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

WF-PPG: A Wrist-finger Dual-Channel Dataset for Studying the Impact of Contact Pressure on PPG Morphology

Photoplethysmography (PPG) is a simple optical technique widely used in wearable devices for continuous cardiac health monitoring. However, the quality of PPG signals, particularly their morphology, is influenced by the contact pressure between the skin and the sensor. This variability in signal quality complicates complex tasks that rely on high-quality signals, such as blood pressure and heart rate variability estimation, making them less reliable or even impossible. To address this issue, we present a novel dataset (termed WF-PPG) comprising PPG signals from the wrist measured under varying contact pressures, along with high-quality PPG signals from the fingertip captured simultaneously. Data collection was conducted using a custom device setup capable of precisely adjusting the contact pressure for wrist PPG signals while also recording additional metrics such as contact pressure, electrocardiogram (ECG), blood pressure, and oxygen saturation. WF-PPG is designed to facilitate the analysis of effects of contact pressure on PPG morphology and to support the development and evaluation of advanced data-driven techniques aimed at enhancing the reliability of PPG-based health monitoring.

Wearable Blood Pressure Monitoring Devices: Understanding Heterogeneity in Design and Evaluation

OBJECTIVE

Rapid advances in cuffless blood pressure (BP) monitoring have the potential to radically transform clinical care for cardiovascular health. However, due to the large heterogeneity in device design and evaluation, it is difficult to critically and quantitatively evaluate research progress. In this two-part manuscript, we provide a principled way of describing and accounting for heterogeneity in device and study design.

METHODS

We first provide an overview of foundational elements and design principles of three critical aspects: 1) sensors and systems, 2) pre-processing and feature extraction, and 3) BP estimation algorithms. Then, we critically analyze the state-of-the-art methods via a systematic review.

RESULTS

First, we find large heterogeneity in study designs, making fair comparisons extremely challenging. Moreover, many study designs have data leakage and are underpowered. We suggest a first open-contribution BP estimation benchmark for standardization. Next, we observe that BP distribution in the study sample and the time between calibration and test in emerging personalized devices confound BP estimation error. We suggest accounting for these using a convenient metric coined "explained deviation". Finally, we complement this manuscript with a website, https://wearablebp.github.io, containing a bibliography, meta-analysis results, datasets, and benchmarks, providing a timely plaWorm to understand state-of-the-art devices.

CONCLUSION

There is large heterogeneity in device and study design, which should be carefully accounted for when designing, comparing, and contrasting studies.

SIGNIFICANCE

Our findings will allow readers to parse out the heterogeneous literature and move toward promising directions for safer and more reliable devices in clinical practice and beyond.

Non-Contact Blood Pressure Estimation From Radar Signals by a Stacked Deformable Convolution Network

This study introduces a contactless blood pressure monitoring approach that combines conventional radar signal processing with novel deep learning architectures. During the preprocessing phase, datasets suitable for synchronization are created by integrating Kalman filtering, multiscale bandpass filters, and a periodic extraction method in the time domain. These data comprise data on chest micro variations, encapsulating a complex array of physiological and biomedical information reflective of cardiac micromotions. The Radar-based Stacked Deformable convolution Network (RSD-Net) integrates channel and spatial self attention mechanisms within a deformable convolutional framework to enhance feature extraction from radar signals. The network architecture systematically employs deformable convolutions for initial deep feature extraction from individual signals. Subsequently, continuous blood pressure estimation is conducted using self attention mechanisms on feature map from single source coupled with multi-feature map channel attention. The performance of model is corroborated via the open-source dataset procured using a non-invasive 24 GHz six-port continuous wave radar system. The dataset, encompassing readings from 30 healthy individuals subjected to diverse conditions including rest, the Valsalva maneuver, apnea, and tilt-table examinations. It serves to substantiate the validity and resilience of the proposed method in the non-contact assessment of continuous blood pressure. Evaluation metrics reveal Pearson correlation coefficients of 0.838 for systolic and 0.797 for diastolic blood pressure predictions. The Mean Error (ME) and Standard Deviation (SD) for systolic and diastolic blood pressure measurements are -0.32 ±6.14 mmHg and -0.20 ±5.50 mmHg, respectively. The ablation study assesses the contribution of different structural components of the RSD-Net, validating their significance in the overall of model performance.

A smartphone application toward detection of systolic hypertension in underserved populations

High systolic blood pressure (BP) is the most important modifiable risk factor for cardiovascular disease. Managing systolic hypertension is especially difficult in underserved populations wherein access to cuff BP devices is limited. We showed that ubiquitous smartphones without force sensing can be converted into absolute pulse pressure (PP) monitors. The concept is for the user to perform guided thumb and hand maneuvers with the phone to induce cuff-like actuation and allow built-in sensors to make cuff-like measurements for computing PP. We developed an Android smartphone PP application. The 'app' could be learned by volunteers and yielded PP with total error < 8 mmHg against cuff PP (N = 24). We also analyzed a large population-level database comprising adults less than 65 years old to show that PP plus other basic information can detect systolic hypertension with ROC AUC of 0.9. The smartphone PP app could ultimately help reduce the burden of systolic hypertension in underserved populations and thus health disparities.

Development of a Personalized Multiclass Classification Model to Detect Blood Pressure Variations Associated with Physical or Cognitive Workload

Comprehending the regulatory mechanisms influencing blood pressure control is pivotal for continuous monitoring of this parameter. Implementing a personalized machine learning model, utilizing data-driven features, presents an opportunity to facilitate tracking blood pressure fluctuations in various conditions. In this work, data-driven photoplethysmograph features extracted from the brachial and digital arteries of 28 healthy subjects were used to feed a random forest classifier in an attempt to develop a system capable of tracking blood pressure. We evaluated the behavior of this latter classifier according to the different sizes of the training set and degrees of personalization used. Aggregated accuracy, precision, recall, and F1-score were equal to 95.1%, 95.2%, 95%, and 95.4% when 30% of a target subject's pulse waveforms were combined with five randomly selected source subjects available in the dataset. Experimental findings illustrated that incorporating a pre-training stage with data from different subjects made it viable to discern morphological distinctions in beat-to-beat pulse waveforms under conditions of cognitive or physical workload.

Evaluating Vascular Depth-Dependent Changes in Multi-Wavelength PPG Signals Due to Contact Force

Photoplethysmography (PPG) is a non-invasive method used for cardiovascular monitoring, with multi-wavelength PPG (MW-PPG) enhancing its efficacy by using multiple wavelengths for improved assessment. This study explores how contact force (CF) variations impact MW-PPG signals. Data from 11 healthy subjects are analyzed to investigate the still understudied specific effects of CF on PPG signals. The obtained dataset includes simultaneous recording of five PPG wavelengths (470, 525, 590, 631, and 940 nm), CF, skin temperature, and the tonometric measurement derived from CF. The evolution of raw signals and the PPG DC and AC components are analyzed in relation to the increasing and decreasing faces of the CF. Findings reveal individual variability in signal responses related to skin and vasculature properties and demonstrate hysteresis and wavelength-dependent responses to CF changes. Notably, all wavelengths except 631 nm showed that the DC component of PPG signals correlates with CF trends, suggesting the potential use of this component as an indirect CF indicator. However, further validation is needed for practical application. The study underscores the importance of biomechanical properties at the measurement site and inter-individual variability and proposes the arterial pressure wave as a key factor in PPG signal formation.

pyPPG: a Python toolbox for comprehensive photoplethysmography signal analysis

Objective.Photoplethysmography is a non-invasive optical technique that measures changes in blood volume within tissues. It is commonly and being increasingly used for a variety of research and clinical applications to assess vascular dynamics and physiological parameters. Yet, contrary to heart rate variability measures, a field which has seen the development of stable standards and advanced toolboxes and software, no such standards and limited open tools exist for continuous photoplethysmogram (PPG) analysis. Consequently, the primary objective of this research was to identify, standardize, implement and validate key digital PPG biomarkers.Approach.This work describes the creation of a standard Python toolbox, denotedpyPPG, for long-term continuous PPG time-series analysis and demonstrates the detection and computation of a high number of fiducial points and digital biomarkers using a standard fingerbased transmission pulse oximeter.Main results.The improved PPG peak detector had an F1-score of 88.19% for the state-of-the-art benchmark when evaluated on 2054 adult polysomnography recordings totaling over 91 million reference beats. The algorithm outperformed the open-source original Matlab implementation by ∼5% when benchmarked on a subset of 100 randomly selected MESA recordings. More than 3000 fiducial points were manually annotated by two annotators in order to validate the fiducial points detector. The detector consistently demonstrated high performance, with a mean absolute error of less than 10 ms for all fiducial points.Significance.Based on these fiducial points,pyPPGengineered a set of 74 PPG biomarkers. Studying PPG time-series variability usingpyPPGcan enhance our understanding of the manifestations and etiology of diseases. This toolbox can also be used for biomarker engineering in training data-driven models.pyPPGis available onhttps://physiozoo.com/.

Effect of transmural pressure on the estimation of arterial stiffness index from the photoplethysmographic waveform

The aim of this study was to find the effect of transmural pressure on the determination of the photoplethysmographic (PPG) waveform arterial stiffness index (PPGAI). The study was conducted on 51 subjects without diagnosis of cardiovascular disease, aged between 24 and 74 years. The relation between the transmural pressure, which is the difference between the arterial blood pressure and the PPG sensor contact pressure, and the PPGAI was determined. PPG, beat-to-beat blood pressure, and sensor contact pressure signals were recorded from the index, middle, and ring finger. The PPG sensor contact pressure of the index finger was increased from 20 to 120 mmHg. The aortic augmentation index (AIx@75) was estimated with a SphygmoCor device as a reference. High correlation coefficients r = 0.79 and r = 0.83 between PPGAI and AIx@75, and low PPGAI standard deviations were observed at the transmural pressures of 10 and 20 mmHg, respectively. Transmural pressure of 20 mmHg can be considered suitable for the PPG signal registration and PPGAI calculation for the assessment of arterial stiffness. In summary, the contact pressure of the sensor should be selected according to theblood pressure of the subject finger in order to achieve the transmural pressure suitable for the assessment of PPGAI and arterial stiffness.

Contact pressure-guided wearable dual-channel bioimpedance device for continuous hemodynamic monitoring

Wearable devices for continuous monitoring of arterial pulse waves have the potential to improve the diagnosis, prognosis, and management of cardiovascular diseases. These pulse wave signals are often affected by the contact pressure between the wearable device and the skin, limiting the accuracy and reliability of hemodynamic parameter quantification. Here, we report a continuous hemodynamic monitoring device that enables the simultaneous recording of dual-channel bioimpedance and quantification of pulse wave velocity (PWV) used to calculate blood pressure (BP). Our investigations demonstrate the effect of contact pressure on bioimpedance and PWV. The pulsatile bioimpedance magnitude reached its maximum when the contact pressure approximated the mean arterial pressure of the subject. We employed PWV to continuously quantify BP while maintaining comfortable contact pressure for prolonged wear. The mean absolute error and standard deviation of the error compared to the reference value were determined to be 0.1 ± 3.3 mmHg for systolic BP, 1.3 ± 3.7 mmHg for diastolic BP, and -0.4 ± 3.0 mmHg for mean arterial pressure when measurements were conducted in the lying down position. This research demonstrates the potential of wearable dual-bioimpedance sensors with contact pressure guidance for reliable and continuous hemodynamic monitoring.

Accuracy and User Acceptability of 24-hour Ambulatory Blood Pressure Monitoring by a Prototype Cuffless Multi-Sensor Device Compared to a Conventional Oscillometric Device

OBJECTIVE

24-hour ambulatory blood pressure monitoring (24ABPM) is state of the art in out-of-office blood pressure (BP) monitoring. Due to discomfort and technical limitations related to cuff-based 24ABPM devices, methods for non-invasive and continuous estimation of BP without the need for a cuff have gained interest. The main aims of the present study were to compare accuracy of a pulse arrival time (PAT) based BP-model and user acceptability of a prototype cuffless multi-sensor device (cuffless device), developed by Aidee Health AS, with a conventional cuff-based oscillometric device (ReferenceBP) during 24ABPM.

METHODS

Ninety-five normotensive and hypertensive adults underwent simultaneous 24ABPM with the cuffless device on the chest and a conventional cuff-based oscillometric device on the non-dominant arm. PAT was calculated using the electrocardiogram (ECG) and photoplethysmography (PPG) sensors incorporated in the chest-worn device. The cuffless device recorded continuously, while ReferenceBP measurements were taken every 20 minutes during daytime and every 30 minutes during nighttime. Two-minute PAT-based BP predictions corresponding to the ReferenceBP measurements were compared with ReferenceBP measurements using paired t-tests, bias, and limits of agreement.

RESULTS

Mean (SD) of ReferenceBP compared to PAT-based daytime and nighttime systolic BP (SBP) were 129.7 (13.8) mmHg versus 133.6 (20.9) mmHg and 113.1 (16.5) mmHg versus 131.9 (23.4) mmHg. Ninety-five % limits of agreements were [-26.7, 34.6 mmHg] and [-20.9, 58.4 mmHg] for daytime and nighttime SBP respectively. The cuffless device was reported to be significantly more comfortable and less disturbing than the ReferenceBP device during 24ABPM.

CONCLUSIONS

In the present study, we demonstrated that a general PAT-based BP model had unsatisfactory agreement with ambulatory BP during 24ABPM, especially during nighttime. If sufficient accuracy can be achieved, cuffless BP devices have promising potential for clinical assessment of BP due to the opportunities provided by continuous BP measurements during real-life conditions and high user acceptability.

Smartphone-Based Blood Pressure Monitoring via the Oscillometric Finger Pressing Method: Analysis of Oscillation Width Variations Can Improve Diastolic Pressure Computation

OBJECTIVE

Oscillometric finger pressing is a potential method for absolute blood pressure (BP) monitoring via a smartphone. The user presses their fingertip against a photoplethysmography-force sensor unit on a smartphone to steadily increase the external pressure on the underlying artery. Meanwhile, the phone guides the finger pressing and computes systolic BP (SP) and diastolic BP (DP) from the measured blood volume oscillations and finger pressure. The objective was to develop and evaluate reliable finger oscillometric BP computation algorithms.

METHODS

The collapsibility of thin finger arteries was exploited in an oscillometric model to develop simple algorithms for computing BP from the finger pressing measurements. These algorithms extract features from "width" oscillograms (oscillation width versus finger pressure functions) and the conventional "height" oscillogram for markers of DP and SP. Finger pressing measurements were obtained using a custom system along with reference arm cuff BP measurements from 22 subjects. Measurements were also obtained during BP interventions in some subjects for 34 total measurements.

RESULTS

An algorithm employing the average of width and height oscillogram features predicted DP with correlation of 0.86 and precision error of 8.6 mmHg with respect to the reference measurements. Analysis of arm oscillometric cuff pressure waveforms from an existing patient database provided evidence that the width oscillogram features are better suited to finger oscillometry.

CONCLUSION

Analysis of oscillation width variations during finger pressing can improve DP computation.

SIGNIFICANCE

The study findings may help in converting widely available devices into truly cuffless BP monitors for improving hypertension awareness and control.

A new method of continuous blood pressure monitoring using multichannel sensing signals on the wrist

Hypertension is a worldwide health problem and a primary risk factor for cardiovascular disease. Continuous monitoring of blood pressure has important clinical value for the early diagnosis and prevention of cardiovascular disease. However, existing technologies for wearable continuous blood pressure monitoring are usually inaccurate, rely on subject-specific calibration and have poor generalization across individuals, which limit their practical applications. Here, we report a new blood pressure measurement method and develop an associated wearable device to implement continuous blood pressure monitoring for new subjects. The wearable device detects cardiac output and pulse waveform features through dual photoplethysmography (PPG) sensors worn on the palmar and dorsal sides of the wrist, incorporating custom-made interface sensors to detect the wearing contact pressure and skin temperature. The detected multichannel signals are fused using a machine-learning algorithm to estimate continuous blood pressure in real time. This dual PPG sensing method effectively eliminates the personal differences in PPG signals caused by different people and different wearing conditions. The proposed wearable device enables continuous blood pressure monitoring with good generalizability across individuals and demonstrates promising potential in personal health care applications.

Filtering-induced changes of pulse transmit time across different ages: a neglected concern in photoplethysmography-based cuffless blood pressure measurement

Background: Pulse transit time (PTT) is a key parameter in cuffless blood pressure measurement based on photoplethysmography (PPG) signals. In wearable PPG sensors, raw PPG signals are filtered, which can change the timing of PPG waveform feature points, leading to inaccurate PTT estimation. There is a lack of comprehensive investigation of filtering-induced PTT changes in subjects with different ages. Objective: This study aimed to quantitatively investigate the effects of aging and PTT definition on the infinite impulse response (IIR) filtering-induced PTT changes. Methods: One hundred healthy subjects in five different ranges of age (i.e., 20-29, 30-39, 40-49, 50-59, and over 60 years old, 20 subjects in each) were recruited. Electrocardiogram (ECG) and PPG signals were recorded simultaneously for 120 s. PTT was calculated from the R wave of ECG and PPG waveform features. Eight PTT definitions were developed from different PPG waveform feature points. The raw PPG signals were preprocessed then further low-pass filtered. The difference between PTTs derived from preprocessed and filtered PPG signals, and the relative difference, were calculated and compared among five age groups and eight PTT definitions using the analysis of variance (ANOVA) or Scheirer-Ray-Hare test with post hoc analysis. Linear regression analysis was used to investigate the relationship between age and filtering-induced PTT changes. Results: Filtering-induced PTT difference and the relative difference were significantly influenced by age and PTT definition (p < 0.001 for both). Aging effect on filtering-induced PTT changes was consecutive with a monotonous trend under all PTT definitions. The age groups with maximum and minimum filtering-induced PTT changes depended on the definition. In all subjects, the PTT defined by maximum peak of PPG had the minimum filtering-induced PTT changes (mean: 16.16 ms and 5.65% for PTT difference and relative difference). The changes of PTT defined by maximum first PPG derivative had the strongest linear relationship with age (R-squared: 0.47 and 0.46 for PTT difference relative difference). Conclusion: The filtering-induced PTT changes are significantly influenced by age and PTT definition. These factors deserve further consideration to improve the accuracy of PPG-based cuffless blood pressure measurement using wearable sensors.

Novel Multi-Parametric Sensor System for Comprehensive Multi-Wavelength Photoplethysmography Characterization

Photoplethysmography (PPG) is widely used to assess cardiovascular health. However, its usage and standardization are limited by the impact of variable contact force and temperature, which influence the accuracy and reliability of the measurements. Although some studies have evaluated the impact of these phenomena on signal amplitude, there is still a lack of knowledge about how these perturbations can distort the signal morphology, especially for multi-wavelength PPG (MW-PPG) measurements. This work presents a modular multi-parametric sensor system that integrates continuous and real-time acquisition of MW-PPG, contact force, and temperature signals. The implemented design solution allows for a comprehensive characterization of the effects of the variations in these phenomena on the contour of the MW-PPG signal. Furthermore, a dynamic DC cancellation circuitry was implemented to improve measurement resolution and obtain high-quality raw multi-parametric data. The accuracy of the MW-PPG signal acquisition was assessed using a synthesized reference PPG optical signal. The performance of the contact force and temperature sensors was evaluated as well. To determine the overall quality of the multi-parametric measurement, an in vivo measurement on the index finger of a volunteer was performed. The results indicate a high precision and accuracy in the measurements, wherein the capacity of the system to obtain high-resolution and low-distortion MW-PPG signals is highlighted. These findings will contribute to developing new signal-processing approaches, advancing the accuracy and robustness of PPG-based systems, and bridging existing gaps in the literature.

Generalization Error of a Regression Model for Non-Invasive Blood Pressure Monitoring using a Single Photoplethysmography (PPG) Signal

Photoplethysmography (PPG) sensors integrated in wearable devices offer the potential to monitor arterial blood pressure (ABP) in patients. Such cuffless, non-invasive, and continuous solution is suitable for remote and ambulatory monitoring. A machine learning model based on PPG signal can be used to detect hypertension, estimate beat-by-beat ABP values, and even reconstruct the shape of the ABP. Overall, models presented in literature have shown good performance, but there is a gap between research and potential real-world use cases. Usually, models are trained and tested on data from the same dataset and same subjects, which may lead to overestimating their accuracy. In this paper: we compare cross-validation, where the test data are from the same dataset as training data, and external validation, where the model is tested on samples from a new dataset, on a regression model which predicts diastolic blood pressure from PPG features. The results show that, in the cross-validation, the predicted and the real values are linearly dependent, while in the external validation, the predicted values are not related to the real ones, but probably just through an average value.

Deep-learning-based blood pressure estimation using multi channel photoplethysmogram and finger pressure with attention mechanism

Recently, several studies have proposed methods for measuring cuffless blood pressure (BP) using finger photoplethysmogram (PPG) signals. This study presents a new BP estimation system that measures PPG signals under progressive finger pressure, making the system relatively robust to errors caused by finger position when using the cuffless oscillometric method. To reduce errors caused by finger position, we developed a sensor that can simultaneously measure multi-channel PPG and force signals in a wide field of view (FOV). We propose a deep-learning-based algorithm that can learn to focus on the optimal PPG channel from multi channel PPG using an attention mechanism. The errors (ME ± STD) of the proposed multi channel system were 0.43±9.35 mmHg and 0.21 ± 7.72 mmHg for SBP and DBP, respectively. Through extensive experiments, we found a significant performance difference depending on the location of the PPG measurement in the BP estimation system using finger pressure.

Remote Blood Pressure Estimation via the Spatiotemporal Mapping of Facial Videos

Blood pressure (BP) monitoring is vital in daily healthcare, especially for cardiovascular diseases. However, BP values are mainly acquired through a contact-sensing method, which is inconvenient and unfriendly for BP monitoring. This paper proposes an efficient end-to-end network for estimating BP values from a facial video to achieve remote BP estimation in daily life. The network first derives a spatiotemporal map of a facial video. Then, it regresses the BP ranges with a designed blood pressure classifier and simultaneously calculates the specific value with a blood pressure calculator in each BP range based on the spatiotemporal map. In addition, an innovative oversampling training strategy was developed to handle the problem of unbalanced data distribution. Finally, we trained the proposed blood pressure estimation network on a private dataset, MPM-BP, and tested it on a popular public dataset, MMSE-HR. As a result, the proposed network achieved a mean absolute error (MAE) and root mean square error (RMSE) of 12.35 mmHg and 16.55 mmHg on systolic BP estimations, and those for diastolic BP were 9.54 mmHg and 12.22 mmHg, which were better than the values obtained in recent works. It can be concluded that the proposed method has excellent potential for camera-based BP monitoring in the indoor scenarios in the real world.

Wearable Two-Dimensional Nanomaterial-Based Flexible Sensors for Blood Pressure Monitoring: A Review

Flexible sensors have been extensively employed in wearable technologies for physiological monitoring given the technological advancement in recent years. Conventional sensors made of silicon or glass substrates may be limited by their rigid structures, bulkiness, and incapability for continuous monitoring of vital signs, such as blood pressure (BP). Two-dimensional (2D) nanomaterials have received considerable attention in the fabrication of flexible sensors due to their large surface-area-to-volume ratio, high electrical conductivity, cost effectiveness, flexibility, and light weight. This review discusses the transduction mechanisms, namely, piezoelectric, capacitive, piezoresistive, and triboelectric, of flexible sensors. Several 2D nanomaterials used as sensing elements for flexible BP sensors are reviewed in terms of their mechanisms, materials, and sensing performance. Previous works on wearable BP sensors are presented, including epidermal patches, electronic tattoos, and commercialized BP patches. Finally, the challenges and future outlook of this emerging technology are addressed for non-invasive and continuous BP monitoring.

New Hemodynamic Parameters in Peri-Operative and Critical Care-Challenges in Translation

Hemodynamic monitoring technologies are evolving continuously-a large number of bedside monitoring options are becoming available in the clinic. Methods such as echocardiography, electrical bioimpedance, and calibrated/uncalibrated analysis of pulse contours are becoming increasingly common. This is leading to a decline in the use of highly invasive monitoring and allowing for safer, more accurate, and continuous measurements. The new devices mainly aim to monitor the well-known hemodynamic variables (e.g., novel pulse contour, bioreactance methods are aimed at measuring widely-used variables such as blood pressure, cardiac output). Even though hemodynamic monitoring is now safer and more accurate, a number of issues remain due to the limited amount of information available for diagnosis and treatment. Extensive work is being carried out in order to allow for more hemodynamic parameters to be measured in the clinic. In this review, we identify and discuss the main sensing strategies aimed at obtaining a more complete picture of the hemodynamic status of a patient, namely: (i) measurement of the circulatory system response to a defined stimulus; (ii) measurement of the microcirculation; (iii) technologies for assessing dynamic vascular mechanisms; and (iv) machine learning methods. By analyzing these four main research strategies, we aim to convey the key aspects, challenges, and clinical value of measuring novel hemodynamic parameters in critical care.

Features from the photoplethysmogram and the electrocardiogram for estimating changes in blood pressure

There is a growing emphasis being placed on the potential for cuffless blood pressure (BP) estimation through modelling of morphological features from the photoplethysmogram (PPG) and electrocardiogram (ECG). However, the appropriate features and models to use remain unclear. We investigated the best features available from the PPG and ECG for BP estimation using both linear and non-linear machine learning models. We conducted a clinical study in which changes in BP ([Formula: see text]BP) were induced by an infusion of phenylephrine in 30 healthy volunteers (53.8% female, 28.0 (9.0) years old). We extracted a large and diverse set of features from both the PPG and the ECG and assessed their individual importance for estimating [Formula: see text]BP through Shapley additive explanation values and a ranking coefficient. We trained, tuned, and evaluated linear (ordinary least squares, OLS) and non-linear (random forest, RF) machine learning models to estimate [Formula: see text]BP in a nested leave-one-subject-out cross-validation framework. We reported the results as correlation coefficient ([Formula: see text]), root mean squared error (RMSE), and mean absolute error (MAE). The non-linear RF model significantly ([Formula: see text]) outperformed the linear OLS model using both the PPG and the ECG signals across all performance metrics. Estimating [Formula: see text]SBP using the PPG alone ([Formula: see text] = 0.86 (0.23), RMSE = 5.66 (4.76) mmHg, MAE = 4.86 (4.29) mmHg) performed significantly better than using the ECG alone ([Formula: see text] = 0.69 (0.45), RMSE = 6.79 (4.76) mmHg, MAE = 5.28 (4.57) mmHg), all [Formula: see text]. The highest ranking features from the PPG largely modelled increasing reflected wave interference driven by changes in arterial stiffness. This finding was supported by changes observed in the PPG waveform in response to the phenylephrine infusion. However, a large number of features were required for accurate BP estimation, highlighting the high complexity of the problem. We conclude that the PPG alone may be further explored as a potential single source, cuffless, blood pressure estimator. The use of the ECG alone is not justified. Non-linear models may perform better as they are able to incorporate interactions between feature values and demographics. However, demographics may not adequately account for the unique and individualised relationship between the extracted features and BP.

Subject-wise model generalization through pooling and patching for regression: Application on non-invasive systolic blood pressure estimation

BACKGROUND

Subject-wise modeling using machine learning is useful in many applications requiring low error and complexity, such as wearable medical devices. However, regression accuracy depends highly on the data available to train the model and the model's generalization ability. Adversely, the prediction error may increase severely if unknown data patterns test the model; such a model is known to be overfitted. In medicine-related applications, such as Non-Invasive Blood Pressure (NIBP) estimation, the high error renders the estimation model useless and dangerous.

METHODS

This paper presents a novel algorithm to handle overfitting by editing the training data to achieve generalization for subject-wise models. The pooling and patching (PaP) algorithms use a relatively short record segment of a subject as a Key-Segment (KS) to search through a larger dataset for similar subjects. Then samples taken from the matched subjects' pool records are used to patch the original subject's KS. Due to the significance of systolic blood pressure (SBP) and the complexity of its variability, non-invasive estimation of SBP from electrocardiography (ECG) and photoplethysmography (PPG) is introduced as an application to assess the algorithm. The study was performed on 2051 subjects with a wide range of age, height, weight, length, and health status. The subjects' records were taken from a large public dataset, VitalDB, which is acquired from subjects undergoing different surgeries. Finally, all the results are obtained without using other model generalization techniques.

RESULTS

The generalization effect of the proposed algorithm, PaP, significantly outperformed cross-validation, which is widely used in regression model generalization. Moreover, the testing results show that a KS of 200 to 2000 samples is sufficient for providing high accuracy for much longer testing data of about 12000 to 24000 samples long, which is less than %10 of the record length on average. Furthermore, compared to other works based on the same dataset, PaP provides a significantly lower mean error of -0.75 ± 5.51 mmHg, with a small training data portion of 15% over 2051 subjects.

Beat-to-Beat Blood Pressure Estimation by Photoplethysmography and Its Interpretation

Blood pressure (BP) is among the most important vital signals. Estimation of absolute BP solely using photoplethysmography (PPG) has gained immense attention over the last years. Available works differ in terms of used features as well as classifiers and bear large differences in their results. This work aims to provide a machine learning method for absolute BP estimation, its interpretation using computational methods and its critical appraisal in face of the current literature. We used data from three different sources including 273 subjects and 259,986 single beats. We extracted multiple features from PPG signals and its derivatives. BP was estimated by xgboost regression. For interpretation we used Shapley additive values (SHAP). Absolute systolic BP estimation using a strict separation of subjects yielded a mean absolute error of 9.456mmHg and correlation of 0.730. The results markedly improve if data separation is changed (MAE: 6.366mmHg, r: 0.874). Interpretation by means of SHAP revealed four features from PPG, its derivation and its decomposition to be most relevant. The presented approach depicts a general way to interpret multivariate prediction algorithms and reveals certain features to be valuable for absolute BP estimation. Our work underlines the considerable impact of data selection and of training/testing separation, which must be considered in detail when algorithms are to be compared. In order to make our work traceable, we have made all methods available to the public.

The Effects of Filtering PPG Signal on Pulse Arrival Time-Systolic Blood Pressure Correlation

Pulse arrival time (PAT), evaluated from electro-cardiogram (ECG) and photoplethysmogram (PPG) signals, has been widely used for cuff-less blood pressure (BP) estimation due to its high correlation with BP. However, the question of whether filtering the PPG signal impacts the extracted PAT values and consequently, the correlation between PAT and BP, has not been investigated before. In this paper, using data from 18 subjects, changes in the PAT values, and in the subject-specific PAT-systolic BP (SBP) correlation caused by filtering the PPG signal with variable cutoff frequencies in the range of 2 to 15 Hz are studied. For PAT extraction, three PPG characteristic points (foot, maximum slope and systolic peak) are considered. Results show that differences in the cutoff frequency can shift the PAT values and introduce a worst-case error of over 8.2 mmHg for SBP estimation, indicating that PPG signal filter settings can impact PAT-based BP estimations. Our study suggests that extracting the PAT from the maximum slope point of PPG signal filtered at 10 Hz provides the most stable correlation with SBP.

Establishing best practices in photoplethysmography signal acquisition and processing

Photoplethysmography is now widely utilised by clinical devices such as pulse oximeters, and wearable devices such as smartwatches. It holds great promise for health monitoring in daily life. This editorial considers whether it would be possible and beneficial to establish best practices for photoplethysmography signal acquisition and processing. It reports progress made towards this, balanced with the challenges of working with a diverse range of photoplethysmography device designs and intended applications, each of which could benefit from different approaches to signal acquisition and processing. It concludes that there are several potential benefits to establishing best practices. However, it is not yet clear whether it is possible to establish best practices which hold across the range of photoplethysmography device designs and applications.

Challenges Presented by Cuffless Measurement of Blood Pressure if Adopted for Diagnosis and Treatment of Hypertension

The global health burden presented by hypertension is providing increased motivation for improved means of collection of blood pressure (BP) data. A growing area of research and commercial activity is the use of wearable devices to provide BP data using non-invasive cuffless techniques. The accelerated progress in recent years, particularly relating to connectivity of smartphone technology, has promoted the availability of consumer devices that provide values of BP. The main types of devices are wrist-worn, watch-type devices with sensors that typically record a photoplethysmography (PPG) signal, sometimes also with an electrocardiography (ECG) signal. The general underlying concept of the cuffless BP measurement in most device types is the association of BP and the travel time of the arterial pulse between two locations, determined from the time delay between the ECG and PPG signals. Other methods may involve additional analysis of the PPG waveform features. Experimental data are presented to illustrate the challenges presented by cuffless BP techniques in obtaining reliable BP measurements when the change in BP is caused by different stimuli affecting cardiac and vascular mechanisms. These effects influence the association of the measured and physiological BP change, thus presenting significant challenges and potential limitations in the use of cuffless BP devices for the diagnosis and treatment of hypertension.

Cuffless Blood Pressure Measurement

Cuffless blood pressure (BP) measurement has become a popular field due to clinical need and technological opportunity. However, no method has been broadly accepted hitherto. The objective of this review is to accelerate progress in the development and application of cuffless BP measurement methods. We begin by describing the principles of conventional BP measurement, outstanding hypertension/hypotension problems that could be addressed with cuffless methods, and recent technological advances, including smartphone proliferation and wearable sensing, that are driving the field. We then present all major cuffless methods under investigation, including their current evidence. Our presentation includes calibrated methods (i.e., pulse transit time, pulse wave analysis, and facial video processing) and uncalibrated methods (i.e., cuffless oscillometry, ultrasound, and volume control). The calibrated methods can offer convenience advantages, whereas the uncalibrated methods do not require periodic cuff device usage or demographic inputs. We conclude by summarizing the field and highlighting potentially useful future research directions. Expected final online publication date for the Annual Review of Biomedical Engineering, Volume 24 is June 2022. Please see http://www.annualreviews.org/page/journal/pubdates for revised estimates.

A Meta-Learning Approach for Fast Personalization of Modality Translation Models in Wearable Physiological Sensing

Modality translation grants diagnostic value to wearable devices by translating signals collected from low-power sensors to their highly-interpretable counterparts that are more familiar to healthcare providers. For instance, bio-impedance (Bio-Z) is a conveniently collected modality for measuring physiological parameters but is not highly interpretable. Thus, translating it to a well-known modality such as electrocardiogram (ECG) improves the usability of Bio-Z in wearables. Deep learning solutions are well-suited for this task given complex relationships between modalities generated by distinct processes. However, current algorithms usually train a single model for all users that results in ignoring cross-user variations. Retraining for new users usually requires collecting abundant labeled data, which is challenging in healthcare applications. In this paper, we build a modality translation framework to translate Bio-Z to ECG by learning personalized user information without training several independent architectures. Furthermore, our framework is able to adapt to new users in testing using very few samples. We design a meta-learning framework that contains shared and user-specific parameters to account for user differences while learning from the similarity amongst user signals. In this model, a meta-learner approximated by a neural network learns how to learn user-specific parameters and can efficiently update them in testing. Our experiments show that the proposed model reduces the percentage root mean square difference (PRD) by 41% compared to training a single model for all users and by 36% compared to training independent models for each user. When adapting the model to new users, our model outperforms fine-tuning a pre-trained model through back-propagation by 40% using as few as two new samples in testing.

Assessing hemodynamics from the photoplethysmogram to gain insights into vascular age: a review from VascAgeNet

The photoplethysmogram (PPG) signal is widely measured by clinical and consumer devices, and it is emerging as a potential tool for assessing vascular age. The shape and timing of the PPG pulse wave are both influenced by normal vascular aging, changes in arterial stiffness and blood pressure, and atherosclerosis. This review summarizes research into assessing vascular age from the PPG. Three categories of approaches are described: 1) those which use a single PPG signal (based on pulse wave analysis), 2) those which use multiple PPG signals (such as pulse transit time measurement), and 3) those which use PPG and other signals (such as pulse arrival time measurement). Evidence is then presented on the performance, repeatability and reproducibility, and clinical utility of PPG-derived parameters of vascular age. Finally, the review outlines key directions for future research to realize the full potential of photoplethysmography for assessing vascular age.

Wearable Photoplethysmography for Cardiovascular Monitoring

Smart wearables provide an opportunity to monitor health in daily life and are emerging as potential tools for detecting cardiovascular disease (CVD). Wearables such as fitness bands and smartwatches routinely monitor the photoplethysmogram signal, an optical measure of the arterial pulse wave that is strongly influenced by the heart and blood vessels. In this survey, we summarize the fundamentals of wearable photoplethysmography and its analysis, identify its potential clinical applications, and outline pressing directions for future research in order to realize its full potential for tackling CVD.

Hybrid Convolutional Networks for End-to-End Event Detection in Concurrent PPG and PCG Signals Affected by Motion Artifacts

The accurate detection of physiologically-related events in photopletismographic (PPG) and phocardiographic (PCG) signals, recorded by wearable sensors, is mandatory to perform the estimation of relevant cardiovascular parameters like the heart rate and the blood pressure. However, the measurement performed in uncontrolled conditions without clinical supervision leaves the detection quality particularly susceptible to noise and motion artifacts. The performed work proposed a new fully-automatic computational framework, based on convolutional networks, to identify and localize fiducial points in time as the foot, maximum slope and peak in PPG signal and the S1 sound in the PCG signal, both acquired by a custom chest sensor, described recently in the literature by our group. The novelty entailing a custom neural architecture to process sequentially the PPG and PCG signals. Tests were performed analysing four different acquisition conditions (rest, cycling, rest recovery and walking). Cross-validation results for the three PPG fiducial points showed identification accuracy greater than 93 % and localization error (RMSE) less than 10 ms. As expected, cycling and walking conditions provided worse results than rest and recovery, however reaching an accuracy greater than 90 % and a localization error lower than 15 ms. Likewise, the identification and localization error for S1 sound were greater than 90 % and lower than 25 ms. Overall, this study showcased the ability of the proposed technique to detect events with high accuracy not only for steady acquisitions but also during subject movements. We also showed that the proposed network outperformed traditional Shannon-energy-envelope method in the detection of S1 sound. Therefore, we argue that coupling chest sensors and deep learning processing techniques may disclose wearable devices to unobtrusively acquire health information, being less affected by noise and motion artifacts.

External factors that affect the photoplethysmography waveforms

Photoplethysmography (PPG) is a simple and inexpensive technology used in many smart devices to monitor cardiovascular health. The PPG sensors use LED lights to penetrate into the bloodstream to detect the different blood volume changes in the tissue through skin contact by sensing the amount of light that hits the sensor. Typically, the data are displayed on a graph and it forms the pulse waveform. The information from the produced pulse waveform can be useful in calculating measurements that help monitor cardiovascular health, such as blood pressure. With many more people beginning to monitor their health status on their smart devices, it is extremely important that the PPG signal is accurate. Designing a simple experiment with standard laboratory equipment and commercial sensors, we wanted to find how external factors influence the results. In this study, it was found that external factors, touch force and temperature, can have a large impact on the resulting waveform, so the effects of those factors need to be considered in order for the information to become more reliable.

Photoplethysmography Fast Upstroke Time Intervals Can Be Useful Features for Cuff-Less Measurement of Blood Pressure Changes in Humans

OBJECTIVE

Photoplethysmography (PPG) waveform analysis is being increasingly investigated for continuous, non-invasive, and cuff-less blood pressure (BP) measurement. However, the efficacy of this approach and the useful features and models remain largely unclear. The objectives were to develop easy-to-understand models relating PPG waveform features to BP changes (after a cuff calibration) and to determine their value in BP measurement accuracy.

METHODS

The study data comprised finger, toe, and ear PPG waveforms, an ECG waveform, and reference manual cuff BP measurements from 32 human subjects (25% hypertensive) before and after slow breathing, mental arithmetic, cold pressor, and nitroglycerin administration. Stepwise linear regression was employed to create parsimonious models for predicting the intervention-induced BP changes from popular PPG waveform features, pulse arrival time (PAT, time delay between ECG R-wave and PPG foot), and subject demographics. Leave-one-subject-out cross validation was applied to compare the BP change prediction root-mean-squared-errors (RMSEs) of the resulting models to reference models in which PPG waveform features were excluded.

RESULTS

Finger b-time (PPG foot to minimum second derivative time interval) and ear "STT" (PPG amplitude divided by maximum derivative), when combined with PAT, reduced the systolic BP change prediction RMSE of reference models by 6-7% (p 0.022). Ear STT together with pulse width reduced the diastolic BP change prediction RMSE of the reference model by 13% (p = 0.003).

CONCLUSION

The two PPG fast upstroke time intervals can offer some added value in cuff-less BP trending.

SIGNIFICANCE

This study offers important information towards achieving non-invasive and passive BP monitoring without a cuff.

Assessment of Calibration Models for Cuff-Less Blood Pressure Measurement After One Year of Aging

OBJECTIVE

Many calibration models for cuff-less blood pressure (BP) measurement must be periodically updated with cuff BP values to account for vascular aging. However, the time period required for these cuff re-calibrations is largely unknown. The impact of one year of aging on several calibration models was assessed.

METHODS

Ten humans (6 males, 5718 years, 3 hypertensives) were studied during multiple recording sessions that occurred one year apart. In each session, electrocardiography (ECG), ear photoplethysmography (PPG), finger PPG, and toe PPG waveforms and manual cuff BP were recorded before and after slow breathing, mental arithmetic, cold pressor, and nitroglycerin. Linear models based on each PPG waveform, which were previously shown to offer value in predicting the intervention-induced BP changes in a larger subject cohort, were employed. The model coefficients were determined for each subject via one session, and the fully-defined, subject-specific calibration models were then evaluated in the corresponding subjects via the session one year later.

RESULTS

Only a linear model relating toe pulse arrival time (PAT) time delay between ECG R-wave and toe PPG foot to systolic BP (SBP) remained useful. After the year, this model changed little on average (root-mean-squared-error (RMSE) = 1.5 mmHg) and predicted the cuff BP values better than the average of the initial cuff BP values of the subject (RMSE = 9.60.8 mmHg vs. 12.71.0 mmHg; p < 0.05).

CONCLUSION

These results suggest annual cuff recalibrations for the toe PAT-SBP model.

SIGNIFICANCE

Toe PAT may offer a practical recalibration period that fosters user adherence.

Assessment of Non-Invasive Blood Pressure Prediction from PPG and rPPG Signals Using Deep Learning

Exploiting photoplethysmography signals (PPG) for non-invasive blood pressure (BP) measurement is interesting for various reasons. First, PPG can easily be measured using fingerclip sensors. Second, camera based approaches allow to derive remote PPG (rPPG) signals similar to PPG and therefore provide the opportunity for non-invasive measurements of BP. Various methods relying on machine learning techniques have recently been published. Performances are often reported as the mean average error (MAE) on the data which is problematic. This work aims to analyze the PPG- and rPPG based BP prediction error with respect to the underlying data distribution. First, we train established neural network (NN) architectures and derive an appropriate parameterization of input segments drawn from continuous PPG signals. Second, we use this parameterization to train NNs with a larger PPG dataset and carry out a systematic evaluation of the predicted blood pressure. The analysis revealed a strong systematic increase of the prediction error towards less frequent BP values across NN architectures. Moreover, we tested different train/test set split configurations which underpin the importance of a careful subject-aware dataset assignment to prevent overly optimistic results. Third, we use transfer learning to train the NNs for rPPG based BP prediction. The resulting performances are similar to the PPG-only case. Finally, we apply different personalization techniques and retrain our NNs with subject-specific data for both the PPG-only and rPPG case. Whilst the particular technique is less important, personalization reduces the prediction errors significantly.

Posture-Dependent Variability in Wrist Ballistocardiogram-Photoplethysmogram Pulse Transit Time: Implication to Cuff-Less Blood Pressure Tracking

OBJECTIVE

Toward the ultimate goal of robust cuff-less blood pressure (BP) tracking with wrist wearables against postural changes, the goal of this work was to investigate posture-dependent variability in pulse transit time (PTT) measured with ballistocardiogram (BCG) and photoplethysmogram (PPG) signal pair at the wrist.

METHODS

BCG and PPG signals were acquired from 25 subjects under the combination of 3 body (standing, sitting, and supine) and 3 arm (vertical in head-to-foot direction, placed on the chest, and holding a shoulder) postures. PTT was computed as the time interval between the BCG J wave and the PPG foot, and the impact of the 9 postures on PTT was analyzed by invoking an array of possible physical mechanisms.

RESULTS

Our work suggests that (i) wrist BCG-PPG PTT is consistent under standing and sitting postures with vertically held arms; and (ii) changes in wrist orientation and height as well as restrictions in body and arm movement may alter wrist BCG-PPG PTT via distortions in the wrist BCG and PPG waveforms. The results indicate that wrist BCG-PPG PTT varies with respect to postures even when BP remains constant.

CONCLUSION

The potential of cuff-less BP tracking via wrist BCG-PPG PTT demonstrated under standing posture with arms vertically down in the head-to-foot direction may not generalize to other body and arm postures.

SIGNIFICANCE

Understanding the physical mechanisms responsible for posture-induced BCG-PPG PTT variability may increase the versatility of the wrist BCG for cuff-less BP tracking.

Conventional pulse transit times as markers of blood pressure changes in humans

Pulse transit time (PTT) represents a potential approach for cuff-less blood pressure (BP) monitoring. Conventionally, PTT is determined by (1) measuring (a) ECG and ear, finger, or toe PPG waveforms or (b) two of these PPG waveforms and (2) detecting the time delay between the waveforms. The conventional PTTs (cPTTs) were compared in terms of correlation with BP in humans. Thirty-two volunteers [50% female; 52 (17) (mean (SD)) years; 25% hypertensive] were studied. The four waveforms and manual cuff BP were recorded before and after slow breathing, mental arithmetic, cold pressor, and sublingual nitroglycerin. Six cPTTs were detected as the time delays between the ECG R-wave and ear PPG foot, R-wave and finger PPG foot [finger pulse arrival time (PAT)], R-wave and toe PPG foot (toe PAT), ear and finger PPG feet, ear and toe PPG feet, and finger and toe PPG feet. These time delays were also detected via PPG peaks. The best correlation by a substantial extent was between toe PAT via the PPG foot and systolic BP [- 0.63 ± 0.05 (mean ± SE); p < 0.001 via one-way ANOVA]. Toe PAT is superior to other cPTTs including the popular finger PAT as a marker of changes in BP and systolic BP in particular.

A Unified Approach to Wearable Ballistocardiogram Gating and Wave Localization

OBJECTIVE

Toward the ultimate goal of cuff-less blood pressure (BP) trend tracking via pulse transit time (PTT) using wearable ballistocardiogram (BCG) signals, we present a unified approach to the gating of wearable BCG and the localization of wearable BCG waves.

METHODS

We present a unified approach to localize wearable BCG waves suited to various gating and localization reference signals. Our approach gates individual wearable BCG beats and identifies candidate waves in each wearable BCG beat using a fiducial point in a reference signal, and exploits a pre-specified probability distribution of the time interval between the BCG wave and the fiducial point in the reference signal to accurately localize the wave in each wearable BCG beat. We tested the validity of our approach using experimental data collected from 17 healthy volunteers.

RESULTS

We showed that our approach could localize the J wave in the wearable wrist BCG accurately with both the electrocardiogram (ECG) and the wearable wrist photoplethysmogram (PPG) signals as reference, and that the wrist BCG-PPG PTT thus derived exhibited high correlation to BP.

CONCLUSION

We demonstrated the proof-of-concept of a unified approach to localize wearable BCG waves suited to various gating and localization reference signals compatible with wearable measurement.

SIGNIFICANCE

Prior work using the BCG itself or the ECG to gate the BCG beats and localize the waves to compute PTT are not ideally suited to the wearable BCG. Our approach may foster the development of cuff-less BP monitoring technologies based on the wearable BCG.