A Novel Deep Learning based Neural Network for Heartbeat Detection in Ballistocardiograph

Ballistocardiography (BCG) is a revamped technology for cardiac function monitoring. Detecting individual heart beats in BCG remains a challenging task due to various artifacts and low signal-to-noise ratio, which are not well addressed by conventional approaches based on intuitive observations of BCG waveforms. In this paper, we propose to employ deep learning networks to capture the characteristics of the variations of BCG waveforms within and across individual subjects. Particularly, we design a neural network that combines Convolutional-Neural-Network (CNN) and Extreme Learning Machine (ELM). We test the new learning method on a real BCG data set and show better detection result compared with a state-of-the-art method. We demonstrate how the advanced machine learning technology can learn and detect BCG waveforms robustly.

Measurement, Prediction, and Control of Individual Heart Rate Responses to Exercise-Basics and Options for Wearable Devices

The use of wearable devices or “wearables” in the physical activity domain has been increasing in the last years. These devices are used as training tools providing the user with detailed information about individual physiological responses and feedback to the physical training process. Advantages in sensor technology, miniaturization, energy consumption and processing power increased the usability of these wearables. Furthermore, available sensor technologies must be reliable, valid, and usable. Considering the variety of the existing sensors not all of them are suitable to be integrated in wearables. The application and development of wearables has to consider the characteristics of the physical training process to improve the effectiveness and efficiency as training tools. During physical training, it is essential to elicit individual optimal strain to evoke the desired adjustments to training. One important goal is to neither overstrain nor under challenge the user. Many wearables use heart rate as indicator for this individual strain. However, due to a variety of internal and external influencing factors, heart rate kinetics are highly variable making it difficult to control the stress eliciting individually optimal strain. For optimal training control it is essential to model and predict individual responses and adapt the external stress if necessary. Basis for this modeling is the valid and reliable recording of these individual responses. Depending on the heart rate kinetics and the obtained physiological data, different models and techniques are available that can be used for strain or training control. Aim of this review is to give an overview of measurement, prediction, and control of individual heart rate responses. Therefore, available sensor technologies measuring the individual heart rate responses are analyzed and approaches to model and predict these individual responses discussed. Additionally, the feasibility for wearables is analyzed.

Comparison of the impedance cardiogram with continuous wave radar using body-contact antennas

This paper describes a continuous wave (CW) radar system with body-contact antennas and basic signal processing. The goal is to assess the signals' reproducibility across different subjects as well as a respiration cycle. Radar signals using body-contact antennas with a carrier frequency of 868 MHz are used to acquire the cardiac activity at the sternum. The radar I and Q channel signals are combined to form their magnitude. Signals are collected from six healthy males during paced breathing conditions. The electrocardiogram (ECG) and impedance cardiogram (ICG) signals are acquired simultaneously as reference. The chosen feature in the radar signal is the maximum of its second derivative, which is closest to the ICG B-point. The median and mean absolute errors in pre-ejection period (PEP) in milliseconds between the ICG's B-point and chosen feature in the radar signal range from -6-119.7 ms and 7.8-62.3 ms for all subjects. The results indicate that a reproducible radar signal is obtained from all six subjects. More work is needed on understanding the origin of the radar signals using ultrasound as a comparison.

Multiclass Classifier based Cardiovascular Condition Detection Using Smartphone Mechanocardiography

Cardiac translational and rotational vibrations induced by left ventricular motions are measurable using joint seismocardiography (SCG) and gyrocardiography (GCG) techniques. Multi-dimensional non-invasive monitoring of the heart reveals relative information of cardiac wall motion. A single inertial measurement unit (IMU) allows capturing cardiac vibrations in sufficient details and enables us to perform patient screening for various heart conditions. We envision smartphone mechanocardiography (MCG) for the use of e-health or telemonitoring, which uses a multi-class classifier to detect various types of cardiovascular diseases (CVD) using only smartphone's built-in internal sensors data. Such smartphone App/solution could be used by either a healthcare professional and/or the patient him/herself to take recordings from their heart. We suggest that smartphone could be used to separate heart conditions such as normal sinus rhythm (SR), atrial fibrillation (AFib), coronary artery disease (CAD), and possibly ST-segment elevated myocardial infarction (STEMI) in multiclass settings. An application could run the disease screening and immediately inform the user about the results. Widespread availability of IMUs within smartphones could enable the screening of patients globally in the future, however, we also discuss the possible challenges raised by the utilization of such self-monitoring systems.

Towards precise tracking of electric-mechanical cardiac time intervals through joint ECG and BCG sensing and signal processing

Automatic tracking of intra-beat cardiac activities in ballistocardiogram (BCG) is a highly interesting yet technically challenging topic for cardiac monitoring, due to the signal's high susceptibility to various forms of distortions. In this paper, we aim to further investigate the BCG waveform detection from a signal processing and analysis viewpoint. We collect synchronized electrocardiography(ECG) and BCG recordings from four healthy human subjects using an in-house built multi-physiological monitoring device. Particularly, we study post-exercise ECG-BCG signals that embed considerable variation in the heart beat during the post-exercise recovery phase. Furthermore, we develop an efficient and interactive tool for detecting and marking ECG-BCG waveforms in each heart beat. Through analyzing the detected time interval signals, we explore new interesting patterns of dynamic associations between different time interval signals. At the same time, we call for development of improved detection algorithms to address robustness and accuracy issues.

Non-Invasive Detection of Respiration and Heart Rate with a Vehicle Seat Sensor

This study demonstrates the feasibility of using a seat sensor designed for occupant classification from a production passenger vehicle to measure an occupant’s respiration rate (RR) and heart rate (HR) in a laboratory setting. Relaying occupant vital signs after a crash could improve emergency response by adding a direct measure of the occupant state to an Advanced Automatic Collision Notification (AACN) system. Data was collected from eleven participants with body weights ranging from 42 to 91 kg using a Ford Mustang passenger seat and seat sensor. Using a ballistocardiography (BCG) approach, the data was processed by time domain filtering and frequency domain analysis using the fast Fourier transform to yield RR and HR in a 1-min sliding window. Resting rates over the 30-min data collection and continuous RR and HR signals were compared to laboratory physiological instruments using the Bland-Altman approach. Differences between the seat sensor and reference sensor were within 5 breaths per minute for resting RR and within 15 beats per minute for resting HR. The time series comparisons for RR and HR were promising with the frequency analysis technique outperforming the peak detection technique. However, future work is necessary for more accurate and reliable real-time monitoring of RR and HR outside the laboratory setting.

Noncontact In-Bed Measurements of Physiological and Behavioral Signals Using an Integrated Fabric-Sheet Sensing Scheme

Home monitoring requires measuring the physiological and behavioral signals without impairing a subject's everyday life. This paper presents an integrated and noncontact approach for obtaining simultaneous physiological and behavioral signals of recumbent humans in beds using a home-monitoring application. In the proposed approach, a fabric-sheet unified sensing electrode (FUSE) obtains physiological signals by recording the electrocardiogram (ECG), chest and abdominal respiratory movements (RMs), and ballistocardiogram (BCG). The FUSE also detects the behavioral signals of body proximity (BPx) and lateral/supine lying postures. A prototype system with FUSE was validated in a short-term experiment and 6-h overnight measurements on two different groups composed of seven lying subjects. The results confirmed that the approach senses each signal independently and records the ECG, RMs, BCG, and BPx signals simultaneously. The mean sensitivities of the R and T waves of the ECG during sleep were 86.1% and 88.0%, respectively, whereas those of the chest and abdominal RMs were 90.7% and 90.1%, respectively. Although our prototype system has room for improvement, the results suggest that our approach enables the unconstrained, nocturnal monitoring of the physiological and behavioral signals in recumbent humans. The at-home monitoring of the physiological and behavioral signals is expected to contribute to cost-effective personalized healthcare in the future. This noncontact and easy-to-install system for in-bed measurements can facilitate a new era of home monitoring.

Heart Rate Estimated from Body Movements at Six Degrees of Freedom by Convolutional Neural Networks

Cardiac activity has been monitored continuously in daily life by virtue of advanced medical instruments with microelectromechanical system (MEMS) technology. Seismocardiography (SCG) has been considered to be free from the burden of measurement for cardiac activity, but it has been limited in its application in daily life. The most important issues regarding SCG are to overcome the limitations of motion artifacts due to the sensitivity of motion sensor. Although novel adaptive filters for noise cancellation have been developed, they depend on the researcher’s subjective decision. Convolutional neural networks (CNNs) can extract significant features from data automatically without a researcher’s subjective decision, so that signal processing has been recently replaced as CNNs. Thus, this study aimed to develop a novel method to enhance heart rate estimation from thoracic movement by CNNs. Thoracic movement was measured by six-axis accelerometer and gyroscope signals using a wearable sensor that can be worn by simply clipping on clothes. The dataset was collected from 30 participants (15 males, 15 females) using 12 measurement conditions according to two physical conditions (i.e., relaxed and aroused conditions), three body postures (i.e., sitting, standing, and supine), and six movement speeds (i.e., 3.2, 4.5, 5.8, 6.4, 8.5, and 10.3 km/h). The motion data (i.e., six-axis accelerometer and gyroscope) and heart rate (i.e., electrocardiogram (ECG)) were determined as the input data and labels in the dataset, respectively. The CNN model was developed based on VGG Net and optimized by testing according to network depth and data augmentation. The ensemble network of the VGG-16 without data augmentation and the VGG-19 with data augmentation was determined as optimal architecture for generalization. As a result, the proposed method showed higher accuracy than the previous SCG method using signal processing in most measurement conditions. The three main contributions are as follows: (1) the CNN model enhanced heart rate estimation with the benefits of automatic feature extraction from the data; (2) the proposed method was compared with the previous SCG method using signal processing; (3) the method was tested in 12 measurement conditions related to daily motion for a more practical application.

Multiple Instance Dictionary Learning for Beat-to-Beat Heart Rate Monitoring From Ballistocardiograms

A multiple instance dictionary learning approach, dictionary learning using functions of multiple instances (DL-FUMI), is used to perform beat-to-beat heart rate estimation and to characterize heartbeat signatures from ballistocardiogram (BCG) signals collected with a hydraulic bed sensor. DL-FUMI estimates a "heartbeat concept" that represents an individual's personal ballistocardiogram heartbeat pattern. DL-FUMI formulates heartbeat detection and heartbeat characterization as a multiple instance learning problem to address the uncertainty inherent in aligning BCG signals with ground truth during training. Experimental results show that the estimated heartbeat concept obtained by DL-FUMI is an effective heartbeat prototype and achieves superior performance over comparison algorithms.

Nonintrusive Remote Monitoring of Sleep in Home-Based Situation

Sleep deprivation can lead to loss of concentration, and risky decision-making. Nevertheless, some people may underestimate the importance of getting quality sleep. The standard health care systems might not be suitable for long-term monitoring of sleep. As an example, the polysomnography, i.e., the gold standard for assessing sleep disorders is cumbersome, expensive, and time-consuming. As a result, portable, nonintrusive and inexpensive systems for monitoring quality of sleep are greatly needed. This paper demonstrates a novel nonintrusive system for monitoring quality of sleep using an optical fiber embedded sensor mat. The proposed system is deployed in real-life conditions over a one-month period. Three senior female residents were enrolled for the study, where the sensor mat is placed under the bed mattress. Sleep quality is assessed based on several parameters, such as duration of sleep, sleep interruption, vital signs (heart rate and respiration). The proposed system shows an agreement with a user's survey collected before the study. Furthermore, the system is integrated within an existing ambient assisted living platform with a user-friendly interface to make it more convenient for the caregivers to follow-up the sleep parameters of the residents.

Toward Closed-Loop Transcutaneous Vagus Nerve Stimulation using Peripheral Cardiovascular Physiological Biomarkers: A Proof-of-Concept Study

Transcutaneous vagus nerve stimulation (t-VNS) is a promising technology for modulating brain function and possibly treating disorders of the central nervous system. While handheld devices are available for t-VNS, stimulation efficacy can only be quantified using expensive imaging or blood biomarker analyses. Additionally, the parameters and "dosage" recommendations for t-VNS are typically fixed, as there are limited biomarkers that can assess downstream effects of the stimulation outside of clinical settings. In this proof-of-concept study, we evaluated non-invasive peripheral cardiovascular measurements as physiological biomarkers of t-VNS efficacy. Specifically, we hypothesized two physiological biomarkers: (1) the pre-ejection period (PEP) of the heart - a parameter closely linked to sympathetic tone - and (2) the amplitude of peripheral photoplethysmogram (PPG) waveforms - representing changes in vasomotor tone and thus parasympathetic / sympathetic activation. A total of six healthy human subjects participated in the multi-day study, half each undergoing active or sham t-VNS stimulus. The three subjects receiving t-VNS had no decrease in PEP and an increase in PPG amplitude following t-VNS, while the subjects receiving sham stimulus had a decrease in PEP and no change in PPG amplitude. When combined with mental stress (a traumatic script being read back to the subjects), the group with t-VNS had no decrease in PEP and only a slight decrease in PPG amplitude following stimulus, while the group receiving sham stimulus had a decrease in PEP and also a slight decrease in PPG amplitude. These studies suggest that PEP and PPG amplitude measures may provide non-invasive physiological biomarkers of t-VNS efficacy, including in the presence of mental stress.

Cardiorespiratory system monitoring using a developed acoustic sensor

This Letter proposes a wireless acoustic sensor for monitoring heartbeat and respiration rate based on phonocardiogram (PCG). The developed sensor comprises a processor, a transceiver which operates at industrial, scientific and medical band and the frequency of 2.54 GHz as well as two capacitor microphones which one for recording the heartbeat and another one for respiration rate. To evaluate the precision of the presented sensor in estimating heartbeat and respiration rate, the sensor is tested on the different volunteers and the obtained results are compared with a gold standard as a reference. The results reveal that root-mean-square error are determined <2.27 beats/min and 0.92 breaths/min for the heartbeat and respiration rate in turn. While the standard deviation of the error is obtained <1.26 and 0.63 for heartbeat and respiration rate, respectively. Also, the sensor estimate sounds of [Formula: see text] to [Formula: see text] obtained PCG signal with sensitivity and specificity 98.1% and 98.3% in turn that make 3% improvement than previous works. The results prove that the sensor can be appropriate candidate for recognising abnormal condition in the cardiorespiratory system.

Wearable ballistocardiogram and seismocardiogram systems for health and performance

Cardiovascular diseases (CVDs) are prevalent in the US, and many forms of CVD primarily affect the mechanical aspects of heart function. Wearable technologies for monitoring the mechanical health of the heart and vasculature could enable proactive management of CVDs through titration of care based on physiological status as well as preventative wellness monitoring to help promote lifestyle choices that reduce the overall risk of developing CVDs. Additionally, such wearable technologies could be used to optimize human performance in austere environments. This review describes our progress in developing wearable ballistocardiogram (BCG)- and seismocardiogram-based systems for monitoring relative changes in cardiac output, contractility, and blood pressure. Our systems use miniature, low-noise accelerometers to measure the movements of the body in response to the heartbeat and novel machine learning algorithms to provide robustness against motion artifacts and sensor misplacement. Moreover, we have mathematically related wearable BCG signals-representing local, cardiogenic movements of a point on the body-to better understood whole body BCG signals, and thereby improved estimation of key health parameters. We validated these systems with experiments in healthy subjects, studies in patients with heart failure, and measurements in austere environments such as water immersion. The systems can be used in future work as a tool for clinicians and physiologists to measure the mechanical aspects of cardiovascular function outside of clinical settings, and to thereby titrate care for patients with CVDs, provide preventative screening, and optimize performance in austere environments by providing real-time in-depth information regarding performance and risk.

Ballistocardiogram-Based Approach to Cuffless Blood Pressure Monitoring: Proof of Concept and Potential Challenges

OBJECTIVE

The goal was to propose and establish the proof of concept of an ultraconvenient cuffless blood pressure monitoring approach based on the ballistocardiogram.

METHODS

The proposed approach monitors blood pressure by exploiting two features in the whole-body head-to-foot ballistocardiogram measured using a force plate: the time interval between the first ("I") and second ("J") major waves ("I-J interval") for diastolic pressure and the amplitude between the J and third major ("K") waves ("J-K amplitude") for pulse pressure. The efficacy of the approach was examined in 22 young healthy volunteers by investigating the diastolic pressure monitoring performance of pulse transit time, pulse arrival time, and ballistocardiogram's I-J interval, and the systolic pressure monitoring performance of pulse transit time and I-J interval in conjunction with ballistocardiogram's J-K amplitude.

RESULTS

The I-J interval was comparable to pulse transit time and pulse arrival time in monitoring diastolic pressure, and the J-K amplitude could provide meaningful improvement to pulse transit time and I-J interval in monitoring systolic pressure.

CONCLUSION

The ballistocardiogram may contribute toward ultraconvenient and more accurate cuffless blood pressure monitoring.

SIGNIFICANCE

The proposed approach has potential to complement the pulse transit time technique for cuffless blood pressure monitoring in two ways. First, it may be integrated with pulse transit time to enable independent monitoring of diastolic and systolic pressures via the J-K amplitude. Second, it may even enable diastolic and systolic pressure monitoring from the ballistocardiogram alone.

An Enhanced Method to Estimate Heart Rate from Seismocardiography via Ensemble Averaging of Body Movements at Six Degrees of Freedom

Continuous cardiac monitoring has been developed to evaluate cardiac activity outside of clinical environments due to the advancement of novel instruments. Seismocardiography (SCG) is one of the vital components that could develop such a monitoring system. Although SCG has been presented with a lower accuracy, this novel cardiac indicator has been steadily proposed over traditional methods such as electrocardiography (ECG). Thus, it is necessary to develop an enhanced method by combining the significant cardiac indicators. In this study, the six-axis signals of accelerometer and gyroscope were measured and integrated by the L2 normalization and multi-dimensional kineticardiography (MKCG) approaches, respectively. The waveforms of accelerometer and gyroscope were standardized and combined via ensemble averaging, and the heart rate was calculated from the dominant frequency. Thirty participants (15 females) were asked to stand or sit in relaxed and aroused conditions. Their SCG was measured during the task. As a result, proposed method showed higher accuracy than traditional SCG methods in all measurement conditions. The three main contributions are as follows: (1) the ensemble averaging enhanced heart rate estimation with the benefits of the six-axis signals; (2) the proposed method was compared with the previous SCG method that employs fewer-axis; and (3) the method was tested in various measurement conditions for a more practical application.

An algorithm for the beat-to-beat assessment of cardiac mechanics during sleep on Earth and in microgravity from the seismocardiogram

Seismocardiogram, SCG, is the measure of precordial vibrations produced by the beating heart, from which cardiac mechanics may be explored on a beat-to-beat basis. We recently collected a large amount of SCG data (>69 recording hours) from an astronaut to investigate cardiac mechanics during sleep aboard the International Space Station and on Earth. SCG sleep recordings are characterized by a prolonged duration and wide heart rate swings, thus a specific algorithm was developed for their analysis. In this article we describe the new algorithm and its performance. The algorithm is composed of three parts: 1) artifacts removal, 2) identification in each SCG waveform of four fiducial points associated with the opening and closure of the aortic and mitral valves, 3) beat-to-beat computation of indexes of cardiac mechanics from the SCG fiducial points. The algorithm was tested on two sleep recordings and yielded the identification of the fiducial points in more than 36,000 beats with a precision, quantified by the Positive Predictive Value, ≥99.2%. These positive findings provide the first evidence that cardiac mechanics may be explored by the automatic analysis of SCG long-lasting recordings, taken out of the laboratory setting, and in presence of significant heart rate modulations.

Lab-on-Skin: A Review of Flexible and Stretchable Electronics for Wearable Health Monitoring

Skin is the largest organ of the human body, and it offers a diagnostic interface rich with vital biological signals from the inner organs, blood vessels, muscles, and dermis/epidermis. Soft, flexible, and stretchable electronic devices provide a novel platform to interface with soft tissues for robotic feedback and control, regenerative medicine, and continuous health monitoring. Here, we introduce the term "lab-on-skin" to describe a set of electronic devices that have physical properties, such as thickness, thermal mass, elastic modulus, and water-vapor permeability, which resemble those of the skin. These devices can conformally laminate on the epidermis to mitigate motion artifacts and mismatches in mechanical properties created by conventional, rigid electronics while simultaneously providing accurate, non-invasive, long-term, and continuous health monitoring. Recent progress in the design and fabrication of soft sensors with more advanced capabilities and enhanced reliability suggest an impending translation of these devices from the research lab to clinical environments. Regarding these advances, the first part of this manuscript reviews materials, design strategies, and powering systems used in soft electronics. Next, the paper provides an overview of applications of these devices in cardiology, dermatology, electrophysiology, and sweat diagnostics, with an emphasis on how these systems may replace conventional clinical tools. The review concludes with an outlook on current challenges and opportunities for future research directions in wearable health monitoring.

Combined Seismo- and Gyro-Cardiography: A More Comprehensive Evaluation of Heart-Induced Chest Vibrations

This paper reports on the combined analysis of seismocardiogram (SCG) and gyrocardiogram (GCG) recordings. An inertial measurement unit (IMU) consisting of a three-axis micro-electromechanical (MEMS) accelerometer and a three-axis MEMS gyroscope is used to record heart-induced mechanical vibrations from the chest wall of the subjects. An electrocardiogram and an impedance cardiogram (ICG) sensor are also used as references for segmenting the cardiac cycles and recording the aortic valve opening and closure (AO and AC) events, respectively. A simplified model is proposed to explain the mechanical coupling of the chest wall to the IMU. Correlations and time differences are analyzed for the annotation of GCG and its first derivative with respect to ICG and SCG as references. Experimental results indicate a precise identification of systolic points such as the AO and AC events. The left ventricular ejection time and pre-ejection period metrics calculated from gyroscope recordings are also shown to accurately track their corresponding trends acquired from ICG signals. Waveform similarity analyses indicate that the first derivative of GCG has a better similarity with SCG than the GCG signal itself. Experimental results also suggest that interdevice differences in GCG recordings would need to be addressed before this technology can gain widespread application.

Unobtrusive Nocturnal Heartbeat Monitoring by a Ballistocardiographic Sensor in Patients with Sleep Disordered Breathing

Sleep disordered breathing (SDB) is known for fluctuating heart rates and an increased risk of developing arrhythmias. The current reference for heartbeat analysis is an electrocardiogram (ECG). As an unobtrusive alternative, we tested a sensor foil for mechanical vibrations to perform a ballistocardiography (BCG) and applied a novel algorithm for beat-to-beat cycle length detection. The aim of this study was to assess the correlation between beat-to-beat cycle length detection by the BCG algorithm and simultaneously recorded ECG. In 21 patients suspected for SDB undergoing polysomnography, we compared ECG to simultaneously recorded BCG data analysed by our algorithm. We analysed 362.040 heartbeats during a total of 93 hours of recording. The baseline beat-to-beat cycle length correlation between BCG and ECG was r_s = 0.77 (n = 362040) with a mean absolute difference of 15 ± 162 ms (mean cycle length: ECG 923 ± 220 ms; BCG 908 ± 203 ms). After filtering artefacts and improving signal quality by our algorithm, the correlation increased to r_s = 0.95 (n = 235367) with a mean absolute difference in cycle length of 4 ± 72 ms (ECG 920 ± 196 ms; BCG 916 ± 194 ms). We conclude that our algorithm, coupled with a BCG sensor foil provides good correlation of beat-to-beat cycle length detection with simultaneously recorded ECG.

A novel dual gating approach using joint inertial sensors: implications for cardiac PET imaging

Positron emission tomography (PET) is a non-invasive imaging technique which may be considered as the state of art for the examination of cardiac inflammation due to atherosclerosis. A fundamental limitation of PET is that cardiac and respiratory motions reduce the quality of the achieved images. Current approaches for motion compensation involve gating the PET data based on the timing of quiescent periods of cardiac and respiratory cycles. In this study, we present a novel gating method called microelectromechanical (MEMS) dual gating which relies on joint non-electrical sensors, i.e. tri-axial accelerometer and gyroscope. This approach can be used for optimized selection of quiescent phases of cardiac and respiratory cycles. Cardiomechanical activity according to echocardiography observations was investigated to confirm whether this dual sensor solution can provide accurate trigger timings for cardiac gating. Additionally, longitudinal chest motions originating from breathing were measured by accelerometric- and gyroscopic-derived respiratory (ADR and GDR) tracking. The ADR and GDR signals were evaluated against Varian real-time position management (RPM) signals in terms of amplitude and phase. Accordingly, high linear correlation and agreement were achieved between the reference electrocardiography, RPM, and measured MEMS signals. We also performed a Ge-68 phantom study to evaluate possible metal artifacts caused by the integrated read-out electronics including mechanical sensors and semiconductors. The reconstructed phantom images did not reveal any image artifacts. Thus, it was concluded that MEMS-driven dual gating can be used in PET studies without an effect on the quantitative or visual accuracy of the PET images. Finally, the applicability of MEMS dual gating for cardiac PET imaging was investigated with two atherosclerosis patients. Dual gated PET images were successfully reconstructed using only MEMS signals and both qualitative and quantitative assessments revealed encouraging results that warrant further investigation of this method.

SeismoWatch: Wearable Cuffless Blood Pressure Monitoring Using Pulse Transit Time

The current norm for measuring blood pressure (BP) at home is using an automated BP cuff based on oscillometry. Despite providing a viable and familiar method of tracking BP at home, oscillometric devices can be both cumbersome and inaccurate with the inconvenience of the hardware typically limiting measurements to once or twice per day. To address these limitations, a wrist-watch BP monitor was developed to measure BP through a simple maneuver: holding the watch against the sternum to detect micro-vibrations of the chest wall associated with the heartbeat. As a pulse wave propagates from the heart to the wrist, an accelerometer and optical sensor on the watch measure the travel time - pulse transit time (PTT) - to estimate BP. In this paper, we conducted a study to test the accuracy and repeatability of our device. After calibration, the diastolic pressure estimations reached a root-mean-square error of 2.9 mmHg. The watch-based system significantly outperformed (p<0.05) conventional pulse arrival time (PAT) based wearable blood pressure estimations - the most commonly used method for wearable BP sensing in the existing literature and commercial devices. Our device can be a convenient means for wearable BP monitoring outside of clinical settings in both health-conscious and hypertensive populations.1.

Heart beat characterization from ballistocardiogram signals using extended functions of multiple instances

A multiple instance learning (MIL) method, extended Function of Multiple Instances (eFUMI), is applied to ballistocardiogram (BCG) signals produced by a hydraulic bed sensor. The goal of this approach is to learn a personalized heartbeat "concept" for an individual. This heartbeat concept is a prototype (or "signature") that characterizes the heartbeat pattern for an individual in ballistocardiogram data. The eFUMI method models the problem of learning a heartbeat concept from a BCG signal as a MIL problem. This approach elegantly addresses the uncertainty inherent in a BCG signal (e. g., misalignment between training data and ground truth, mis-collection of heartbeat by some transducers, etc.). Given a BCG training signal coupled with a ground truth signal (e.g., a pulse finger sensor), training "bags" labeled with only binary labels denoting if a training bag contains a heartbeat signal or not can be generated. Then, using these bags, eFUMI learns a personalized concept of heartbeat for a subject as well as several non-heartbeat background concepts. After learning the heartbeat concept, heartbeat detection and heart rate estimation can be applied to test data. Experimental results show that the estimated heartbeat concept found by eFUMI is more representative and a more discriminative prototype of the heartbeat signals than those found by comparison MIL methods in the literature.

Robust Sensing of Distal Pulse Waveforms on a Modified Weighing Scale for Ubiquitous Pulse Transit Time Measurement

The measurement of aortic pulse transit time (PTT), the time for the arterial pulse wave to travel from the carotid to the femoral artery, can provide valuable insight into cardiovascular health, specifically regarding arterial stiffness and blood pressure (BP). To measure aortic PTT, both proximal and distal arterial pulse timings are required. Recently, our group has demonstrated that the ballistocardiogram signal measured on a modified weighing scale can provide an unobtrusive, yet accurate, means of obtaining a proximal timing reference; however, there are no convenient, reliable methods to extract the distal timing from a subject standing on the modified weighing scale. It is common to use a photoplethysmograph (PPG) attached to a toe to measure this distal pulse, but we discovered that this signal is greatly deteriorated as the subject stands on the scale. In this paper, we propose a novel method to measure the distal pulse using a custom reflective PPG array attached to the dorsum side of the foot (D-PPG). A total of 12 subjects of varying skin tones were recruited to assess the preliminary validation of this approach. Pulse measurements using the D-PPG were taken from seated and standing subjects, and the commercially available PPG were measured for facilitating comparison of timing measurements. We show that the D-PPG was the only sensor to retain the high detection rate of feasible timing values. To further test and optimize the system, various factors such as applied pressure, measurement location, and LED/photodiode configuration were tested.

Gyrocardiography: A New Non-invasive Monitoring Method for the Assessment of Cardiac Mechanics and the Estimation of Hemodynamic Variables

Gyrocardiography (GCG) is a new non-invasive technique for assessing heart motions by using a sensor of angular motion – gyroscope – attached to the skin of the chest. In this study, we conducted simultaneous recordings of electrocardiography (ECG), GCG, and echocardiography in a group of subjects consisting of nine healthy volunteer men. Annotation of underlying fiducial points in GCG is presented and compared to opening and closing points of heart valves measured by a pulse wave Doppler. Comparison between GCG and synchronized tissue Doppler imaging (TDI) data shows that the GCG signal is also capable of providing temporal information on the systolic and early diastolic peak velocities of the myocardium. Furthermore, time intervals from the ECG Q-wave to the maximum of the integrated GCG (angular displacement) signal and maximal myocardial strain curves obtained by 3D speckle tracking are correlated. We see GCG as a promising mechanical cardiac monitoring tool that enables quantification of beat-by-beat dynamics of systolic time intervals (STI) related to hemodynamic variables and myocardial contractility.

Systolic Time Interval Estimation Using Continuous Wave Radar With On-Body Antennas

The estimation of systolic time intervals (STIs) is done using continuous wave (CW) radar at 2.45 GHz with an on-body antenna.

MOTIVATION

In the state of the art, typically bioimpedance, heart sounds and/or ultrasound are used to measure STIs. All three methods suffer from insufficient accuracy of STI estimation due to various reasons. CW radar is investigated for its ability to overcome the deficiencies in the state of the art.

METHODS

Ten healthy male subjects aged 25-45 were asked to lie down at a 30 incline. Recordings of 60 s were taken without breathing and with paced breathing. Heart sounds, electrocardiogram, respiration, and impedance cardiogram were measured simultaneously as reference. The radar antennas were placed at two positions on the chest. The antennas were placed directly on the body as well as with cotton textile in between. The beat to beat STIs have been determined from the reference signals as well as CW radar signals.

RESULTS

The results indicate that CW radar can be used to estimate STIs in ambulatory monitoring.

SIGNIFICANCE

The results pave way to a potentially more compact method of estimating STIs, which can be integrated into a wearable device.

Evaluation of respiratory- and postural-induced changes on the ballistocardiogram signal by time warping averaging

OBJECTIVE

The aim of this work was to evaluate the potential changes in the ballistocardiogram (BCG) signal induced by different respiratory patterns and posture, by using the dynamic time warping (DTW) technique.

APPROACH

BCG signals were recorded in a group of 20 healthy volunteers, simultaneously with an electrocardiogram (ECG). Two recordings, one in a supine (SUP) and one in a standing (ST) position, including spontaneous breathing, two 1 min apneas (at full and empty-lungs, respectively) and 30 s of Valsalva, were analyzed. A warped averaged waveform was computed for each phase, from which amplitude and temporal parameters were extracted to characterize each condition.

MAIN RESULTS

Variations were found in both amplitude and duration of BCG-derived parameters among manoeuvres, especially when compared to spontaneous breathing, suggesting a complex interaction between intra-thoracic pressure changes acting on venous return, together with the autonomic nervous system modulation on heart rate. The effect of a hydrostatic pressure gradient elicited by postural conditions was also evident.

SIGNIFICANCE

Posture and respiratory manoeuvres affect the BCG signal in different ways, probably as a result of changes induced in preload and afterload. This supports the need to define separate normality ranges for each posture and/or breathing conditions, as well as the importance of applying specific manoeuvres to highlight any pathological response in the computed BCG parameters.

Bed-based instrumentation for unobtrusive sleep quality assessment in severely disabled autistic children

The relationship between sleep quality and daytime wellness and performance in severely disabled, autistic children is not well understood. While polysomnography and, more recently, actigraphy serve as means to obtain sleep assessment data from neurotypical children and adults, these techniques are not well-suited to severely autistic children. This paper presents recent progress on a bed sensor suite that can unobtrusively track physiological and behavioral parameters used to assess sleep quality. Electromechanical films and load cells provide data that yield heart rate, respiration rate, center of position, in-and-out-of-bed activity, and general movement, while thermocouples are used to detect bed-wetting events.

Automatic Detection of Seismocardiogram Sensor Misplacement for Robust Pre-Ejection Period Estimation in Unsupervised Settings

Seismocardiography (SCG), the measurement of the local chest vibrations due to the movements of blood and the heart, is a non-invasive technique for assessing myocardial contractility via the pre-ejection period (PEP). Recently, SCG-based extraction of PEP has been shown to be an effective means of classifying decompensated from compensated heart failure patients, and thus can be potentially used for monitoring such patients at home. Accurate extraction of PEP from SCG signals hinges on lab-based population data (i.e., regression curves) linking particular time-domain features of the SCG signal to corresponding features from reference standard bulky instruments such as impedance cardiography (ICG). Such regression curves, in the case of SCG, have always been estimated based on the "ideal" positioning of the SCG sensor on the chest. However, in settings such as the home where users may position the SCG measurement hardware on the chest without supervision, it is likely that the sensor will not always be placed exactly on this "ideal" location on the sternum, but rather on other positions on the chest as well. In this study, we show for the first time that the regression curve for estimating PEP from SCG signals differs significantly as the position of the sensor changes. We further devise a method to automatically detect when the sensor is placed in any position other than the desired one in order to avoid inaccurate systolic time interval estimation. Our classification algorithm for this purpose resulted in 0.83 precision and 0.82 recall when classifying whether the sensor is placed in the desired position or not. The classifier was tested with heartbeats taken both at rest, and also during exercise recovery to ensure that waveform changes due to positioning could be accurately discriminated from those due to physiological effects.

Time-Frequency Distribution of Seismocardiographic Signals: A Comparative Study

Accurate estimation of seismocardiographic (SCG) signal features can help successful signal characterization and classification in health and disease. This may lead to new methods for diagnosing and monitoring heart function. Time-frequency distributions (TFD) were often used to estimate the spectrotemporal signal features. In this study, the performance of different TFDs (e.g., short-time Fourier transform (STFT), polynomial chirplet transform (PCT), and continuous wavelet transform (CWT) with different mother functions) was assessed using simulated signals, and then utilized to analyze actual SCGs. The instantaneous frequency (IF) was determined from TFD and the error in estimating IF was calculated for simulated signals. Results suggested that the lowest IF error depended on the TFD and the test signal. STFT had lower error than CWT methods for most test signals. For a simulated SCG, Morlet CWT more accurately estimated IF than other CWTs, but Morlet did not provide noticeable advantages over STFT or PCT. PCT had the most consistently accurate IF estimations and appeared more suited for estimating IF of actual SCG signals. PCT analysis showed that actual SCGs from eight healthy subjects had multiple spectral peaks at 9.20 ± 0.48, 25.84 ± 0.77, 50.71 ± 1.83 Hz (mean ± SEM). These may prove useful features for SCG characterization and classification.

A Cardiac Early Warning System with Multi Channel SCG and ECG Monitoring for Mobile Health

Use of information and communication technology such as smart phone, smart watch, smart glass and portable health monitoring devices for healthcare services has made Mobile Health (mHealth) an emerging research area. Coronary Heart Disease (CHD) is considered as a leading cause of death world wide and an increasing number of people die prematurely due to CHD. Under such circumstances, there is a growing demand for a reliable cardiac monitoring system to catch the intermittent abnormalities and detect critical cardiac behaviors which lead to sudden death. Use of mobile devices to collect Electrocardiography (ECG), Seismocardiography (SCG) data and efficient analysis of those data can monitor a patient’s cardiac activities for early warning. This paper presents a novel cardiac data acquisition method and combined analysis of Electrocardiography (ECG) and multi channel Seismocardiography (SCG) data. An early warning system is implemented to monitor the cardiac activities of a person and accuracy assessment of the early warning system is conducted for the ECG data only. The assessment shows 88% accuracy and effectiveness of our proposed analysis, which implies the viability and applicability of the proposed early warning system.

Assessment of pre-ejection period in ambulatory subjects using seismocardiogram in a wearable blood pressure monitor

The ability to monitor arterial blood pressure continuously with unobtrusive body worn sensors may provide a unique and potentially valuable assessment of a patient's cardiovascular health. Pulse wave velocity (PWV) offers an attractive method to continuously monitoring blood pressure. However, PWV technologies based on timing measurements between the ECG and a distal PPG suffer from inaccuracies on mobile patients due to the confounding influence of pre-ejection period (PEP). In this paper, we presented a wearable, continuous blood pressure monitor (ViSi Mobile) that can measure and track changes in PEP. PEP is determined from precordial vibrations captured by an accelerometer coupled to the patient's sternum. The performance of the PEP measurements was evaluated on test subjects with postural change and patient activity. Results showed potential to improve cNIBP accuracy in active patients.

Sensor data quality processing for vital signs with opportunistic ambient sensing

Opportunistic ambient sensing involves placement of sensors appropriately so that intermittent contact can be made unobtrusively for gathering physiological signals for vital signs. In this paper, we discuss the results of our quality processing system used to extract heart rate from ballisto-cardiogram signals obtained from a micro-bending fiber optic sensor pressure mat. Visual inspection is used to label data into informative and non-informative classes based on their heart rate information. Five classifiers are employed for the classification process, i.e., random forest, support vector machine, multilayer, feedforward neural network, linear discriminant analysis, and decision tree. To compute the overall effectiveness of quality processing, the informative signals are processed to estimate interbeat intervals. The system was used to process, data collected from 50 human subjects sitting in a massage chair while performing different activities. Opportunistically collected data was obtained from the fiber optic sensor mat placed on the headrest of the massage chair. Using our classification approach, 57.37% of the dataset was able to provide informative signals. On the informative signals, random forest classifier achieves the best classification accuracy with a mean accuracy of 98.99%. The average of the mean absolute error between the estimated heart rate and the reference ECG is reduced from 13.2 to 8.47. Therefore, the proposed system shows a good robustness for opportunistic ambient sensing.

An adaptive approach for heartbeat detection based on S-transform in seismocardiograms

This study presents a new technique which allows identification of individual heartbeats from seismocardiograms (SCG) with high accuracy. Our method is electrocardiogram (ECG) independent and designed based upon S-transform and Shannon energy. The S-transform which is a time-frequency (TF) representation first provides frequency-dependent resolution while preserving a direct relationship with Fourier spectrum. Subsequently, individual heartbeats are detected in the time domain by calculating the Shannon energy (SSE) of each obtained local spectrum and employing other techniques such as successive mean quantization transform (SMQT) and adaptive thresholding. A total of 30 recordings were analysed in this study by measuring SCG and simultaneous electrocardiogram (ECG) in supine position. The performance of the algorithm was tested using the standard ECGs obtained from each test subject. The obtained results were as follows (sensitivity, precision, and detection error rate): (98.0%, 98.4% and 0.2%). In conclusion, the results confirmed that combination of S-transform, Shannon energy, and other techniques considerably enhanced the efficiency for the heartbeat detection in seismocardiograms.

Using multiple placements of accelerometers to measure cardiovascular pulse transit times

To measure pulse transit times (PTT) and calculate pulse wave velocities (PWV), tri-axial and uniaxial accelerometers were placed in groups of 2 to 4 over the manubrium, xiphoid process, forehead, wrist and ankle, and/or over the carotid, femoral, and posterior auricular arteries in 11 consented supine subjects. Signals were sampled at 1 kHz and filtered. Radial vectors were calculated from the tri-axial measurements. A 3-lead ECG was simultaneously collected over the same 180 s window, as well as respiratory rate. Ensemble averages (with ±S.D.) and raster plots were generated for each filtered time series from 200 ms before to 800 ms after the peak of each ECG R-wave. Lag times between the R-wave peak (taken as t=0) and one or more prominent peaks (or valleys if appropriate) of the various accelerometer signals were calculated, by using the signal from the axis (or the radial vector) with the best signal to noise ratio. PWV was calculated from the regression of the distance measured versus the PTT between pairs, especially of the clinically-relevant carotid-femoral PTT. A spectral analysis of each ensemble was taken, with the hypothesis that in young adults the measures at the periphery would have less energy at higher frequencies than those of an older adult because of age-related changes in arterial distensibility.

Wearable Sensing of Cardiac Timing Intervals from Cardiogenic Limb Vibration Signals

In this paper we describe a new method to measure aortic valve opening (AVO) and closing (AVC) from cardiogenic limb vibrations (i.e., wearable ballistocardiogram [BCG] signals). AVO and AVC were detected for each heartbeat with accelerometers on the upper arm (A), wrist (W), and knee (K) of 22 subjects following isometric exercise. Exercise-induced changes were recorded with impedance cardiography. The method, Filter BCG, detects peaks in distal vibrations after filtering with individually-tuned bandpass filters. In agreement with recent studies, we did not find peaks at AVO and AVC in limb vibrations directly. Interestingly, distal vibrations filtered with FilterBCG yielded reliable peaks at AVO (r² = 0.95 A, 0.94 W, 0.77 K) and AVC (r²= 0.92 A, 0.89 W, 0.68 K). FilterBCG measures AVO and AVC accurately from arm, wrist, and knee vibrations, and it outperforms the standard R-J interval method.

REM sleep estimation based on autonomic dynamics using R-R intervals

OBJECTIVE

We developed an automatic algorithm to determine rapid eye movement (REM) sleep on the basis of the autonomic activities reflected in heart rate variations.

APPROACH

The heart rate variability (HRV) parameters were calculated using the R-R intervals from an electrocardiogram (ECG). A major autonomic variation associated with the sleep cycle was extracted from a combination of the obtained parameters. REM sleep was determined with an adaptive threshold applied to the acquired feature. The algorithm was optimized with the data from 26 healthy subjects and obstructive sleep apnea (OSA) patients and was validated with data from a separate group of 25 healthy and OSA subjects.

MAIN RESULTS

According to an epoch-by-epoch (30 s) analysis, the average of Cohen's kappa and the accuracy were respectively 0.63 and 87% for the training set and 0.61 and 87% for the validation set. In addition, the REM sleep-related information extracted from the results of the proposed method revealed a significant correlation with those from polysomnography (PSG).

SIGNIFICANCE

The current algorithm only using R-R intervals can be applied to mobile and wearable devices that acquire heart-rate-related signals; therefore, it is appropriate for sleep monitoring in the home and ambulatory environments. Further, long-term sleep monitoring could provide useful information to clinicians and patients for the diagnosis and treatments of sleep-related disorders and individual health care.

Fiber bragg grating sensor based device for simultaneous measurement of respiratory and cardiac activities

This paper reports a novel optical ballistocardiography technique, which is non-invasive, for the simultaneous measurement of cardiac and respiratory activities using a Fiber Bragg Grating Heart Beat Device (FBGHBD). The unique design of FBGHBD offers additional capabilities such as monitoring nascent morphology of cardiac and breathing activity, heart rate variability, heart beat rhythm, etc., which can assist in early clinical diagnosis of many conditions associated with heart and lung malfunctioning. The results obtained from the FBGHBD positioned around the pulmonic area on the chest have been evaluated against an electronic stethoscope which detects and records sound pulses originated from the cardiac activity. In order to evaluate the performance of the FBGHBD, quantitative and qualitative studies have been carried out and the results are found to be reliable and accurate, validating its potential as a standalone medical diagnostic device. The developed FBGHBD is simple in design, robust, portable, EMI proof, shock proof and non-electric in its operation which are desired features for any clinical diagnostic tool used in hospital environment.

Automatic and Robust Delineation of the Fiducial Points of the Seismocardiogram Signal for Non-invasive Estimation of Cardiac Time Intervals

OBJECTIVE

The purpose of this research was to design a delineation algorithm that could detect specific fiducial points of the seismocardiogram (SCG) signal with or without using the electrocardiogram (ECG) R-wave as the reference point. The detected fiducial points were used to estimate cardiac time intervals. Due to complexity and sensitivity of the SCG signal, the algorithm was designed to robustly discard the low-quality cardiac cycles, which are the ones that contain unrecognizable fiducial points.

METHOD

The algorithm was trained on a dataset containing 48,318 manually annotated cardiac cycles. It was then applied to three test datasets: 65 young healthy individuals (dataset 1), 15 individuals above 44 years old (dataset 2), and 25 patients with previous heart conditions (dataset 3).

RESULTS

The algorithm accomplished high prediction accuracy with the rootmean- square-error of less than 5 ms for all the test datasets. The algorithm overall mean detection rate per individual recordings (DRI) were 74, 68, and 42 percent for the three test datasets when concurrent ECG and SCG were used. For the standalone SCG case, the mean DRI was 32, 14 and 21 percent.

CONCLUSION

When the proposed algorithm applied to concurrent ECG and SCG signals, the desired fiducial points of the SCG signal were successfully estimated with a high detection rate. For the standalone case, however, the algorithm achieved high prediction accuracy and detection rate for only the young individual dataset.

SIGNIFICANCE

The presented algorithm could be used for accurate and non-invasive estimation of cardiac time intervals.

A Hidden Markov Model for Seismocardiography

We propose a hidden Markov model approach for processing seismocardiograms. The seismocardiogram morphology is learned using the expectation-maximization algorithm, and the state of the heart at a given time instant is estimated by the Viterbi algorithm. From the obtained Viterbi sequence, it is then straightforward to estimate instantaneous heart rate, heart rate variability measures, and cardiac time intervals (the latter requiring a small number of manual annotations). As is shown in the conducted experimental study, the presented algorithm outperforms the state-of-the-art in seismocardiogram-based heart rate and heart rate variability estimation. Moreover, the isovolumic contraction time and the left ventricular ejection time are estimated with mean absolute errors of about 5 [ms] and [Formula: see text], respectively. The proposed algorithm can be applied to any set of inertial sensors; does not require access to any additional sensor modalities; does not make any assumptions on the seismocardiogram morphology; and explicitly models sensor noise and beat-to-beat variations (both in amplitude and temporal scaling) in the seismocardiogram morphology. As such, it is well suited for low-cost implementations using off-the-shelf inertial sensors and targeting, e.g., at-home medical services.

Weighing Scale-Based Pulse Transit Time is a Superior Marker of Blood Pressure than Conventional Pulse Arrival Time

Pulse transit time (PTT) is being widely pursued for cuff-less blood pressure (BP) monitoring. Most efforts have employed the time delay between ECG and finger photoplethysmography (PPG) waveforms as a convenient surrogate of PTT. However, these conventional pulse arrival time (PAT) measurements include the pre-ejection period (PEP) and the time delay through small, muscular arteries and may thus be an unreliable marker of BP. We assessed a bathroom weighing scale-like system for convenient measurement of ballistocardiography and foot PPG waveforms – and thus PTT through larger, more elastic arteries – in terms of its ability to improve tracking of BP in individual subjects. We measured “scale PTT”, conventional PAT, and cuff BP in humans during interventions that increased BP but changed PEP and smooth muscle contraction differently. Scale PTT tracked the diastolic BP changes well, with correlation coefficient of −0.80 ± 0.02 (mean ± SE) and root-mean-squared-error of 7.6 ± 0.5 mmHg after a best-case calibration. Conventional PAT was significantly inferior in tracking these changes, with correlation coefficient of −0.60 ± 0.04 and root-mean-squared-error of 14.6 ± 1.5 mmHg (p < 0.05). Scale PTT also tracked the systolic BP changes better than conventional PAT but not to an acceptable level. With further development, scale PTT may permit reliable, convenient measurement of BP.

Automatic Identification of Systolic Time Intervals in Seismocardiogram

Continuous and non-invasive monitoring of hemodynamic parameters through unobtrusive wearable sensors can potentially aid in early detection of cardiac abnormalities, and provides a viable solution for long-term follow-up of patients with chronic cardiovascular diseases without disrupting the daily life activities. Electrocardiogram (ECG) and siesmocardiogram (SCG) signals can be readily acquired from light-weight electrodes and accelerometers respectively, which can be employed to derive systolic time intervals (STI). For this purpose, automated and accurate annotation of the relevant peaks in these signals is required, which is challenging due to the inter-subject morphological variability and noise prone nature of SCG signal. In this paper, an approach is proposed to automatically annotate the desired peaks in SCG signal that are related to STI by utilizing the information of peak detected in the sliding template to narrow-down the search for the desired peak in actual SCG signal. Experimental validation of this approach performed in conventional/controlled supine and realistic/challenging seated conditions, containing over 5600 heart beat cycles shows good performance and robustness of the proposed approach in noisy conditions. Automated measurement of STI in wearable configuration can provide a quantified cardiac health index for long-term monitoring of patients, elderly people at risk and health-enthusiasts.