relationship for neural sources from across the brain surface has not yet been established. In addition, modeling of the ear-EEG sensitivity to sources of artifacts is still missing. Through volume conductor modeling, the sensitivity of various configurations of ear-EEG is established for a range of neural sources, in addition to ocular artifact sources for the blink, vertical saccade, and horizontal saccade eye movements. Results conclusively support the introduction of ear-EEG into conventional EEG paradigms for monitoring neural activity that originates from within the temporal lobes, while also revealing the extent to which ear-EEG can be used for sources further away from these regions. The use of ear-EEG in scenarios prone to ocular artifacts is also supported, through the demonstration of proportional scaling of artifacts and neural signals in various configurations of ear-EEG. The results from this study can be used to support both existing and prospective experimental ear-EEG studies and applications in the context of sensitivity to both neural sources and ocular artifacts.

A Pilot Study of Heart Rate Variability Synchrony as a Marker of Intraoperative Surgical Teamwork and Its Correlation to the Length of Procedure

Objective: Quality of intraoperative teamwork may have a direct impact on patient outcomes. Heart rate variability (HRV) synchrony may be useful for objective assessment of team cohesion and good teamwork. The primary aim of this study was to investigate the feasibility of using HRV synchrony in surgical teams. Secondary aims were to investigate the association of HRV synchrony with length of procedure (LOP), complications, number of intraoperative glitches and length of stay (LOS). We also investigated the correlation between HRV synchrony and team familiarity, pre- and intraoperative stress levels (STAI questionnaire), NOTECHS score and experience of team members. Methods: Ear, nose and throat (ENT) and vascular surgeons (consultant and registrar team members) were recruited into the study. Baseline demographics including level of team members’ experience were gathered before each procedure. For each procedure, continuous electrocardiogram (ECG) recording was performed and questionnaires regarding pre- and intraoperative stress levels and non-technical skills (NOTECHS) scores were collected for each team member. An independent observer documented the time of each intraoperative glitch. Statistical analysis was conducted using stepwise multiple linear regression. Results: Four HRV synchrony metrics which may be markers of efficient surgical collaboration were identified from the data: 1. number of HRV synchronies per hour of procedure, 2. number of HRV synchrony trends per hour of procedure, 3. length of HRV synchrony trends per hour of procedure, 4. area under the HRV synchrony trend curve per hour of procedure. LOP was inversely correlated with number of HRV synchrony trends per hour of procedure (p < 0.0001), area under HRV synchrony trend curve per hour of procedure (p = 0.001), length of HRV synchrony trends per hour of procedure (p = 0.002) and number of HRV synchronies per hour of procedure (p < 0.0001). LOP was positively correlated with: FS (p = 0.043; R = 0.358) and intraoperative STAI score of the whole team (p = 0.007; R = 0.493). Conclusions: HRV synchrony metrics within operating teams may be used as an objective marker to quantify surgical teamwork. We have shown that LOP is shorter when the intraoperative surgical teams’ HRV is more synchronised.

Supervised Learning for Nonsequential Data: A Canonical Polyadic Decomposition Approach

Efficient modeling of feature interactions underpins supervised learning for nonsequential tasks, characterized by a lack of inherent ordering of features (variables). The brute force approach of learning a parameter for each interaction of every order comes at an exponential computational and memory cost (curse of dimensionality). To alleviate this issue, it has been proposed to implicitly represent the model parameters as a tensor, the order of which is equal to the number of features; for efficiency, it can be further factorized into a compact tensor train (TT) format. However, both TT and other tensor networks (TNs), such as tensor ring and hierarchical Tucker, are sensitive to the ordering of their indices (and hence to the features). To establish the desired invariance to feature ordering, we propose to represent the weight tensor through the canonical polyadic (CP) decomposition (CPD) and introduce the associated inference and learning algorithms, including suitable regularization and initialization schemes. It is demonstrated that the proposed CP-based predictor significantly outperforms other TN-based predictors on sparse data while exhibiting comparable performance on dense nonsequential tasks. Furthermore, for enhanced expressiveness, we generalize the framework to allow feature mapping to arbitrarily high-dimensional feature vectors. In conjunction with feature vector normalization, this is shown to yield dramatic improvements in performance for dense nonsequential tasks, matching models such as fully connected neural networks.

Multivariate Multiscale Cosine Similarity Entropy and Its Application to Examine Circularity Properties in Division Algebras

The extension of sample entropy methodologies to multivariate signals has received considerable attention, with traditional univariate entropy methods, such as sample entropy (SampEn) and fuzzy entropy (FuzzyEn), introduced to measure the complexity of chaotic systems in terms of irregularity and randomness. The corresponding multivariate methods, multivariate multiscale sample entropy (MMSE) and multivariate multiscale fuzzy entropy (MMFE), were developed to explore the structural richness within signals at high scales. However, the requirement of high scale limits the selection of embedding dimension and thus, the performance is unavoidably restricted by the trade-off between the data size and the required high scale. More importantly, the scale of interest in different situations is varying, yet little is known about the optimal setting of the scale range in MMSE and MMFE. To this end, we extend the univariate cosine similarity entropy (CSE) method to the multivariate case, and show that the resulting multivariate multiscale cosine similarity entropy (MMCSE) is capable of quantifying structural complexity through the degree of self-correlation within signals. The proposed approach relaxes the prohibitive constraints between the embedding dimension and data length, and aims to quantify the structural complexity based on the degree of self-correlation at low scales. The proposed MMCSE is applied to the examination of the complex and quaternion circularity properties of signals with varying correlation behaviors, and simulations show the MMCSE outperforming the standard methods, MMSE and MMFE.

Tracking Cognitive Workload in Gaming with In-Ear [Formula: see text]

The feasibility of using in-ear [Formula: see text] to track cognitive workload induced by gaming is investigated. This is achieved by examining temporal variations in cognitive workload through the game Geometry Dash, with 250 trials across 7 subjects. The relationship between performance and cognitive load in Dark Souls III boss fights is also investigated followed by a comparison of the cognitive workload responses across three different genres of game. A robust decrease in in-ear [Formula: see text] is observed in response to cognitive workload induced by gaming, which is consistent with existing results from memory tasks. The results tentatively suggest that in-ear [Formula: see text] may be able to distinguish cognitive workload alone, whereas heart rate and breathing rate respond similarly to both cognitive workload and stress. This study demonstrates the feasibility of low cost wearable cognitive workload tracking in gaming with in-ear [Formula: see text], with applications to the play testing of games and biofeedback in games of the future.

Wearable In-Ear PPG: Detailed Respiratory Variations Enable Classification of COPD

An ability to extract detailed spirometry-like breath-ing waveforms from wearable sensors promises to greatly improve respiratory health monitoring. Photoplethysmography (PPG) has been researched in depth for estimation of respiration rate, given that it varies with respiration through overall intensity, pulse amplitude and pulse interval. We compare and contrast the extraction of these three respiratory modes from both the ear canal and finger and show a marked improvement in the respiratory power for respiration induced intensity variations and pulse amplitude variations when recording from the ear canal. We next employ a data driven multi-scale method, noise assisted multivariate empirical mode decomposition (NA-MEMD), which allows for simultaneous analysis of all three respiratory modes to extract detailed respiratory waveforms from in-ear PPG. For rigour, we considered in-ear PPG recordings from healthy subjects, both older and young, patients with chronic obstructive pulmonary disease (COPD) and idiopathic pulmonary fibrosis (IPF) and healthy subjects with artificially obstructed breathing. Specific in-ear PPG waveform changes are observed for COPD, such as a decreased inspiratory duty cycle and an increased inspiratory magnitude, when compared with expiratory magnitude. These differences are used to classify COPD from healthy and IPF waveforms with a sensitivity of 87% and an overall accuracy of 92%. Our findings indicate the promise of in-ear PPG for COPD screening and unobtrusive respiratory monitoring in ambulatory scenarios and in consumer wearables.

Variational Embedding Multiscale Sample Entropy: A Tool for Complexity Analysis of Multichannel Systems

Entropy-based methods have received considerable attention in the quantification of structural complexity of real-world systems. Among numerous empirical entropy algorithms, conditional entropy-based methods such as sample entropy, which are associated with amplitude distance calculation, are quite intuitive to interpret but require excessive data lengths for meaningful evaluation at large scales. To address this issue, we propose the variational embedding multiscale sample entropy (veMSE) method and conclusively demonstrate its ability to operate robustly, even with several times shorter data than the existing conditional entropy-based methods. The analysis reveals that veMSE also exhibits other desirable properties, such as the robustness to the variation in embedding dimension and noise resilience. For rigor, unlike the existing multivariate methods, the proposed veMSE assigns a different embedding dimension to every data channel, which makes its operation independent of channel permutation. The veMSE is tested on both stimulated and real world signals, and its performance is evaluated against the existing multivariate multiscale sample entropy methods. The proposed veMSE is also shown to exhibit computational advantages over the existing amplitude distance-based entropy methods.

Hearables: Making Sense from Motion Artefacts in Ear-EEG for Real-Life Human Activity Classification

Ear-worn devices are rapidly gaining popularity as they provide the means for measuring vital signals in an unobtrusive, 24/7 wearable and discrete fashion. Naturally, these devices are prone to motion artefacts when used in out-of-lab environments, various movements and activities cause relative movement between user's skin and the electrodes. Historically, these artefacts are seen as nuisance resulting in discarding the segments of signal wherever such artefacts are present. In this work, we make use of such artefacts to classify different daily activities that include sitting, speaking aloud, chewing and walking. To this end, multiple classification techniques are employed to identify these activities using 8 features calculated from the electrode and microphone signal embedded in a generic multimodal in-ear sensor. The results show an overall training accuracy of 93% and 90% and a testing accuracy of 85% and 80% when using a KNN and a 2-layer neural network respectively, thus providing a much needed, simple and reliable framework for real-life human activity classification.

Tensor Networks for Latent Variable Analysis: Novel Algorithms for Tensor Train Approximation

Decompositions of tensors into factor matrices, which interact through a core tensor, have found numerous applications in signal processing and machine learning. A more general tensor model that represents data as an ordered network of subtensors of order-2 or order-3 has, so far, not been widely considered in these fields, although this so-called tensor network (TN) decomposition has been long studied in quantum physics and scientific computing. In this article, we present novel algorithms and applications of TN decompositions, with a particular focus on the tensor train (TT) decomposition and its variants. The novel algorithms developed for the TT decomposition update, in an alternating way, one or several core tensors at each iteration and exhibit enhanced mathematical tractability and scalability for large-scale data tensors. For rigor, the cases of the given ranks, given approximation error, and the given error bound are all considered. The proposed algorithms provide well-balanced TT-decompositions and are tested in the classic paradigms of blind source separation from a single mixture, denoising, and feature extraction, achieving superior performance over the widely used truncated algorithms for TT decomposition.

In-Ear SpO2: A Tool for Wearable, Unobtrusive Monitoring of Core Blood Oxygen Saturation

The non-invasive estimation of blood oxygen saturation (SpO2) by pulse oximetry is of vital importance clinically, from the detection of sleep apnea to the recent ambulatory monitoring of hypoxemia in the delayed post-infective phase of COVID-19. In this proof of concept study, we set out to establish the feasibility of SpO2 measurement from the ear canal as a convenient site for long term monitoring, and perform a comprehensive comparison with the right index finger-the conventional clinical measurement site. During resting blood oxygen saturation estimation, we found a root mean square difference of 1.47% between the two measurement sites, with a mean difference of 0.23% higher SpO2 in the right ear canal. Using breath holds, we observe the known phenomena of time delay between central circulation and peripheral circulation with a mean delay between the ear and finger of 12.4 s across all subjects. Furthermore, we document the lower photoplethysmogram amplitude from the ear canal and suggest ways to mitigate this issue. In conjunction with the well-known robustness to temperature induced vasoconstriction, this makes conclusive evidence for in-ear SpO2 monitoring being both convenient and superior to conventional finger measurement for continuous non-intrusive monitoring in both clinical and everyday-life settings.

Hearables: Feasibility and Validation of In-Ear Electrocardiogram

Out-of-clinic, continuous monitoring of vital signs is envisaged to become the backbone of future e-health. The emerging wrist worn devices have already proven to be a success in the measurement of pulse, however, a susceptibility to artefacts and missing data caused by regular motion in everyday activities, and the inability to continuously acquire the electrocardiogram call into question the utility of this technology in future e-Health. With this in mind, the head, and in particular the ear canals, have been investigated as possible locations for wearable devices. The ears offer a stable position relative to the vital signs during everyday activities, such as sitting, walking, running and sleeping, as well as being a practical and widely accepted base for wearable accessories. This all suggests that the ear canals are a most natural location for physiological sensing in the community. This work addresses the feasibility of recording the ECG from the ear canals, from a one-fits-all, user-friendly device. For rigour and clarity, we quantitatively compare the timings of the identified P-, QRS-, and T-waves within Ear-ECG and standard arm-ECG. Finally, to depict a future e-Health scenario for the Ear-ECG technology, a case study with an abnormal heart condition, the ventricular bigeminy, is presented. A comprehensive study over ten subjects demonstrates conclusively the possibility of in-ear cardiac monitoring in normal daily life.

Tensor Networks for Latent Variable Analysis: Higher Order Canonical Polyadic Decomposition

The canonical polyadic decomposition (CPD) is a convenient and intuitive tool for tensor factorization; however, for higher order tensors, it often exhibits high computational cost and permutation of tensor entries, and these undesirable effects grow exponentially with the tensor order. Prior compression of tensor in-hand can reduce the computational cost of CPD, but this is only applicable when the rank R of the decomposition does not exceed the tensor dimensions. To resolve these issues, we present a novel method for CPD of higher order tensors, which rests upon a simple tensor network of representative inter-connected core tensors of orders not higher than 3. For rigor, we develop an exact conversion scheme from the core tensors to the factor matrices in CPD and an iterative algorithm of low complexity to estimate these factor matrices for the inexact case. Comprehensive simulations over a variety of scenarios support the proposed approach.

Scalable automatic sleep staging in the era of Big Data

Numerous automatic sleep staging approaches have been proposed to provide an eHealth alternative to the current gold-standard - hypnogram scoring by human experts. However, a majority of such studies exploit data of limited scale, which compromises both the validation and the reproducibility and transferability of such automatic sleep staging systems in real clinical settings. In addition, the computational issues and physical meaningfulness of the analysis are typically neglected, yet affordable computation is a key criterion in Big Data analytics. To this end, we establish a comprehensive analysis framework to rigorously evaluate the feasibility of automatic sleep staging from multiple perspectives, including robustness with respect to the number of training subjects, model complexity, and different classifiers. This is achieved for a large collection of publicly accessible polysomnography (PSG) data, recorded over 515 subjects. The trade-off between affordable computation and satisfactory accuracy is shown to be fulfilled by an extreme learning machine (ELM) classifier, which in conjunction with the physically meaningful hidden Markov model (HMM) of the transition between the different sleep stages (smoothing model) is shown to achieve both fast computation and the highest average Cohen's kappa value of κ = 0.73 (Substantial Agreement). Finally, it is shown that for accurate and robust automatic sleep staging, a combination of structural complexity (multi-scale entropy) and frequency-domain (spectral edge frequency) features is both computationally affordable and physically meaningful.

A physiology based model of heart rate variability

Heart rate variability (HRV) is governed by the autonomic nervous system (ANS) and is routinely used to estimate the state of body and mind. At the same time, recorded HRV features can vary substantially between people. A model for HRV that (1) correctly simulates observed HRV, (2) reliably functions for multiple scenarios, and (3) can be personalised using a manageable set of parameters, would be a significant step forward toward understanding individual responses to external influences, such as physical and physiological stress. Current HRV models attempt to reproduce HRV characteristics by mimicking the statistical properties of measured HRV signals. The model presented here for the simulation of HRV follows a radically different approach, as it is based on an approximation of the physiology behind the triggering of a heart beat and the biophysics mechanisms of how the triggering process-and thereby the HRV-is governed by the ANS. The model takes into account the metabolisation rates of neurotransmitters and the change in membrane potential depending on transmitter and ion concentrations. It produces an HRV time series that not only exhibits the features observed in real data, but also explains a reduction of low frequency band-power for physically or psychologically high intensity scenarios. Furthermore, the proposed model enables the personalisation of input parameters to the physiology of different people, a unique feature not present in existing methods. All these aspects are crucial for the understanding and application of future wearable health.

A Transition Probability Based Classification Model for Enhanced N1 Sleep stage Identification During Automatic Sleep Stage Scoring

Automatic sleep staging provides a cheaper, faster and more accessible alternative for evaluating sleep patterns and quality compared with manual hypnogram scoring performed by a clinician. Traditionally, classification methods treat sleep stages independently of their temporal order, despite sleep patterns themselves being highly sequential. Such independent sleep stage classification can result in poor sensitivity and precision, in particular when attempting to classify the sleep stage N1, otherwise known as the transition stage of sleep which links periods of wakefulness to periods of deep sleep. To this end, we propose a novel transition sleep classification method which aims to improve classification accuracy. This is achieved by utilising both the temporal information of previous stages and treating the transitions between stages as classes in their own right. Simulations on publicly available polysomnography (PSG) data and a comprehensive performance comparison with standard classifiers demonstrate a marked improvement achieved by the proposed method in both N1 sensitivity and precision across all considered classifiers. This includes an increase in N1 precision from 0.01% to 36.75% in an MLP classifier, and an increase in both accuracy and Cohen's kappa value in two of the three classifiers. Overall best mean performance is obtained by transition classification with a random forest classifier (RF) which achieved a kappa value of κ = 0.75 (substantial agreement), and an N1 stage precision of 58%.

The ClassA Framework: HRV Based Assessment of SNS and PNS Dynamics Without LF-HF Controversies

The powers of the low frequency (LF) and high frequency (HF) components of heart rate variability (HRV) have become the de facto standard metrics in the assessment of the stress response, and the related activities of the sympathetic nervous system (SNS) and the parasympathetic nervous system (PNS). However, the widely adopted physiological interpretations of the LF and HF components in SNS /PNS balance are now questioned, which puts under serious scrutiny stress assessments which employ the LF and HF components. To avoid these controversies, we here introduce the novel Classification Angle (ClassA) framework, which yields a family of metrics which quantify cardiac dynamics in three-dimensions. This is achieved using a finite-difference plot of HRV, which displays successive rates of change of HRV, and is demonstrated to provide sufficient degrees of freedom to determine cardiac deceleration and/or acceleration. The robustness and accuracy of the novel ClassA framework is verified using HRV signals from ten males, recorded during standardized stress tests, consisting of rest, mental arithmetic, meditation, exercise and further meditation. Comparative statistical testing demonstrates that unlike the existing LF-HF metrics, the ClassA metrics are capable of distinguishing both the physical and mental stress epochs from the epochs of no stress, with statistical significance (Bonferroni corrected p-value ≤ 0.025); HF was able to distinguish physical stress from no stress, but was not able to identify mental stress. The ClassA results also indicated that at moderate levels of stress, the extent of parasympathetic withdrawal was greater than the extent of sympathetic activation. Finally, the analyses and the experimental results provide conclusive evidence that the proposed nonlinear approach to quantify cardiac activity from HRV resolves three critical obstacles to current HRV stress assessments: (i) it is not based on controversial assumptions of balance between the LF and HF powers; (ii) its temporal resolution when estimating parasympathetic dominance is as little as 10 s of HRV data, while only 60 s to estimate sympathetic dominance; (iii) unlike LF and HF analyses, the ClassA framework does not require the prohibitive assumption of signal stationarity. The ClassA framework is unique in offering HRV based stress analysis in three-dimensions.

Hearables: Automatic Overnight Sleep Monitoring With Standardized In-Ear EEG Sensor

OBJECTIVE

Advances in sensor miniaturization and computational power have served as enabling technologies for monitoring human physiological conditions in real-world scenarios. Sleep disruption may impact neural function, and can be a symptom of both physical and mental disorders. This study proposes wearable in-ear electroencephalography (ear-EEG) for overnight sleep monitoring as a 24/7 continuous and unobtrusive technology for sleep quality assessment in the community.

METHODS

A total of 22 healthy participants took part in overnight sleep monitoring with simultaneous ear-EEG and conventional full polysomnography recordings. The ear-EEG data were analyzed in the both structural complexity and spectral domains. The extracted features were used for automatic sleep stage prediction through supervized machine learning, whereby the PSG data were manually scored by a sleep clinician.

RESULTS

The agreement between automatic sleep stage prediction based on ear-EEG from a single in-ear sensor and the hypnogram based on the full PSG was 74.1% in the accuracy over five sleep stage classification. This is supported by a substantial agreement in the kappa metric (0.61).

CONCLUSION

The in-ear sensor is feasible for monitoring overnight sleep outside the sleep laboratory and also mitigates technical difficulties associated with PSG. It, therefore, represents a 24/7 continuously wearable alternative to conventional cumbersome and expensive sleep monitoring.

SIGNIFICANCE

The "standardized" one-size-fits-all viscoelastic in-ear sensor is a next generation solution to monitor sleep-this technology promises to be a viable method for readily wearable sleep monitoring in the community, a key to affordable healthcare and future eHealth.

A pilot study of preoperative heart rate variability predicting pain during local anesthetic varicose vein surgery

OBJECTIVE

Local anesthetic endovenous procedures were shown to reduce recovery time, to decrease postoperative pain, and to more quickly return the patient to baseline activities. However, a substantial number of patients experience pain during these procedures. The autonomic nervous system modulates pain perception, and its influence on stress response can be noninvasively quantified using heart rate variability (HRV) indices. The aim of our study was to evaluate whether preoperative baseline HRV can predict intraoperative pain during local anesthetic varicose vein surgery.

METHODS

Patients scheduled for radiofrequency ablation were included in the study. They had their electrocardiograms recorded from a single channel of a custom-made amplifier. Each patient preoperatively filled in forms Y-1 and Y-2 of Spielberger's State and Trait Anxiety Inventory, completed the Aberdeen Varicose Vein Questionnaire, and rated anxiety level on a numeric scale. Postoperatively, patients filled in the pain they felt during the procedure on the numeric pain intensity scale. MATLAB software (MathWorks, Natick, Mass) was used to extract R waves and to generate HRV signals, and a mathematical model was created to predict the pain score for each patient.

RESULTS

In multivariable analysis, we looked into correlation between reported patient's pain score (rPPS) and Aberdeen Varicose Vein Questionnaire score, preoperative forms Y-1 and Y-2, preoperative anxiety level, and predicted patient's pain (pPPS) score. Multivariable analysis found association only between rPPS and pPPS. The pPPS was significantly correlated with rPPS (R = 0.807; P < .001) with accuracy of prediction of 65.2%, which was calculated from R² on a linear regression model.

CONCLUSIONS

This preliminary study shows that preoperative HRV can accurately predict patients' pain, allowing patients with higher predicted score to have the procedure under general anesthesia.

Multiple-Model Adaptive Estimation for 3-D and 4-D Signals: A Widely Linear Quaternion Approach

Quaternion state estimation techniques have been used in various applications, yet they are only suitable for dynamical systems represented by a single known model. In order to deal with model uncertainty, this paper proposes a class of widely linear quaternion multiple-model adaptive estimation (WL-QMMAE) algorithms based on widely linear quaternion Kalman filters and Bayesian inference. The augmented second-order quaternion statistics is employed to capture complete second-order statistical information in improper quaternion signals. Within the WL-QMMAE framework, a widely linear quaternion interacting multiple-model algorithm is proposed to track time-variant model uncertainty, while a widely linear quaternion static multiple-model algorithm is proposed for time-invariant model uncertainty. A performance analysis of the proposed algorithms shows that, as expected, the WL-QMMAE reduces to semiwidely linear QMMAE for [Formula: see text]-improper signals and further reduces to strictly linear QMMAE for proper signals. Simulation results indicate that for improper signals, the proposed WL-QMMAE algorithms exhibit an enhanced performance over their strictly linear counterparts. The effectiveness of the proposed recursive performance analysis algorithm is also validated.

The Female Heart: Sex Differences in the Dynamics of ECG in Response to Stress

Sex differences in the study of the human physiological response to mental stress are often erroneously ignored. To this end, we set out to show that our understanding of the stress response is fundamentally altered once sex differences are taken into account. This is achieved by comparing the heart rate variability (HRV) signals acquired during mental maths tests from ten females and ten males of similar maths ability; all females were in the follicular phase of their menstrual cycle. For rigor, the HRV signals from this pilot study were analyzed using temporal, spectral and nonlinear signal processing techniques, which all revealed significant statistical differences between the sexes, with the stress-induced increases in the heart rates from the males being significantly larger than those from the females (p-value = 4.4 × 10⁻⁴). In addition, mental stress produced an overall increase in the power of the low frequency component of HRV in the males, but caused an overall decrease in the females. The stress-induced changes in the power of the high frequency component were even more profound; it greatly decreased in the males, but increased in the females. We also show that mental stress was followed by the expected decrease in sample entropy, a nonlinear measure of signal regularity, computed from the males' HRV signals, while overall, stress manifested in an increase in the sample entropy computed from the females' HRV signals. This finding is significant, since mental stress is commonly understood to be manifested in the decreased entropy of HRV signals. The significant difference (p-value = 2.1 × 10⁻⁹) between the changes in the entropies from the males and females highlights the pitfalls in ignoring sex in the formation of a physiological hypothesis. Furthermore, it has been argued that estrogen attenuates the effect of catecholamine stress hormones; the findings from this investigation suggest for the first time that the conventionally cited cardiac changes, attributed to the fight-or-flight stress response, are not universally applicable to females. Instead, this pilot study provides an alternative interpretation of cardiac responses to stress in females, which indicates a closer alignment to the evolutionary tend-and-befriend response.

Self-assessment of surgical ward crisis management using video replay augmented with stress biofeedback

Background

We aimed to explore the feasibility and attitudes towards using video replay augmented with real time stress quantification for the self-assessment of clinical skills during simulated surgical ward crisis management.

Methods

Twenty two clinicians participated in 3 different simulated ward based scenarios of deteriorating post-operative patients. Continuous ECG recordings were made for all participants to monitor stress levels using heart rate variability (HRV) indices. Video recordings of simulated scenarios augmented with real time stress biofeedback were replayed to participants. They were then asked to self-assess their performance using an objective assessment tool. Participants attitudes were explored using a post study questionnaire.

Results

Using HRV stress indices, we demonstrated higher stress levels in novice participants. Self-assessment scores were significantly higher in more experienced participants. Overall, participants felt that video replay and augmented stress biofeedback were useful in self-assessment.

Conclusion

Self-assessment using an objective self-assessment tool alongside video replay augmented with stress biofeedback is feasible in a simulated setting and well liked by participants.

Electronic supplementary material

The online version of this article (10.1186/s13037-018-0153-5) contains supplementary material, which is available to authorized users.

A novel in-ear sensor to determine sleep latency during the Multiple Sleep Latency Test in healthy adults with and without sleep restriction

OBJECTIVES

Detecting sleep latency during the Multiple Sleep Latency Test (MSLT) using electroencephalogram (scalp-EEG) is time-consuming. The aim of this study was to evaluate the efficacy of a novel in-ear sensor (in-ear EEG) to detect the sleep latency, compared to scalp-EEG, during MSLT in healthy adults, with and without sleep restriction.

METHODS

We recruited 25 healthy adults (28.5±5.3 years) who participated in two MSLTs with simultaneous recording of scalp and in-ear EEG. Each test followed a randomly assigned sleep restriction (≤5 hours sleep) or usual night sleep (≥7 hours sleep). Reaction time and Stroop test were used to assess the functional impact of the sleep restriction. The EEGs were scored blind to the mode of measurement and study conditions, using American Academy of Sleep Medicine 2012 criteria. The Agreement between the scalp and in-ear EEG was assessed using Bland-Altman analysis.

RESULTS

Technically acceptable data were obtained from 23 adults during 69 out of 92 naps in the sleep restriction condition and 25 adults during 85 out of 100 naps in the usual night sleep. Meaningful sleep restrictions were confirmed by an increase in the reaction time (mean ± SD: 238±30 ms vs 228±27 ms; P=0.045). In the sleep restriction condition, the in-ear EEG exhibited a sensitivity of 0.93 and specificity of 0.80 for detecting sleep latency, with a substantial agreement (κ=0.71), whereas after the usual night's sleep, the in-ear EEG exhibited a sensitivity of 0.91 and specificity of 0.89, again with a substantial agreement (κ=0.79).

CONCLUSION

The in-ear sensor was able to detect reduced sleep latency following sleep restriction, which was sufficient to impair both the reaction time and cognitive function. Substantial agreement was observed between the scalp and in-ear EEG when measuring sleep latency. This new in-ear EEG technology is shown to have a significant value as a convenient measure for sleep latency.

Quantifying team cooperation through intrinsic multi-scale measures: respiratory and cardiac synchronization in choir singers and surgical teams

A highly localized data-association measure, termed intrinsic synchrosqueezing transform (ISC), is proposed for the analysis of coupled nonlinear and non-stationary multivariate signals. This is achieved based on a combination of noise-assisted multivariate empirical mode decomposition and short-time Fourier transform-based univariate and multivariate synchrosqueezing transforms. It is shown that the ISC outperforms six other combinations of algorithms in estimating degrees of synchrony in synthetic linear and nonlinear bivariate signals. Its advantage is further illustrated in the precise identification of the synchronized respiratory and heart rate variability frequencies among a subset of bass singers of a professional choir, where it distinctly exhibits better performance than the continuous wavelet transform-based ISC. We also introduce an extension to the intrinsic phase synchrony (IPS) measure, referred to as nested intrinsic phase synchrony (N-IPS), for the empirical quantification of physically meaningful and straightforward-to-interpret trends in phase synchrony. The N-IPS is employed to reveal physically meaningful variations in the levels of cooperation in choir singing and performing a surgical procedure. Both the proposed techniques successfully reveal degrees of synchronization of the physiological signals in two different aspects: (i) precise localization of synchrony in time and frequency (ISC), and (ii) large-scale analysis for the empirical quantification of physically meaningful trends in synchrony (N-IPS).

Discrimination of emotional states from scalp- and intracranial EEG using multiscale Rényi entropy

A data-adaptive, multiscale version of Rényi's quadratic entropy (RQE) is introduced for emotional state discrimination from EEG recordings. The algorithm is applied to scalp EEG recordings of 30 participants watching 4 emotionally-charged video clips taken from a validated public database. Krippendorff's inter-rater statistic reveals that multiscale RQE of the mid-frontal scalp electrodes best discriminates between five emotional states. Multiscale RQE is also applied to joint scalp EEG, amygdala- and occipital pole intracranial recordings of an implanted patient watching a neutral and an emotionally charged video clip. Unlike for the neutral video clip, the RQEs of the mid-frontal scalp electrodes and the amygdala-implanted electrodes are observed to coincide in the time range where the crux of the emotionally-charged video clip is revealed. In addition, also during this time range, phase synchrony between the amygdala and mid-frontal recordings is maximal, as well as our 30 participants' inter-rater agreement on the same video clip. A source reconstruction exercise using intracranial recordings supports our assertion that amygdala could contribute to mid-frontal scalp EEG. On the contrary, no such contribution was observed for the occipital pole's intracranial recordings. Our results suggest that emotional states discriminated from mid-frontal scalp EEG are likely to be mirrored by differences in amygdala activations in particular when recorded in response to emotionally-charged scenes.

Hearables: feasibility of recording cardiac rhythms from head and in-ear locations

Mobile technologies for the recording of vital signs and neural signals are envisaged to underpin the operation of future health services. For practical purposes, unobtrusive devices are favoured, such as those embedded in a helmet or incorporated onto an earplug. However, these locations have so far been underexplored, as the comparably narrow neck impedes the propagation of vital signals from the torso to the head surface. To establish the principles behind electrocardiogram (ECG) recordings from head and ear locations, we first introduce a realistic three-dimensional biophysics model for the propagation of cardiac electric potentials to the head surface, which demonstrates the feasibility of head-ECG recordings. Next, the proposed biophysics propagation model is verified over comprehensive real-world experiments based on head- and in-ear-ECG measurements. It is shown both that the proposed model is an excellent match for the recordings, and that the quality of head- and ear-ECG is sufficient for a reliable identification of the timing and shape of the characteristic P-, Q-, R-, S- and T-waves within the cardiac cycle. This opens up a range of new possibilities in the identification and management of heart conditions, such as myocardial infarction and atrial fibrillation, based on 24/7 continuous in-ear measurements. The study therefore paves the way for the incorporation of the cardiac modality into future 'hearables', unobtrusive devices for health monitoring.

Pain Prediction From ECG in Vascular Surgery

Varicose vein surgeries are routine outpatient procedures, which are often performed under local anaesthesia. The use of local anaesthesia both minimises the risk to patients and is cost effective, however, a number of patients still experience pain during surgery. Surgical teams must therefore decide to administer either a general or local anaesthetic based on their subjective qualitative assessment of patient anxiety and sensitivity to pain, without any means to objectively validate their decision. To this end, we develop a 3-D polynomial surface fit, of physiological metrics and numerical pain ratings from patients, in order to model the link between the modulation of cardiovascular responses and pain in varicose vein surgeries. Spectral and structural complexity features found in heart rate variability signals, recorded immediately prior to 17 varicose vein surgeries, are used as pain metrics. The so obtained pain prediction model is validated through a leave-one-out validation, and achieved a Kappa coefficient of 0.72 (substantial agreement) and an area below a receiver operating characteristic curve of 0.97 (almost perfect accuracy). This proof-of-concept study conclusively demonstrates the feasibility of the accurate classification of pain sensitivity, and introduces a mathematical model to aid clinicians in the objective administration of the safest and most cost-effective anaesthetic to individual patients.

Physiological artifacts in scalp EEG and ear-EEG

Background

A problem inherent to recording EEG is the interference arising from noise and artifacts. While in a laboratory environment, artifacts and interference can, to a large extent, be avoided or controlled, in real-life scenarios this is a challenge. Ear-EEG is a concept where EEG is acquired from electrodes in the ear.

Methods

We present a characterization of physiological artifacts generated in a controlled environment for nine subjects. The influence of the artifacts was quantified in terms of the signal-to-noise ratio (SNR) deterioration of the auditory steady-state response. Alpha band modulation was also studied in an open/closed eyes paradigm.

Results

Artifacts related to jaw muscle contractions were present all over the scalp and in the ear, with the highest SNR deteriorations in the gamma band. The SNR deterioration for jaw artifacts were in general higher in the ear compared to the scalp. Whereas eye-blinking did not influence the SNR in the ear, it was significant for all groups of scalps electrodes in the delta and theta bands. Eye movements resulted in statistical significant SNR deterioration in both frontal, temporal and ear electrodes. Recordings of alpha band modulation showed increased power and coherence of the EEG for ear and scalp electrodes in the closed-eyes periods.

Conclusions

Ear-EEG is a method developed for unobtrusive and discreet recording over long periods of time and in real-life environments. This study investigated the influence of the most important types of physiological artifacts, and demonstrated that spontaneous activity, in terms of alpha band oscillations, could be recorded from the ear-EEG platform. In its present form ear-EEG was more prone to jaw related artifacts and less prone to eye-blinking artifacts compared to state-of-the-art scalp based systems.

Hearables: Multimodal physiological in-ear sensing

Future health systems require the means to assess and track the neural and physiological function of a user over long periods of time, and in the community. Human body responses are manifested through multiple, interacting modalities – the mechanical, electrical and chemical; yet, current physiological monitors (e.g. actigraphy, heart rate) largely lack in cross-modal ability, are inconvenient and/or stigmatizing. We address these challenges through an inconspicuous earpiece, which benefits from the relatively stable position of the ear canal with respect to vital organs. Equipped with miniature multimodal sensors, it robustly measures the brain, cardiac and respiratory functions. Comprehensive experiments validate each modality within the proposed earpiece, while its potential in wearable health monitoring is illustrated through case studies spanning these three functions. We further demonstrate how combining data from multiple sensors within such an integrated wearable device improves both the accuracy of measurements and the ability to deal with artifacts in real-world scenarios.

Automatic Sleep Monitoring Using Ear-EEG

The monitoring of sleep patterns without patient's inconvenience or involvement of a medical specialist is a clinical question of significant importance. To this end, we propose an automatic sleep stage monitoring system based on an affordable, unobtrusive, discreet, and long-term wearable in-ear sensor for recording the electroencephalogram (ear-EEG). The selected features for sleep pattern classification from a single ear-EEG channel include the spectral edge frequency and multi-scale fuzzy entropy, a structural complexity feature. In this preliminary study, the manually scored hypnograms from simultaneous scalp-EEG and ear-EEG recordings of four subjects are used as labels for two analysis scenarios: 1) classification of ear-EEG hypnogram labels from ear-EEG recordings; and 2) prediction of scalp-EEG hypnogram labels from ear-EEG recordings. We consider both 2-class and 4-class sleep scoring, with the achieved accuracies ranging from 78.5% to 95.2% for ear-EEG labels predicted from ear-EEG, and 76.8% to 91.8% for scalp-EEG labels predicted from ear-EEG. The corresponding Kappa coefficients range from 0.64 to 0.83 for Scenario 1, and indicate substantial to almost perfect agreement, while for Scenario 2 the range of 0.65-0.80 indicates substantial agreement, thus further supporting the feasibility of in-ear sensing for sleep monitoring in the community.

Resolving Ambiguities in the LF/HF Ratio: LF-HF Scatter Plots for the Categorization of Mental and Physical Stress from HRV

It is generally accepted that the activities of the autonomic nervous system (ANS), which consists of the sympathetic (SNS) and parasympathetic nervous systems (PNS), are reflected in the low- (LF) and high-frequency (HF) bands in heart rate variability (HRV)—while, not without some controversy, the ratio of the powers in those frequency bands, the so called LF-HF ratio (LF/HF), has been used to quantify the degree of sympathovagal balance. Indeed, recent studies demonstrate that, in general: (i) sympathovagal balance cannot be accurately measured via the ratio of the LF- and HF- power bands; and (ii) the correspondence between the LF/HF ratio and the psychological and physiological state of a person is not unique. Since the standard LF/HF ratio provides only a single degree of freedom for the analysis of this 2D phenomenon, we propose a joint treatment of the LF and HF powers in HRV within a two-dimensional representation framework, thus providing the required degrees of freedom. By virtue of the proposed 2D representation, the restrictive assumption of the linear dependence between the activity of the autonomic nervous system (ANS) and the LF-HF frequency band powers is demonstrated to become unnecessary. The proposed analysis framework also opens up completely new possibilities for a more comprehensive and rigorous examination of HRV in relation to physical and mental states of an individual, and makes possible the categorization of different stress states based on HRV. In addition, based on instantaneous amplitudes of Hilbert-transformed LF- and HF-bands, a novel approach to estimate the markers of stress in HRV is proposed and is shown to improve the robustness to artifacts and irregularities, critical issues in real-world recordings. The proposed approach for resolving the ambiguities in the standard LF/HF-ratio analyses is verified over a number of real-world stress-invoking scenarios.

Synchrosqueezing an effective method for analyzing Doppler radar physiological signals

Doppler radar can monitor vital sign wirelessly. Respiratory and heart rate have time-varying behavior. Capturing the rate variability provides crucial physiological information. However, the common time-frequency methods fail to detect key information. We investigate Synchrosqueezing method to extract oscillatory components of the signal with time varying spectrum. Simulation and experimental result shows the potential of the proposed method for analyzing signals with complex time-frequency behavior like physiological signals. Respiration and heart signals and their components are extracted with higher resolution and without any pre-filtering and signal conditioning.

Is a Complex-Valued Stepsize Advantageous in Complex-Valued Gradient Learning Algorithms?

Complex gradient methods have been widely used in learning theory, and typically aim to optimize real-valued functions of complex variables. The stepsize of complex gradient learning methods (CGLMs) is a positive number, and little is known about how a complex stepsize would affect the learning process. To this end, we undertake a comprehensive analysis of CGLMs with a complex stepsize, including the search space, convergence properties, and the dynamics near critical points. Furthermore, several adaptive stepsizes are derived by extending the Barzilai-Borwein method to the complex domain, in order to show that the complex stepsize is superior to the corresponding real one in approximating the information in the Hessian. A numerical example is presented to support the analysis.

Quaternion Common Spatial Patterns

A novel quaternion-valued common spatial patterns (QCSP) algorithm is introduced to model co-channel coupling of multi-dimensional processes. To cater for the generality of quaternion-valued non-circular data, we propose a generalized QCSP (G-QCSP) which incorporates the information on power difference between the real and imaginary parts of data channels. As an application, we demonstrate how G-QCSP can be used to provide high classification rates, even at a signal-to-noise ratio (SNR) as low as -10 dB. To illustrate the usefulness of our method in EEG analysis, we employ G-QCSP to extract features for discriminating between imagery left and right hand movements. The classification accuracy using these features is 70%. Furthermore, the proposed method is used to distinguish between Parkinson's disease (PD) patients and healthy control subjects, providing an accuracy of 87%.