Aim: to evaluate the ability of the nonlinear 24-HRV as a predictor of MI using Machine Learning Methods: The sample was composed of 218 patients divided into two groups (Healthy, n=128; MI n=90). The sample dataset is part of the Telemetric and Holter Electrocardiogram Warehouse (THEW) database, from the University of Rochester Medical Center. We used the most common ML algorithms for accuracy comparison with a setting of 10-fold cross-validation (briefly, Linear Regression, Linear Discriminant Analysis, k-Nearest Neighbour, Random Forest, Supporting Vector Machine, Na\"ive Bayes, C 5.0 and Stochastic Gradient Boosting). Results: The main findings of this study show that the combination of SD1nu + SD2nu has greater predictive power for MI in comparison to other HRV indexes. Conclusion: The ML model using nonlinear HRV indexes showed to be more effective than the linear domain, evidenced through the application of ML, represented by a good precision of the Stochastic Gradient Boosting model. Keywords: heart rate variability, machine learning, nonlinear domain, cardiovascular disease
The presence of recurring arrhythmic events (also known as cardiac dysrhythmia or irregular heartbeats), as well as erroneous beat detection due to low signal quality, significantly affects estimation of both time and frequency domain indices of heart rate variability (HRV). A reliable, real-time classification and correction of ECG-derived heartbeats is a necessary prerequisite for an accurate online monitoring of HRV and cardiovascular control. We have developed a novel point-process-based method for real-time R-R interval error detection and correction. Given an R-wave event, we assume that the length of the next R-R interval follows a physiologically motivated, time-varying inverse Gaussian probability distribution. We then devise an instantaneous automated detection and correction procedure for erroneous and arrhythmic beats by using the information on the probability of occurrence of the observed beat provided by the model. We test our algorithm over two datasets from the PhysioNet archive. The Fantasia normal rhythm database is artificially corrupted with known erroneous beats to test both the detection procedure and correction procedure. The benchmark MIT-BIH Arrhythmia database is further considered to test the detection procedure of real arrhythmic events and compare it with results from previously published algorithms. Our automated algorithm represents an improvement over previous procedures, with best specificity for the detection of correct beats, as well as highest sensitivity to missed and extra beats, artificially misplaced beats, and for real arrhythmic events. A near-optimal heartbeat classification and correction, together with the ability to adapt to time-varying changes of heartbeat dynamics in an online fashion, may provide a solid base for building a more reliable real-time HRV monitoring device.
In this paper, we define the term 'DigitalExposome' as a conceptual framework that takes us closer towards understanding the relationship between environment, personal characteristics, behaviour and wellbeing using multimodel mobile sensing technology. Specifically, we simultaneously collected (for the first time) multi-sensor data including urban environmental factors (e.g. air pollution including: PM1, PM2.5, PM10, Oxidised, Reduced, NH3 and Noise, People Count in the vicinity), body reaction (physiological reactions including: EDA, HR, HRV, Body Temperature, BVP and movement) and individuals' perceived responses (e.g. self-reported valence) in urban settings. Our users followed a pre-specified urban path and collected the data using a comprehensive sensing edge devices. The data is instantly fused, time-stamped and geo-tagged at the point of collection. A range of multivariate statistical analysis techniques have been applied including Principle Component Analysis, Regression and spatial visualisations to unravel the relationship between the variables. Results showed that EDA and Heart Rate Variability HRV are noticeably impacted by the level of Particulate Matters (PM) in the environment well with the environmental variables. Furthermore, we adopted Deep Belief Network to extract features from the multimodel data feed which outperformed Convolutional Neural Network and achieved up to (a=80.8%, {\sigma}=0.001) accuracy.
Remote photo-plethysmography (rPPG) uses a camera to estimate a person's heart rate (HR). Similar to how heart rate can provide useful information about a person's vital signs, insights about the underlying physio/psychological conditions can be obtained from heart rate variability (HRV). HRV is a measure of the fine fluctuations in the intervals between heart beats. However, this measure requires temporally locating heart beats with a high degree of precision. We introduce a refined and efficient real-time rPPG pipeline with novel filtering and motion suppression that not only estimates heart rates, but also extracts the pulse waveform to time heart beats and measure heart rate variability. This unsupervised method requires no rPPG specific training and is able to operate in real-time. We also introduce a new multi-modal video dataset, VicarPPG 2, specifically designed to evaluate rPPG algorithms on HR and HRV estimation. We validate and study our method under various conditions on a comprehensive range of public and self-recorded datasets, showing state-of-the-art results and providing useful insights into some unique aspects. Lastly, we make available CleanerPPG, a collection of human-verified ground truth peak/heart-beat annotations for existing rPPG datasets. These verified annotations should make future evaluations and benchmarking of rPPG algorithms more accurate, standardized and fair.
We describe techniques and specifications of MATLAB software to process ambulatory electrocardiogram (ECG) data. Through template-based beat identification and simple pattern recognition models on the intervals between regular heart beats, we filter noisy sections of waveform and ectopic beats. Our end-to-end process can be used towards analysis of ECG and calculation of heart rate variability metrics after beat adjustments, removals and interpolation. Classification and noise detection is assessed on the human-annotated MIT-BIH Arrythmia and Noise Stress Test Databases.
Thermal comfort is an assessment of one's satisfaction with the surroundings; yet, most mechanisms that are used to provide thermal comfort are based on approaches that preclude physiological, psychological, and personal psychophysics that are precursors to thermal comfort. This leads to many people feeling either cold or hot in an environment that was supposed to be thermally comfortable to most users. To address this problem, this paper proposes to use heart rate variability (HRV) as an alternative indicator of thermal comfort status. Since HRV is linked to homeostasis, we conjectured that people's thermal comfort could be more accurately estimated based on their heart rate variability (HRV). To test our hypothesis, we analyzed statistical, spectral, and nonlinear HRV indices of 17 human subjects doing light office work in a cold, neutral, and hot environment. The resulting HRV indices were used as inputs to machine learning classification algorithms. We observed that HRV is distinctively different depending on the thermal environment and that it is possible to reliably predict each subject's thermal state (cold, neutral, and hot) with up to 93.7% accuracy. The result of this study suggests that it could be possible to design automatic real-time thermal comfort controllers based on people's HRV.
Respiratory sinus arrhythmia (RSA) is heart rate variability in synchrony with respiration although its functional significance not clear. The loss of sinus arrhythmia may indicate underlying heart failure or disease; therefore, there would be a great advantage of knowing how it works and affects the cardio-respiratory system, especially by providing a mathematical model. To this end, Windkessel model and cardiovascular partial differential equations are used to obtain cardiac output based on the elasticity of left ventricle, which is related to RSA. By solving the corresponding equations, it would be possible to propose a new model to predict the RSA effects on cardiac output.
Ischemic brain injuries are frequent and difficult to detect reliably or early. We present the multi-modal data set containing cardiovascular (blood pressure, blood flow, electrocardiogram) and brain electrical activities to derive electroencephalogram (EEG) biomarkers of corticothalamic communication under normal, sedation and hypoxic/ischemic conditions with ensuing recovery. We provide technical validation using EEGLAB. We also delineate the corresponding changes in the electrocardiogram (ECG)-derived heart rate variability (HRV) with the potential for future in-depth analyses of joint EEG-ECG dynamics. We review an open-source methodology to derive signatures of coupling between the ECoG and electrothalamogram (EThG) signals contained in the presented data set to better characterize the dynamics of thalamocortical communication during these clinically relevant states. The data set is presented in full band sampled at 2000 Hz, so the additional potential exists for insights from the full-band EEG and high-frequency oscillations under the bespoke experimental conditions. Future studies on the dataset may contribute to the development of new brain monitoring technologies, which will facilitate the prevention of neurological injuries.
Thermal comfort is a personal assessment of one's satisfaction with the surroundings. Yet, most thermal comfort delivery mechanisms preclude physiological and psychological precursors to thermal comfort. Accordingly, many people feel either cold or hot in an environment that is supposedly thermally comfortable to most people. To address this issue, this paper proposes to use people's heart rate variability (HRV) as an alternative indicator of thermal comfort. Since HRV is linked to homeostasis, we hypothesize that it could be used to predict people's thermal comfort status. To test our hypothesis, we analyzed statistical, spectral, and nonlinear HRV indices of 17 human subjects doing light office work in a cold, a neutral, and a hot environment. The resulting HRV indices were used as inputs to machine learning classification algorithms. We observed that HRV is distinctively altered depending on the thermal environment and that it is possible to steadfastly predict each subject's thermal environment (cold, neutral, and hot) with up to a 93.7% prediction accuracy. The result of this study implies that it could be possible to design automatic real-Time thermal comfort controllers based on people's HRV.
An electrocardiogram (ECG)-derived heart rate variability (HRV) index reliably tracks the inflammatory response induced by low-dose lipopolysaccharide (LPS) in near-term sheep fetuses. We evaluated the effect of vagus nerve stimulation (VNS) on vagus nerve electroneurogram (VENG) and the systemic inflammatory response induced by a high dose of LPS in neonatal piglets to mimic late-onset neonatal sepsis. We tested if our HRV inflammatory index tracks inflammation in piglets and its relationship to VENG. Following anesthesia, electrodes were attached to the left vagal nerve; ECG and blood pressure (BP) were recorded throughout the experiment. Following baseline, the piglets were administered LPS as 2mg/kg IV bolus. In the VNS treated piglet, the vagus nerve was stimulated for 10 minutes prior to and 10 min after the injection of LPS. In both groups, every 15 min post LPS, the arterial blood sample was drawn for blood gas, metabolites, and inflammatory cytokines. At the end of the experiment, the piglets were euthanized. BP and HRV measures were calculated. The piglets developed a potent inflammatory response to the LPS injection with TNF-alpha, IL-1beta, IL-6 and IL-8 peaking between 45 and 90 min post-injection. VNS diminished the LPS-induced systemic inflammatory response varying across the measured cytokines from two to ten-fold. The HRV index tracked accurately the temporal profile of cytokines and VENG changes. This novel model allows manipulating and tracking neonatal sepsis: The HRV inflammatory index 1) applies across species pre- and postnatally and 2) performs well at different degrees of sepsis (i.e., nanogram and milligram doses of LPS); 3) the present VNS paradigm effectively suppresses LPS-induced inflammation, even at high doses of LPS. The potential of early postnatal VNS to counteract sepsis and of HRV monitoring to early detect and track it deserve further study.
Objectives: This study examines human Photoplethysmogram (PPG) along with Electrocardiogram (ECG) signals to study cardiac autonomic imbalance in epileptic seizures. The significance and the prevalence of changes in PPG morphological parameters have been investigated to find common patterns among subjects. Alterations in cardiovascular parameters measured by PPG/ECG signals are used to train a neural network based on LSTM for automatic seizure detection. Methods: Electroencephalogram (EEG), ECG, and PPG signals from 12 different subjects ( 8 males;4 females;age 34.3$\pm$ 13.8) were recorded including 57 seizures and 101 hours of inter-ictal data. 12 PPG features significantly changing due to epileptic seizures were extracted and normalized based on a proposed z-score metric. 7 feature are heart rate variability related and 5 features hemodynamic related. Results: A consistent pattern of ictal change was observed for all the features across the subjects/seziures. The proposed seizure detector is subject independent and works for both nocturnal and diurnal seizures. With an average of 0.52 false alarms per hour, positive predictive value of $43\%$ and sensitivity of $92\%$, the new proposed hemodynamic based seizure detector shows improvement over the the heart rate variability based detector. Conclusion: The cardiac autonomic imbalance due to seizure manifests itself in variations of peripheral hemodynamics measured by PPG signal, suggesting vasoconstriction in limbs. These variations can be used on a consumer seizure detecting devices with optical sensors for seizure detection. Significance: The stereotyped pattern is common among all the subjects which can help understand the mechanism of cardiac autonomic imbalance induced by epileptic seizures.
Chaos, or exponential sensitivity to small perturbations, appears everywhere in nature. Moreover, chaos is predicted to play diverse functional roles in living systems. A method for detecting chaos from empirical measurements should therefore be a key component of the biologist's toolkit. But, classic chaos-detection tools are highly sensitive to measurement noise and break down for common edge cases, making it difficult to detect chaos in domains, like biology, where measurements are noisy. However, newer tools promise to overcome these limitations. Here, we combine several such tools into an automated processing pipeline, and show that our pipeline can detect the presence (or absence) of chaos in noisy recordings, even for difficult edge cases. As a first-pass application of our pipeline, we show that heart rate variability is not chaotic as some have proposed, and instead reflects a stochastic process in both health and disease. Our tool is easy-to-use and freely available.
Heart rate variability studies depend on the robust calculation of the tachogram, the heart rate times series, usually by the detection of R peaks in the electrocardiogram (ECG). ECGs however are subject to a number of sources of noise which are difficult to filter and therefore reduce the tachogram accuracy. We describe a pipeline for fast calculation of tachograms from noisy ECGs of several hours' length. The pipeline consists of three stages. A neural network (NN) trained to detect R peaks and distinguish these from noise; a measure to robustly detect false positives (FPs) and negatives (FNs) produced by the NN; a simple "alarm" algorithm for automatically removing FPs and interpolating FNs. In addition, we introduce the approach of encoding ECGs, tachograms and other cardiac time series in the form of raster images, which greatly speeds and eases their visual inspection and analysis.
Automated sleep stage classification using heart-rate variability is an active field of research. In this work limitations of the current state-of-the-art are addressed through the use of deep learning techniques and their efficacy is demonstrated. First, a temporal model is proposed for the inference of sleep stages from electrocardiography using a deep long- and short-term (LSTM) classifier and it is shown that this model outperforms previous approaches which were often limited to non-temporal or Markovian classifiers on a comprehensive benchmark data set (292 participants, 541214 samples) comprising a wide range of ages and pathological profiles, achieving a Cohen's $\kappa$ of $0.61\pm0.16$ and accuracy of $76.30\pm10.17$ annotated according to the Rechtschaffen & Kales annotation standard. Subsequently, it is demonstrated how knowledge learned on this large benchmark data set can be re-used through transfer learning for the classification of photoplethysmography (PPG) data. This is done using a smaller data set (60 participants, 91479 samples) that is annotated with the more recent American Association of Sleep Medicine annotation standard, achieving a Cohen's $\kappa$ of $0.63\pm0.13$ and accuracy of $74.65\pm8.63$ for wrist-mounted PPG-based sleep stage classification, higher than any previously reported performance using this sensor modality. This demonstrates the feasibility of knowledge transfer in sleep staging to adapt models for new sensor modalities as well as different annotation strategies.
Mindfulness has become increasingly popular as a method for building resilience against stress in both clinical and healthy populations. This study sought to investigate the effects of mindfulness training on perceived levels of stress and heart rate variability in students.
$\bullet$ Background : In Heart rate variability analysis, the rate-rate time series suffer often from aperiodic non-stationarity, presence of ectopic beats etc. It would be hard to extract helpful information from the original signals. 10 $\bullet$ Problem : Trend removal methods are commonly practiced to reduce the influence of the low frequency and aperiodic non-stationary in RR data. This can unfortunately affect the signal and make the analysis on detrended data less appropriate. $\bullet$ Objective : Investigate the detrending effect (linear \& nonlinear) in temporal / nonliear analysis of heart rate variability of long-term RR data (in normal sinus rhythm, atrial fibrillation, 15 congestive heart failure and ventricular premature arrhythmia conditions). $\bullet$ Methods : Temporal method : standard measure SDNN; Nonlinear methods : multi-scale Fractal Dimension (FD), Detrended Fluctuation Analysis (DFA) \& Sample Entropy (Sam-pEn) analysis. $\bullet$ Results : The linear detrending affects little the global characteristics of the RR data, either 20 in temporal analysis or in nonlinear complexity analysis. After linear detrending, the SDNNs are just slightly shifted and all distributions are well preserved. The cross-scale complexity remained almost the same as the ones for original RR data or correlated. Nonlinear detrending changed not only the SDNNs distribution, but also the order among different types of RR data. After this processing, the SDNN became indistinguishable be-25 tween SDNN for normal sinus rhythm and ventricular premature beats. Different RR data has different complexity signature. Nonlinear detrending made the all RR data to be similar , in terms of complexity. It is thus impossible to distinguish them. The FD showed that nonlinearly detrended RR data has a dimension close to 2, the exponent from DFA is close to zero and SampEn is larger than 1.5 -- these complexity values are very close to those for 30 random signal. $\bullet$ Conclusions : Pre-processing by linear detrending can be performed on RR data, which has little influence on the corresponding analysis. Nonlinear detrending could be harmful and it is not advisable to use this type of pre-processing. Exceptions do exist, but only combined with other appropriate techniques to avoid complete change of the signal's intrinsic dynamics. 35 Keywords $\bullet$ heart rate variability $\bullet$ linear / nonlinear detrending $\bullet$ complexity analysis $\bullet$ mul-tiscale analysis $\bullet$ detrended fluctuation analysis $\bullet$ fractal dimension $\bullet$ sample entropy;
Ischemic heart disease (IHD), particularly in its chronic stable form, is a subtle pathology due to its silent behavior before developing in unstable angina, myocardial infarction or sudden cardiac death. Machine learning techniques applied to parameters extracted form heart rate variability (HRV) signal seem to be a valuable support in the early diagnosis of some cardiac diseases. However, so far, IHD patients were identified using Artificial Neural Networks (ANNs) applied to a limited number of HRV parameters and only to very few subjects. In this study, we used several linear and non-linear HRV parameters applied to ANNs, in order to confirm these results on a large cohort of 965 sample of subjects and to identify which features could discriminate IHD patients with high accuracy. By using principal component analysis and stepwise regression, we reduced the original 17 parameters to five, used as inputs, for a series of ANNs. The highest accuracy of 82% was achieved using meanRR, LFn, SD1, gender and age parameters and two hidden neurons.
The diagnosis of heart diseases is a difficult task generally addressed by an appropriate examination of patients clinical data. Recently, the use of heart rate variability (HRV) analysis as well as of some machine learning algorithms, has proved to be a valuable support in the diagnosis process. However, till now, ischemic heart disease (IHD) has been diagnosed on the basis of Artificial Neural Networks (ANN) applied only to signs, symptoms and sequential ECG and coronary angiography, an invasive tool, while could be probably identified in a non-invasive way by using parameters extracted from HRV, a signal easily obtained from the ECG. In this study, 18 non-invasive features (age, gender, left ventricular ejection fraction and 15 obtained from HRV) of 243 subjects (156 normal subjects and 87 IHD patients) were used to train and validate a series of several ANN, different for number of input and hidden nodes. The best result was obtained using 7 input parameters and 7 hidden nodes with an accuracy of 98.9% and 82% for the training and validation dataset, respectively.
The 12-lead electrocardiogram (ECG) is a commonly used tool for detecting cardiac abnormalities such as atrial fibrillation, blocks, and irregular complexes. For the PhysioNet/CinC 2020 Challenge, we built an algorithm using gradient boosted tree ensembles fitted on morphology and signal processing features to classify ECG diagnosis. For each lead, we derive features from heart rate variability, PQRST template shape, and the full signal waveform. We join the features of all 12 leads to fit an ensemble of gradient boosting decision trees to predict probabilities of ECG instances belonging to each class. We train a phase one set of feature importance determining models to isolate the top 1,000 most important features to use in our phase two diagnosis prediction models. We use repeated random sub-sampling by splitting our dataset of 43,101 records into 100 independent runs of 85:15 training/validation splits for our internal evaluation results. Our methodology generates us an official phase validation set score of 0.476 and test set score of -0.080 under the team name, CVC, placing us 36 out of 41 in the rankings.
Remote physiological measurements, e.g., remote photoplethysmography (rPPG) based heart rate (HR), heart rate variability (HRV) and respiration frequency (RF) measuring, are playing more and more important roles under the application scenarios where contact measurement is inconvenient or impossible. Since the amplitude of the physiological signals is very small, they can be easily affected by head movements, lighting conditions, and sensor diversities. To address these challenges, we propose a cross-verified feature disentangling strategy to disentangle the physiological features with non-physiological representations, and then use the distilled physiological features for robust multi-task physiological measurements. We first transform the input face videos into a multi-scale spatial-temporal map (MSTmap), which can suppress the irrelevant background and noise features while retaining most of the temporal characteristics of the periodic physiological signals. Then we take pairwise MSTmaps as inputs to an autoencoder architecture with two encoders (one for physiological signals and the other for non-physiological information) and use a cross-verified scheme to obtain physiological features disentangled with the non-physiological features. The disentangled features are finally used for the joint prediction of multiple physiological signals like average HR values and rPPG signals. Comprehensive experiments on different large-scale public datasets of multiple physiological measurement tasks as well as the cross-database testing demonstrate the robustness of our approach.