Cuffless Blood Pressure Estimation Algorithms for Continuous Health-Care Monitoring

Novel Deep Convolutional Neural Network for Cuff-less Blood Pressure Measurement Using ECG and PPG Signals

Cuff-less blood pressure (BP) is a potential method for BP monitoring because it is undisturbed and continuous monitoring. Existing cuff-less estimation methods are easily influenced by signal noise and non-ideal signal morphology. In this study we propose a novel well-designed Convolutional Neural Network (CNN) model named Deep-BP for BP estimation task. The structure of Deep-BP can help to capture more underlying data features associated with BP than handcrafted features, thus improving the robustness and estimation accuracy. We carry out experiments with and without calibration procedure in training stage to evaluate the performance of new method in different application scenarios. The experiment results show that the Deep-BP model achieves high accuracy and outperforms existing methods, in the experiments both with and without calibration.

End-to-End Deep Learning Architecture for Continuous Blood Pressure Estimation Using Attention Mechanism

Blood pressure (BP) is a vital sign that provides fundamental health information regarding patients. Continuous BP monitoring is important for patients with hypertension. Various studies have proposed cuff-less BP monitoring methods using pulse transit time. We propose an end-to-end deep learning architecture using only raw signals without the process of extracting features to improve the BP estimation performance using the attention mechanism. The proposed model consisted of a convolutional neural network, a bidirectional gated recurrent unit, and an attention mechanism. The model was trained by a calibration-based method, using the data of each subject. The performance of the model was compared to the model that used each combination of the three signals, and the model with the attention mechanism showed better performance than other state-of-the-art methods, including conventional linear regression method using pulse transit time (PTT). A total of 15 subjects were recruited, and electrocardiogram, ballistocardiogram, and photoplethysmogram levels were measured. The 95% confidence interval of the reference BP was [86.34, 143.74] and [51.28, 88.74] for systolic BP (SBP) and diastolic BP (DBP), respectively. The R 2 values were 0.52 and 0.49, and the mean-absolute-error values were 4.06 ± 4.04 and 3.33 ± 3.42 for SBP and DBP, respectively. In addition, the results complied with global standards. The results show the applicability of the proposed model as an analytical metric for BP estimation.

A Revised Point-to-Point Calibration Approach with Adaptive Errors Correction to Weaken Initial Sensitivity of Cuff-Less Blood Pressure Estimation

Initial calibration is a great challenge for cuff-less blood pressure (BP) measurement. The traditional one point-to-point (oPTP) calibration procedure only uses one sample/point to obtain unknown parameters of a specific model in a calm state. In fact, parameters such as pulse transit time (PTT) and BP still have slight fluctuations at rest for each subject. The conventional oPTP method had a strong sensitivity in the selection of initial value. Yet, the initial sensitivity of calibration has not been reported and investigated in cuff-less BP motoring. In this study, a mean point-to-point (mPTP) paring calibration method through averaging and balancing calm or peaceful states was proposed for the first time. Thus, based on mPTP, a factor point-to-point (fPTP) paring calibration method through introducing the penalty factor was further proposed to improve and optimize the performance of BP estimation. Using the oPTP, mPTP, and fPTP methods, a total of more than 100,000 heartbeat samples from 21 healthy subjects were tested and validated in the PTT-based BP monitoring technologies. The results showed that the mPTP and fPTP methods significantly improved the performance of estimating BP compared to the conventional oPTP method. Moreover, the mPTP and fPTP methods could be widely popularized and applied, especially the fPTP method, on estimating cuff-less diastolic blood pressure (DBP). To this extent, the fPTP method weakens the initial calibration sensitivity of cuff-less BP estimation and fills in the ambiguity for individualized calibration procedure.

Uncertainty in Blood Pressure Measurement Estimated Using Ensemble-Based Recursive Methodology

Automated oscillometric blood pressure monitors are commonly used to measure blood pressure for many patients at home, office, and medical centers, and they have been actively studied recently. These devices usually provide a single blood pressure point and they are not able to indicate the uncertainty of the measured quantity. We propose a new technique using an ensemble-based recursive methodology to measure uncertainty for oscillometric blood pressure measurements. There are three stages we consider: the first stage is pre-learning to initialize good parameters using the bagging technique. In the second stage, we fine-tune the parameters using the ensemble-based recursive methodology that is used to accurately estimate blood pressure and then measure the uncertainty for the systolic blood pressure and diastolic blood pressure in the third stage.

Flexible Hybrid Electronics for Digital Healthcare

Recent advances in material innovation and structural design provide routes to flexible hybrid electronics that can combine the high-performance electrical properties of conventional wafer-based electronics with the ability to be stretched, bent, and twisted to arbitrary shapes, revolutionizing the transformation of traditional healthcare to digital healthcare. Here, material innovation and structural design for the preparation of flexible hybrid electronics are reviewed, a brief chronology of these advances is given, and biomedical applications in bioelectrical monitoring and stimulation, optical monitoring and treatment, acoustic imitation and monitoring, bionic touch, and body-fluid testing are described. In conclusion, some remarks on the challenges for future research of flexible hybrid electronics are presented.

PPG Sensor Contact Pressure Should Be Taken Into Account for Cuff-Less Blood Pressure Measurement

OBJECTIVE

Photo-plethysmography (PPG) sensors are often used to detect pulse transit time (PTT) for potential cuff-less blood pressure (BP) measurement. It is known that the contact pressure (CP) of the PPG sensor markedly alters the PPG waveform amplitude. The objective was to test the hypothesis that PTT detected via PPG sensors is likewise impacted by CP.

METHODS

A device was built to measure the time delay between ECG and finger PPG waveforms (i.e., pulse arrival time (PAT) - a popular surrogate of PTT) and the PPG sensor CP at different CP levels. These measurements and finger cuff BP were recorded while the CP was slowly varied in 17 healthy subjects.

RESULTS

Over a physiologic range of CP, the maximum deviations of PAT detected at the PPG foot and peak were 22±2 and 40±7 ms (p<0.05), which translate to ∼11 and ∼20 mmHg BP error based on the literature. The curve relating PAT detected at the PPG foot to CP was U-shaped with minimum near finger diastolic BP. A conceptual model accounting for finger artery viscoelasticity and nonlinearity explained this curve.

CONCLUSION

Since the regulatory bias error for BP measurement is limited to 5 mmHg, PPG sensor CP should be taken into account for cuff-less BP measurement via PTT.

SIGNIFICANCE

This study suggests that widely pursued PPG-based BP measurement devices including those that detect PTT should maintain the CP or include a CP measurement in the calibration equation for deriving BP.

A Machine Learning Approach to Classifying Self-Reported Health Status in a Cohort of Patients With Heart Disease Using Activity Tracker Data

Constructing statistical models using personal sensor data could allow for tracking health status over time, thereby enabling the possibility of early intervention. The goal of this study was to use machine learning algorithms to classify patient-reported outcomes (PROs) using activity tracker data in a cohort of patients with stable ischemic heart disease (SIHD). A population of 182 patients with SIHD were monitored over a period of 12 weeks. Each subject received a Fitbit Charge 2 device to record daily activity data, and each subject completed eight Patient-Reported Outcomes Measurement Information Systems short form at the end of each week as a self-assessment of their health status. Two models were built to classify PRO scores using activity tracker data. The first model treated each week independently, whereas the second used a hidden Markov model (HMM) to take advantage of correlations between successive weeks. Retrospective analysis compared the classification accuracy of the two models and the importance of each feature. In the independent model, a random forest classifier achieved a mean area under curve (AUC) of 0.76 for classifying the physical function PRO. The HMM model achieved significantly better AUCs for all PROs (p < 0.05) other than Fatigue and Sleep Disturbance, with a highest mean AUC of 0.79 for the physical function-short form 10a. Our study demonstrates the ability of activity tracker data to classify health status over time. These results suggest that patient outcomes can be monitored in real time using activity trackers.

Deep learning for prediction of cardiac indices from photoplethysmographic waveform: A virtual database approach

Deep learning methods combined with large datasets have recently shown significant progress in solving several medical tasks. However, collecting and annotating large datasets can be a very cumbersome and expensive task. We tackle these problems with a virtual database approach where training data is generated using computer simulations of related phenomena. Specifically, we concentrate on the following problem: can cardiovascular indices such as aortic elasticity, diastolic and systolic blood pressures, and blood flow from heart be predicted continuously using wearable photoplethysmographic sensors? We simulate the blood flow using a haemodynamic model consisting of the entire human circulation. Repeated evaluation of the simulator allows us to create a database of "virtual subjects" with size that is only limited by available computational resources. Using this database, we train neural networks to predict the cardiac indices from photoplethysmographic signal waveform. We consider two approaches: neural networks based on predefined input features and deep convolutional neural networks taking waveform directly as the input. The performance of the methods is demonstrated using numerical examples, thus carrying out a preliminary assessment of the approaches. The results show improvements in accuracy compared with the previous methods. The improvements are especially significant with indices related to aortic elasticity and maximum blood flow. The proposed approach would provide new means to measure cardiovascular health continuously, for example, with a simple wrist device.

Increase of Input Resistance of a Normal-Mode Helical Antenna (NMHA) in Human Body Application

In recent years, the development of healthcare monitoring devices requires high performance and compact in-body sensor antennas. A normal-mode helical antenna (NMHA) is one of the most suitable candidates that meets the criteria, especially with the ability to achieve high efficiency when the antenna structure is in self-resonant mode. It was reported that when the antenna was placed in a human body, the antenna efficiency was decreased due to the increase of its input resistance (Rin). However, the reason for Rin increase was not clarified. In this paper, the increase of Rin is ensured through experiments and the physical reasons are validated through electromagnetic simulations. In the simulation, the Rin is calculated by placing the NMHA inside a human's stomach, skin and fat. The dependency of Rin to conductivity (σ) is significant. Through current distribution calculation, it is verified that the reason of the increase in Rin is due to the decrease of antenna current. The effects of Rin to bandwidth (BW) and electrical field are also numerically clarified. Furthermore, by using the fabricated human body phantom, the measured Rin and bandwidth are also obtained. From the good agreement between the measured and simulated results, the condition of Rin increment is clarified.

An Optimization Study of Estimating Blood Pressure Models Based on Pulse Arrival Time for Continuous Monitoring

Continuous blood pressure (BP) monitoring has a significant meaning for the prevention and early diagnosis of cardiovascular disease. However, under different calibration methods, it is difficult to determine which model is better for estimating BP. This study was firstly designed to reveal a better BP estimation model by evaluating and optimizing different BP models under a justified and uniform criterion, i.e., the advanced point-to-point pairing method (PTP). Here, the physical trial in this study caused the BP increase largely. In addition, the PPG and ECG signals were collected while the cuff bps were measured for each subject. The validation was conducted on four popular vascular elasticity (VE) models (MK-EE, L-MK, MK-BH, and dMK-BH) and one representative elastic tube (ET) model, i.e., M-M. The results revealed that the VE models except for L-MK outperformed the ET model. The linear L-MK as a VE model had the largest estimated error, and the nonlinear M-M model had a weaker correlation between the estimated BP and the cuff BP than MK-EE, MK-BH, and dMK-BH models. Further, in contrast to L-MK, the dMK-BH model had the strongest correlation and the smallest difference between the estimated BP and the cuff BP including systolic blood pressure (SBP) and diastolic blood pressure (DBP) than others. In this study, the simple MK-EE model showed the best similarity to the dMK-BH model. There were no significant changes between MK-EE and dMK-BH models. These findings indicated that the nonlinear MK-EE model with low estimated error and simple mathematical expression was a good choice for application in wearable sensor devices for cuff-less BP monitoring compared to others.

A Novel Method for Calibration-Based Cuff-Less Blood Pressure Estimation

Cuff-less blood pressure estimation technology is useful for cardiovascular disease monitoring. However, without calibration, cuff-less blood pressure estimation is hard to achieve clinical acceptable performance. The traditional methods are always calibrated with retraining. With the increases of the parameters number, the cost of model retraining increases several times. So we propose a novel blood pressure estimation method, which can be calibrated with reference inputs rather than with retraining. The experiment results suggest that the method we proposed can achieve clinical performance (SBP:-0.004 ± 5.869 mmHg, DBP:-0.004±4.511 mmHg) with low calibration cost.

A Survey of Healthcare Internet-of-Things (HIoT): A Clinical Perspective

In combination with current sociological trends, the maturing development of IoT devices is projected to revolutionize healthcare. A network of body-worn sensors, each with a unique ID, can collect health data that is orders-of-magnitude richer than what is available today from sporadic observations in clinical/hospital environments. When databased, analyzed, and compared against information from other individuals using data analytics, HIoT data enables the personalization and modernization of care with radical improvements in outcomes and reductions in cost. In this paper, we survey existing and emerging technologies that can enable this vision for the future of healthcare, particularly in the clinical practice of healthcare. Three main technology areas underlie the development of this field: (a) sensing, where there is an increased drive for miniaturization and power efficiency; (b) communications, where the enabling factors are ubiquitous connectivity, standardized protocols, and the wide availability of cloud infrastructure, and (c) data analytics and inference, where the availability of large amounts of data and computational resources is revolutionizing algorithms for individualizing inference and actions in health management. Throughout the paper, we use a case study to concretely illustrate the impact of these trends. We conclude our paper with a discussion of the emerging directions, open issues, and challenges.

Cuffless Blood Pressure Monitoring from an Array of Wrist Bio-Impedance Sensors Using Subject-Specific Regression Models: Proof of Concept

Continuous and beat-to-beat monitoring of blood pressure (BP), compared to office-based BP measurement, provides significant advantages in predicting future cardiovascular disease. Traditional BP measurement methods are based on a cuff, which is bulky, obtrusive and not applicable to continuous monitoring. Measurement of pulse transit time (PTT) is one of the prominent cuffless methods for continuous BP monitoring. PTT is the time taken by the pressure pulse to travel between two points in an arterial vessel, which is correlated with the BP. In this paper, we present a new cuffless BP method using an array of wrist-worn bio-impedance sensors placed on the radial and the ulnar arteries of the wrist to monitor the arterial pressure pulse from the blood volume changes at each sensor site. BP is accurately estimated by using AdaBoost regression model based on selected arterial pressure pulse features such as transit time, amplitude and slope of the pressure pulse, which are dependent on the cardiac activity and the vascular properties of the wrist arteries. A separate model is developed for each subject based on calibration data to capture the individual variations of BP parameters. In this pilot study, data was collected from 10 healthy participants with age ranges from 18 to 30 years after exercising using our custom low-noise bio-impedance sensing hardware. Post-exercise BP was accurately estimated with an average correlation coefficient and root mean square error (RMSE) of 0.77 and 2.6 mmHg for the diastolic BP and 0.86 and 3.4 mmHg for the systolic BP.

Highly Sensitive and Wearable Liquid Metal-Based Pressure Sensor for Health Monitoring Applications: Integration of a 3D-Printed Microbump Array with the Microchannel

Wearable pressure sensors capable of sensitive, precise, and continuous measurement of physiological and physical signals have great potential for the monitoring of health status and the early diagnosis of diseases. This work introduces a 3D-printed rigid microbump-integrated liquid metal-based soft pressure sensor (3D-BLiPS) for wearable and health-monitoring applications. Using a 3D-printed master mold based on multimaterial fused deposition modeling, the fabrication of a liquid metal microchannel and the integration of a rigid microbump array above the microchannel are achieved in a one-step, direct process. The microbump array enhances the sensitivity of the pressure sensor (0.158 kPa^-1 ) by locally concentrating the deformation of the microchannel with negligible hysteresis and a stable signal response under cyclic loading. The 3D-BLiPS also demonstrates excellent robustness to 10 000 cycles of multidirectional stretching/bending, changes in temperature, and immersion in water. Finally, these characteristics are suitable for a wide range of applications in health monitoring systems, including a wristband for the continuous monitoring of the epidermal pulse rate for cuffless blood pressure estimation and a wireless wearable device for the monitoring of body pressure using a multiple pressure sensor array for the prevention of pressure ulcers.

Analysis of Pulse Arrival Time as an Indicator of Blood Pressure in a Large Surgical Biosignal Database: Recommendations for Developing Ubiquitous Blood Pressure Monitoring Methods

As non-invasive continuous blood pressure monitoring (NCBPM) has gained wide attraction in the recent decades, many pulse arrival time (PAT) or pulse transit time (PTT) based blood pressure (BP) estimation studies have been conducted. However, most of the studies have used small homogeneous subject pools to generate models of BP based on particular interventions for induced hemodynamic change. In this study, a large open biosignal database from a diverse group of 2309 surgical patients was analyzed to assess the efficacy of PAT, PTT, and confounding factors on the estimation of BP. After pre-processing the dataset, a total of 6,777,308 data pairs of BP and temporal features between electrocardiogram (ECG) and photoplethysmogram (PPG) were extracted and analyzed. Correlation analysis revealed that PAT or PTT extracted from the intersecting-tangent (IT) point of PPG showed the highest mean correlation to BP. The mean correlation between PAT and systolic blood pressure (SBP) was -0.37 and the mean correlation between PAT and diastolic blood pressure (DBP) was -0.30, outperforming the correlation between BP and PTT at -0.12 for SBP and -0.11 for DBP. A linear model of BP with a simple calibration method using PAT as a predictor was developed which satisfied international standards for automatic oscillometric BP monitors in the case of DBP, however, SBP could not be predicted to a satisfactory level due to higher errors. Furthermore, multivariate regression analyses showed that many confounding factors considered in previous studies had inconsistent effects on the degree of correlation between PAT and BP.

Sex-Related Differences in Photoplethysmography Signals Measured From Finger and Toe

Sex plays an important role in the normal cardiovascular system function including resting heart rate and arterial blood pressure. In addition, it has been reported that men and women are at different levels of risk for cardiovascular diseases. The aim of this study was to evaluate and compare the temporal and morphological features of both finger and toe photoplethysmography (PPG), and anthropometric and biological parameters with respect to sex. A customized PPG and electrocardiography (ECG) combo device was developed to measure the signals of interest. ECG/PPG features in addition to subjects’ information were compared regarding finger and toe PPGs. Eighty-eight subjects participated in the study. Linear regression and Student’s t-test were used for statistical analysis. Our results revealed that pulse arrival time (PAT), pulse transit time (PTT), systolic pulse transit time (SPTT), and the ratio of areas under the PPG waveform from the onset to the inflection point and the inflection point to the end of the waveform (S2/S1), are dependent on sex. The highest dependence was shown for the finger PTT while the toe PTT did not indicate any significant dependence on sex. This is the first study that evaluates the effect of sex on cardiovascular system function using finger and toe PPG based features which can help to understand sex-based risk factors for cardiovascular diseases and to improve related disease management and treatments.

Blood Pressure Estimation from Photoplethysmogram Using a Spectro-Temporal Deep Neural Network

Blood pressure (BP) is a direct indicator of hypertension, a dangerous and potentially deadly condition. Regular monitoring of BP is thus important, but many people have aversion towards cuff-based devices, and their limitation is that they can only be used at rest. Using just a photoplethysmogram (PPG) to estimate BP is a potential solution investigated in our study. We analyzed the MIMIC III database for high-quality PPG and arterial BP waveforms, resulting in over 700 h of signals after preprocessing, belonging to 510 subjects. We then used the PPG alongside its first and second derivative as inputs into a novel spectro-temporal deep neural network with residual connections. We have shown in a leave-one-subject-out experiment that the network is able to model the dependency between PPG and BP, achieving mean absolute errors of 9.43 for systolic and 6.88 for diastolic BP. Additionally we have shown that personalization of models is important and substantially improves the results, while deriving a good general predictive model is difficult. We have made crucial parts of our study, especially the list of used subjects and our neural network code, publicly available, in an effort to provide a solid baseline and simplify potential comparison between future studies on an explicit MIMIC III subset.

Dynamic Feature Acquisition Using Denoising Autoencoders

In real-world scenarios, different features have different acquisition costs at test time which necessitates cost-aware methods to optimize the cost and performance tradeoff. This paper introduces a novel and scalable approach for cost-aware feature acquisition at test time. The method incrementally asks for features based on the available context that are known feature values. The proposed method is based on sensitivity analysis in neural networks and density estimation using denoising autoencoders with binary representation layers. In the proposed architecture, a denoising autoencoder is used to handle unknown features (i.e., features that are yet to be acquired), and the sensitivity of predictions with respect to each unknown feature is used as a context-dependent measure of informativeness. We evaluated the proposed method on eight different real-world data sets as well as one synthesized data set and compared its performance with several other approaches in the literature. According to the results, the suggested method is capable of efficiently acquiring features at test time in a cost- and context-aware fashion.

The use of photoplethysmography for assessing hypertension

The measurement of blood pressure (BP) is critical to the treatment and management of many medical conditions. High blood pressure is associated with many chronic disease conditions, and is a major source of mortality and morbidity around the world. For outpatient care as well as general health monitoring, there is great interest in being able to accurately and frequently measure BP outside of a clinical setting, using mobile or wearable devices. One possible solution is photoplethysmography (PPG), which is most commonly used in pulse oximetry in clinical settings for measuring oxygen saturation. PPG technology is becoming more readily available, inexpensive, convenient, and easily integrated into portable devices. Recent advances include the development of smartphones and wearable devices that collect pulse oximeter signals. In this article, we review (i) the state-of-the-art and the literature related to PPG signals collected by pulse oximeters, (ii) various theoretical approaches that have been adopted in PPG BP measurement studies, and (iii) the potential of PPG measurement devices as a wearable application. Past studies on changes in PPG signals and BP are highlighted, and the correlation between PPG signals and BP are discussed. We also review the combined use of features extracted from PPG and other physiological signals in estimating BP. Although the technology is not yet mature, it is anticipated that in the near future, accurate, continuous BP measurements may be available from mobile and wearable devices given their vast potential.

A Non-Invasive Continuous Blood Pressure Estimation Approach Based on Machine Learning

Considering the existing issues of traditional blood pressure (BP) measurement methods and non-invasive continuous BP measurement techniques, this study aims to establish the systolic BP and diastolic BP estimation models based on machine learning using pulse transit time and characteristics of pulse waveform. In the process of model construction, the mean impact value method was introduced to investigate the impact of each feature on the models and the genetic algorithm was introduced to implement parameter optimization. The experimental results showed that the proposed models could effectively describe the nonlinear relationship between the features and BP and had higher accuracy than the traditional methods with the error of 3.27 ± 5.52 mmHg for systolic BP and 1.16 ± 1.97 mmHg for diastolic BP. Moreover, the estimation errors met the requirements of the Advancement of Medical Instrumentation and British Hypertension Society criteria. In conclusion, this study was helpful in promoting the practical application of methods for non-invasive continuous BP estimation models.

Multimodal Wrist Biosensor for Wearable Cuff-less Blood Pressure Monitoring System

We propose a multimodal biosensor for use in continuous blood pressure (BP) monitoring system. Our proposed novel configuration measures photo-plethysmography (PPG) and impedance plethysmography (IPG) signals simultaneously from the subject wrist. The proposed biosensor system enables a fully non-intrusive system that is cuff-less, also utilize a single measurement site for maximum wearability and convenience of the patients. The efficacy of the proposed technique was evaluated on 10 young healthy subjects. Experimental results indicate that the pulse transit time (PTT)-based features calculated from an IPG peak and PPG maximum second derivative (f14) had a relatively high correlation coefficient (r) to the reference BP, with -0.81 ± 0.08 and -0.78 ± 0.09 for systolic BP (SBP) and diastolic BP (DBP), respectively. Moreover, here we proposed two BP estimation models that utilize six- and one-point calibration models. The six-point model is based on the PTT, whereas the one-point model is based on the combined PTT and radial impedance (Z). Thus, in both models, we observed an adequate root-mean-square-error estimation performance, with 4.20 ± 1.66 and 2.90 ± 0.90 for SBP and DBP, respectively, with the PTT BP model; and 6.86 ± 1.65 and 6.67 ± 1.75 for SBP and DBP, respectively, with the PTT-Z BP model. This study suggests the possibility of estimating a subject's BP from only wrist bio-signals. Thus, the six- and one-point PTT-Z calibration models offer adequate performance for practical applications.

Predicting adverse hemodynamic events in critically ill patients

PURPOSE OF REVIEW

The art of predicting future hemodynamic instability in the critically ill has rapidly become a science with the advent of advanced analytical processed based on computer-driven machine learning techniques. How these methods have progressed beyond severity scoring systems to interface with decision-support is summarized.

RECENT FINDINGS

Data mining of large multidimensional clinical time-series databases using a variety of machine learning tools has led to our ability to identify alert artifact and filter it from bedside alarms, display real-time risk stratification at the bedside to aid in clinical decision-making and predict the subsequent development of cardiorespiratory insufficiency hours before these events occur. This fast evolving filed is primarily limited by linkage of high-quality granular to physiologic rationale across heterogeneous clinical care domains.

SUMMARY

Using advanced analytic tools to glean knowledge from clinical data streams is rapidly becoming a reality whose clinical impact potential is great.

Multi-Sensor Fusion Approach for Cuff-Less Blood Pressure Measurement

Ambulatory blood pressure (BP) provides valuable information for cardiovascular risk assessment. The present cuff-based devices are intrusive for long-term BP monitoring, whereas cuff-less BP measurement methods based on pulse transit time or multi-parameter are inferior in robustness and reliability by using electrocardiogram (ECG) and photoplethysmogram signals. This study examined a multi-sensor fusion-based platform and algorithm for systolic BP (SBP), mean arterial pressure (MAP), and diastolic BP (DBP) estimation. The proposed multi-sensor platform was comprised of one ECG sensor and two pulse pressure wave sensors for simultaneous signal collection. After extracting 35 features from the collected signals, a weakly supervised feature selection method was proposed for dimension reduction because the reference oscillometric technique-based BP are intermittent and can be redeemed as coarse-grained labels. BP models were then established using a multi-instance regression algorithm. A total of 85 participants including 17 hypertensive and 12 hypotensive patients were enrolled. Experimental results showed that the proposed approach exhibited good accuracy for diverse population with an estimation error of 1.62 ± 7.76 mmHg for SBP, 1.53 ± 6.03 mmHg for MAP, and 1.49 ± 5.52 for DBP, which complied with the association for the advancement of medical instrumentation standards in BP estimation. Moreover, the estimation accuracy is with random daily fluctuations rather than long-term degradation through a maximum two-month follow-up period indicated good robustness performance. These results suggest that the proposed approach is with high reliability and robustness and thus provides a novel insight for cuff-less BP measurement.

How Effective Is Pulse Arrival Time for Evaluating Blood Pressure? Challenges and Recommendations from a Study Using the MIMIC Database

Cardiovascular disease (CVD) is the number one cause of non-infectious morbidity and mortality in the world. The detection, measurement, and management of high blood pressure play an essential role in the prevention and control of CVDs. However, owing to the limitations and discomfort of traditional blood pressure (BP) detection techniques, many new cuff-less blood pressure approaches have been proposed and explored. Most of these involve arterial wave propagation theory, which is based on pulse arrival time (PAT), the time interval needed for a pulse wave to travel from the heart to some distal place on the body, such as the finger or earlobe. For this study, the Medical Information Mart for Intensive Care (MIMIC) database was used as a benchmark for PAT analysis. Many researchers who use the MIMIC database make the erroneous assumption that all the signals are synchronized. Therefore, we decided to investigate the calculation of PAT intervals in the MIMIC database and check its usefulness for evaluating BP. Our findings have important implications for the future use of the MIMIC database, especially for BP evaluation.

A Highly Sensitive Pressure-Sensing Array for Blood Pressure Estimation Assisted by Machine-Learning Techniques

This work describes the development of a pressure-sensing array for noninvasive continuous blood pulse-wave monitoring. The sensing elements comprise a conductive polymer film and interdigital electrodes patterned on a flexible Parylene C substrate. The polymer film was patterned with microdome structures to enhance the acuteness of pressure sensing. The proposed device uses three pressure-sensing elements in a linear array, which greatly facilitates the blood pulse-wave measurement. The device exhibits high sensitivity (−0.533 kPa⁻¹) and a fast dynamic response. Furthermore, various machine-learning algorithms, including random forest regression (RFR), gradient-boosting regression (GBR), and adaptive boosting regression (ABR), were employed for estimating systolic blood pressure (SBP) and diastolic blood pressure (DBP) from the measured pulse-wave signals. Among these algorithms, the RFR-based method gave the best performance, with the coefficients of determination for the reference and estimated blood pressures being R² = 0.871 for SBP and R² = 0.794 for DBP, respectively.

Pulse transit time technique for cuffless unobtrusive blood pressure measurement: from theory to algorithm

Cuffless technique holds great promise to measure blood pressure (BP) in an unobtrusive way, improving diagnostics and monitoring of hypertension and its related cardiovascular diseases, and maximizing the independence and participation of individual. Pulse transit time (PTT) has been the most commonly employed techniques for cuffless BP estimation. Many studies have been conducted to explore its feasibility and validate its performance in the clinical settings. However, there is still issues and challenges ahead before its wide application. This review will investigate the understanding and development of the PTT technique in depth, with a focus on the physiological regulation of arterial BP, the relationship between PTT and BP, and the summaries of the PTT-based models for BP estimation.

A Chair-Based Unconstrained/Nonintrusive Cuffless Blood Pressure Monitoring System Using a Two-Channel Ballistocardiogram

Hypertension is a well-known chronic disease that causes complications such as cardiovascular diseases or stroke, and thus needs to be continuously managed by using a simple system for measuring blood pressure. The existing method for measuring blood pressure uses a wrapping cuff, which makes measuring difficult for patients. To address this problem, cuffless blood pressure measurement methods that detect the peak pressure via signals measured using photoplethysmogram (PPG) and electrocardiogram (ECG) sensors and use it to calculate the pulse transit time (PTT) or pulse wave velocity (PWV) have been studied. However, a drawback of these methods is that a user must be able to recognize and establish contact with the sensor. Furthermore, the peak of the PPG or ECG cannot be detected if the signal quality drops, leading to a decrease in accuracy. In this study, a chair-type system that can monitor blood pressure using polyvinylidene fluoride (PVDF) films in a nonintrusive manner to users was developed. The proposed method also uses instantaneous phase difference (IPD) instead of PTT as the feature value for estimating blood pressure. Experiments were conducted using a blood pressure estimation model created via an artificial neural network (ANN), which showed that IPD could estimate more accurate readings of blood pressure compared to PTT, thus demonstrating the possibility of a nonintrusive blood pressure monitoring system.

Continuous Blood Pressure Monitoring as a Basis for Ambient Assisted Living (AAL) - Review of Methodologies and Devices

Blood pressure (BP) is a bio-physiological signal that can provide very useful information regarding human's general health. High or low blood pressure or its rapid fluctuations can be associated to various diseases or conditions. Nowadays, high blood pressure is considered to be an important health risk factor and major cause of various health problems worldwide. High blood pressure may precede serious heart diseases, stroke and kidney failure. Accurate blood pressure measurement and monitoring plays fundamental role in diagnosis, prevention and treatment of these diseases. Blood pressure is usually measured in the hospitals, as a part of a standard medical routine. However, there is an increasing demand for methodologies, systems as well as accurate and unobtrusive devices that will permit continuous blood pressure measurement and monitoring for a wide variety of patients, allowing them to perform their daily activities without any disturbance. Technological advancements in the last decade have created opportunities for using various devices as a part of ambient assisted living for improving quality of life for people in their natural environment. The main goal of this paper is to provide a comprehensive review of various methodologies for continuous cuff-less blood pressure measurement, as well as to evidence recently developed devices and systems for continuous blood pressure measurement that can be used in ambient assisted living applications.

Hypertension Assessment Using Photoplethysmography: A Risk Stratification Approach

Hypertension is a common chronic cardiovascular disease (CVD). Early screening and diagnosis of hypertension plays a major role in its prevention and in the control of CVDs. Our study discusses the early screening of hypertension while using the morphological features of photoplethysmography (PPG). Numerous morphological features of PPG and its derivative waves were defined and extracted. Six types of feature selection methods were chosen to screen and evaluate these PPG morphological features. The optimal features were comprehensively analyzed in relation to the physiological processes of the cardiovascular circulatory system. Particularly, the intrinsic relation and physiological significance between the formation process of systolic blood pressure (SBP) and PPG morphology features were analyzed in depth. A variety of linear and nonlinear classification models were established for the comparison trials. The F1 scores for the normotension versus prehypertension, normotension and prehypertension versus hypertension, and normotension versus hypertension trials were 72.97%, 81.82%, and 92.31%, respectively. In summary, this study established a PPG characteristic analysis model and established the intrinsic relationship between SBP and PPG characteristics. Finally, the risk stratification of hypertension at different stages was examined and compared based on the optimal feature subset.

Cuffless Blood Pressure Estimation Using Pressure Pulse Wave Signals

Pulse transit time (PTT) has received considerable attention for noninvasive cuffless blood pressure measurement. However, this approach is inconvenient to deploy in wearable devices because two sensors are required for collecting two-channel physiological signals, such as electrocardiogram and pulse wave signals. In this study, we investigated the pressure pulse wave (PPW) signals collected from one piezoelectric-induced sensor located at a single site for cuffless blood pressure estimation. Twenty-one features were extracted from PPW that collected from the radial artery, and then a linear regression method was used to develop blood pressure estimation models by using the extracted PPW features. Sixty-five middle-aged and elderly participants were recruited to evaluate the performance of the constructed blood pressure estimation models, with oscillometric technique-based blood pressure as a reference. The experimental results indicated that the mean ± standard deviation errors for the estimated systolic blood pressure and diastolic blood pressure were 0.70 ± 7.78 mmHg and 0.83 ± 5.45 mmHg, which achieved a decrease of 1.33 ± 0.37 mmHg in systolic blood pressure and 1.14 ± 0.20 mmHg in diastolic blood pressure, compared with the conventional PTT-based method. The proposed model also demonstrated a high level of robustness in a maximum 60-day follow-up study. These results indicated that PPW obtained from the piezoelectric sensor has great feasibility for cuffless blood pressure estimation, and could serve as a promising method in home healthcare settings.

Blood Pressure Estimation Using Photoplethysmography Only: Comparison between Different Machine Learning Approaches

Introduction

Blood pressure (BP) has been a potential risk factor for cardiovascular diseases. BP measurement is one of the most useful parameters for early diagnosis, prevention, and treatment of cardiovascular diseases. At present, BP measurement mainly relies on cuff-based techniques that cause inconvenience and discomfort to users. Although some of the present prototype cuffless BP measurement techniques are able to reach overall acceptable accuracies, they require an electrocardiogram (ECG) and a photoplethysmograph (PPG) that make them unsuitable for true wearable applications. Therefore, developing a single PPG-based cuffless BP estimation algorithm with enough accuracy would be clinically and practically useful.

Methods

The University of Queensland vital sign dataset (online database) was accessed to extract raw PPG signals and its corresponding reference BPs (systolic BP and diastolic BP). The online database consisted of PPG waveforms of 32 cases from whom 8133 (good quality) signal segments (5 s for each) were extracted, preprocessed, and normalised in both width and amplitude. Three most significant pulse features (pulse area, pulse rising time, and width 25%) with their corresponding reference BPs were used to train and test three machine learning algorithms (regression tree, multiple linear regression (MLR), and support vector machine (SVM)). A 10-fold cross-validation was applied to obtain overall BP estimation accuracy, separately for the three machine learning algorithms. Their estimation accuracies were further analysed separately for three clinical BP categories (normotensive, hypertensive, and hypotensive). Finally, they were compared with the ISO standard for noninvasive BP device validation (average difference no greater than 5 mmHg and SD no greater than 8 mmHg).

Results

In terms of overall estimation accuracy, the regression tree achieved the best overall accuracy for SBP (mean and SD of difference: −0.1 ± 6.5 mmHg) and DBP (mean and SD of difference: −0.6 ± 5.2 mmHg). MLR and SVM achieved the overall mean difference less than 5 mmHg for both SBP and DBP, but their SD of difference was >8 mmHg. Regarding the estimation accuracy in each BP categories, only the regression tree achieved acceptable ISO standard for SBP (−1.1 ± 5.7 mmHg) and DBP (−0.03 ± 5.6 mmHg) in the normotensive category. MLR and SVM did not achieve acceptable accuracies in any BP categories.

Conclusion

This study developed and compared three machine learning algorithms to estimate BPs using PPG only and revealed that the regression tree algorithm was the best approach with overall acceptable accuracy to ISO standard for BP device validation. Furthermore, this study demonstrated that the regression tree algorithm achieved acceptable measurement accuracy only in the normotensive category, suggesting that future algorithm development for BP estimation should be more specific for different BP categories.

Analysis for the Influence of ABR Sensitivity on PTT-Based Cuff-Less Blood Pressure Estimation before and after Exercise

An accurate and continuous measurement of blood pressure (BP) is of great importance for the prognosis of some cardiovascular diseases in out-of-hospital settings. Pulse transit time (PTT) is a well-known cardiovascular parameter which is highly correlated with BP and has been widely applied in the estimation of continuous BP. However, due to the complexity of cardiovascular system, the accuracy of PTT-based BP estimation is still unsatisfactory. Recent studies indicate that, for the subjects before and after exercise, PTT can track the high-frequency BP oscillation (HF-BP) well, but is inadequate to follow the low-frequency BP variance (LF-BP). Unfortunately, the cause for this failure of PTT in LF-BP estimation is still unclear. Based on these previous researches, we investigated the cause behind this failure of PTT in LF-BP estimation. The heart rate- (HR-) related arterial baroreflex (ABR) model was introduced to analyze the failure of PTT in LF-BP estimation. Data from 42 healthy volunteers before and after exercise were collected to evaluate the correlation between the ABR sensitivity and the estimation error of PTT-based BP in LF and HF components. In the correlation plot, an obvious difference was observed between the LF and HF groups. The correlation coefficient r for the ABR sensitivity with the estimation error of systolic BP (SBP) and diastolic BP (DBP) in LF was 0.817 ± 0.038 and 0.757 ± 0.069, respectively. However, those correlation coefficient r for the ABR sensitivity with the estimation error of SBP and DBP in HF was only 0.403 ± 0.145 and 0.274 ± 0.154, respectively. These results indicated that there is an ABR-related complex LF autonomic regulation mechanism on BP, PTT, and HR, which influences the effect of PTT in LF-BP estimation.

Continuous Blood Pressure Monitoring using Wrist-worn Bio-impedance Sensors with Wet Electrodes

Continuous blood pressure (BP) monitoring is essential for diagnosis and management of cardiovascular disorders. Currently, BP is measured using cuff-based methods, which are obtrusive and not suitable for continuous monitoring. Estimation of BP using pulse transit time (PTT) is a prominent method that eliminates the need for a cuff. In this paper, we present a new method to estimate BP based on PTT measurements from an array of 2×2 bio-impedance sensors placed on the wrist, which can be integrated into a small wearable device such as a smart watch for continuous BP monitoring. Diastolic and systolic BP were estimated using AdaBoost regression model based on PTT features extracted from the wrist bio-impedance signals. Data was collected from three participants using our custom bio-impedance sensors. Our method can estimate BP accurately with correlation coefficient, mean absolute error (MAE) and standard deviation (STD) of 0.92, 1.71 and 2.46 mmHg for the diastolic BP and 0.94, 2.57 and 4.35 mmHg for the systolic BP.

Reply to Comment on 'New photoplethysmogram indicators for improving cuffless and continuous blood pressure estimation accuracy'

OBJECTIVE

This article was written by the invitation of the editorial board of Physiological Measurement. It is a Reply to the Comment regarding our recently published paper entitled 'New photoplethysmogram indicators for improving cuffless and continuous blood pressure estimation accuracy' (Lin et al 2018 Physiol. Meas. 29 025005).

APPROACH

We appreciate van Helmond and Joseph's (2018 Physiol. Meas. 098001) interests and comments on our previous paper. In the Comment, they discussed in detail the physiology underlying the pulse arrive time (PAT)-based methods for blood pressure (BP) measurement, and concluded that there are inherent physiological reasons precluding the development of an accurate continuous cuffless BP measurement using PAT-based methods. We could agree with the comments of van Helmond and Joseph about the physiology underlying PAT-based methods for BP measurement. It may be difficult to minimize the confounding effects of physiological factors such as pre-ejection period and smooth muscle tone, etc. However, in this Reply, we discuss some potential solutions to deal with these problems from an engineering point of view.

MAIN RESULTS

When heart rate, more photoplethysmogram (PPG) features, PAT, robust machine learning models, and other techniques were adopted for BP estimation, it is promising for improving the accuracy of BP estimation to an acceptable range that can meet professional standards (e.g. Advancement of Medical Instrumentation (AAMI) standard, British Hypertension Society (BHS) protocol).

SIGNIFICANCE

PAT- and/or PPG-based methods may be a promising technique for continuous and unobtrusive blood pressure measurement.

Hypertension Assessment via ECG and PPG Signals: An Evaluation Using MIMIC Database

Cardiovascular diseases (CVDs) have become the biggest threat to human health, and they are accelerated by hypertension. The best way to avoid the many complications of CVDs is to manage and prevent hypertension at an early stage. However, there are no symptoms at all for most types of hypertension, especially for prehypertension. The awareness and control rates of hypertension are extremely low. In this study, a novel hypertension management method based on arterial wave propagation theory and photoplethysmography (PPG) morphological theory was researched to explore the physiological changes in different blood pressure (BP) levels. Pulse Arrival Time (PAT) and photoplethysmogram (PPG) features were extracted from electrocardiogram (ECG) and PPG signals to represent the arterial wave propagation theory and PPG morphological theory, respectively. Three feature sets, one containing PAT only, one containing PPG features only, and one containing both PAT and PPG features, were used to classify the different BP categories, defined as normotension, prehypertension, and hypertension. PPG features were shown to classify BP categories more accurately than PAT. Furthermore, PAT and PPG combined features improved the BP classification performance. The F1 scores to classify normotension versus prehypertension reached 84.34%, the scores for normotension versus hypertension reached 94.84%, and the scores for normotension plus prehypertension versus hypertension reached 88.49%. This indicates that the simultaneous collection of ECG and PPG signals could detect hypertension.

Monitoring the Relative Blood Pressure Using a Hydraulic Bed Sensor System

We propose a nonwearable hydraulic bed sensor system that is placed underneath the mattress to estimate the relative systolic blood pressure of a subject, which only differs from the actual blood pressure by a scaling and an offset factor. Two types of features are proposed to obtain the relative blood pressure, one based on the strength and the other on the morphology of the bed sensor ballistocardiogram pulses. The relative blood pressure is related to the actual by a scale and an offset factor that can be obtained through calibration. The proposed system is able to extract the relative blood pressure more accurately with a less sophisticated sensor system compared to those from the literature. We tested the system using a dataset collected from 48 subjects right after active exercises. Comparison with the ground truth obtained from the blood pressure cuff validates the promising performance of the proposed system, where the mean correlation between the estimate and the ground truth is near to 90% for the strength feature and 83% for the morphology feature.

Cuffless blood pressure estimation using only a smartphone

Cuffless blood pressure (BP) measurement is an all-inclusive term for a method that aims to measure BP without using a cuff. Recent cuffless technology has made it possible to estimate BP with reasonable accuracy. However, mainstream methods require an electrocardiogram and photoplethysmogram measurements, and frequent calibration procedures using a cuff sphygmomanometer. We therefore developed a far simpler cuffless method, using only heart rate (HR) and modified normalized pulse volume (mNPV) that can be measured using a smartphone, based on the knowledge that ln BP = ln cardiac output (CO) + ln total peripheral resistance (TPR), where CO and TPR are correlated with HR and mNPV, respectively. Here, we show that mean arterial pressure (MAP), systolic BP (SBP), and diastolic BP (DBP) could be estimated using the exponential transformation of linear polynomial equation, (a × ln HR) + (b × ln mNPV) + constant, using only a smartphone, with an accuracy of R > 0.70. This implies that our cuffless method could convert a large number of smartphones or smart watches into simplified sphygmomanometers.

Deep learning ensemble with asymptotic techniques for oscillometric blood pressure estimation

BACKGROUND AND OBJECTIVE

This paper proposes a deep learning based ensemble regression estimator with asymptotic techniques, and offers a method that can decrease uncertainty for oscillometric blood pressure (BP) measurements using the bootstrap and Monte-Carlo approach. While the former is used to estimate SBP and DBP, the latter attempts to determine confidence intervals (CIs) for SBP and DBP based on oscillometric BP measurements.

METHOD

This work originally employs deep belief networks (DBN)-deep neural networks (DNN) to effectively estimate BPs based on oscillometric measurements. However, there are some inherent problems with these methods. First, it is not easy to determine the best DBN-DNN estimator, and worthy information might be omitted when selecting one DBN-DNN estimator and discarding the others. Additionally, our input feature vectors, obtained from only five measurements per subject, represent a very small sample size; this is a critical weakness when using the DBN-DNN technique and can cause overfitting or underfitting, depending on the structure of the algorithm. To address these problems, an ensemble with an asymptotic approach (based on combining the bootstrap with the DBN-DNN technique) is utilized to generate the pseudo features needed to estimate the SBP and DBP. In the first stage, the bootstrap-aggregation technique is used to create ensemble parameters. Afterward, the AdaBoost approach is employed for the second-stage SBP and DBP estimation. We then use the bootstrap and Monte-Carlo techniques in order to determine the CIs based on the target BP estimated using the DBN-DNN ensemble regression estimator with the asymptotic technique in the third stage.

RESULTS

The proposed method can mitigate the estimation uncertainty such as large the standard deviation of error (SDE) on comparing the proposed DBN-DNN ensemble regression estimator with the DBN-DNN single regression estimator, we identify that the SDEs of the SBP and DBP are reduced by 0.58 and 0.57  mmHg, respectively. These indicate that the proposed method actually enhances the performance by 9.18% and 10.88% compared with the DBN-DNN single estimator.

CONCLUSION

The proposed methodology improves the accuracy of BP estimation and reduces the uncertainty for BP estimation.

Toward Ubiquitous Blood Pressure Monitoring via Pulse Transit Time: Predictions on Maximum Calibration Period and Acceptable Error Limits

OBJECTIVE

Pulse transit time (PTT) is being widely pursued for ubiquitous blood pressure (BP) monitoring. PTT-based systems may require periodic cuff calibrations but can still be useful for hypertension screening by affording numerous out-of-clinic measurements that can be averaged. The objective was to predict the maximum calibration period that would not compromise accuracy and acceptable error limits in light of measurement averaging for PTT-based systems.

METHODS

Well-known mathematical models and vast BP data were leveraged. Models relating PTT, age, and gender to BP were employed to determine the maximum time period for the PTT-BP calibration curve to change by <1 mmHg over physiological BP ranges for each age and gender. A model of within-person BP variability was employed to establish the screening accuracy of the conventional cuff-based approach. These models were integrated to investigate the screening accuracy of the average of numerous measurements of a PTT-based system in relation to the accuracy of its individual measurements.

RESULTS

The maximum calibration period was about 1 year for a 30 year old and declined linearly to about 6 months for a 70 year old. A PTT-based system with a precision error of >12 mmHg for systolic BP could achieve the screening accuracy of the cuff-based approach via measurement averaging.

CONCLUSION

This theoretical study indicates that PTT-based BP monitoring is viable even with periodic calibration and seemingly high measurement errors.

SIGNIFICANCE

The predictions may help guide the implementation, evaluation, and application of PTT-based BP monitoring systems in practice.