Cuffless Blood Pressure Measurement Research Articles

Accurate and 30-plus days reliable cuffless blood pressure measurements with 9-minutes personal photoplethysmograph data and mixed deduction learning

Cuffless blood pressure (BP) measurements have long been anticipated, and the PPG (Photoplethysmography)-only method is the most promising one since already embedded in many wearable devices. To further meet the clinical accuracy requirements, PPG-only BP predictions with personalized modeling for overcoming personal deviations have been widely studied, but all required tens to hundreds of minutes of personal PPG measurements for training. Moreover, their accurate test periods without calibration practice were not reported. In this work, we collected records of PPG data from our recruited subjects in real-life scenarios instead of relying on the openly available MIMIC dataset obtained from intensive care unit (ICU) patients. Since our objective is commercial application and a substantial reduction in training data, we tailored our model training to closely mimic real-world usage. To achieve this, we developed a training approach that only requires 9-minutes of personal PPG signal recordings and mixed with other PPG data from our recruited 364 subjects. The modeling is conducted with two-channel paired inputs to the convolutional neural network (CNN)-based model, which we called Mixed Deduction Learning (MDL). The test results of 88 samples from 15 subjects, under testing period up to 30-plus days without extra calibration, revealed that MDL meets most of the standards of AAMI, BHS, and IEEE 1708–2014 (for static test only) for BP measurement devices, which indicates MDL’s long-term stability and consistency. Furthermore, we found that the model with two-channel inputs presents a trend of improving performance as the pool of mixed training data increased, while the conventional one-channel input revealed degraded performance. The outperformance of MDL is attributed to many significant features remained in the first CNN layer even when mixing personal 9-minutes data with the other 364 subjects. Consequently, PPG-only with MDL introduces a new avenue for overcoming challenges in training due to personal physiological variations. Given our consideration of real-life usage, this technology can be seamlessly translated to commercial applications.

A highly elastic, moisturizing, and adhesion conductive hydrogel designed for cuffless blood pressure measurement

Real-time monitoring of blood pressure in athletes during exercise is of great importance, however, conventional methods pose challenges in achieving portable and instantaneous measurements. While, the electrodes used in emerging wearable blood pressure monitoring instruments still have certain defects. In this study, a novel cuffless blood pressure monitoring system was developed using a self-developed polyacrylamide/trehalose/LiCl (PAM/Trehalose/LiCl) conductive hydrogel as the electrode material. The hydrogel electrode has high tensile (strain: 3568 %) and conductive properties (7.57 S/m), while also exhibiting low toughness, moisture retention, and adhesion. While ensuring the comfort of the monitoring system, it can also accurately detect strain over a wide range of deformations (GF value of up to 6.31 in the strain interval of 2000–3400 %). Finally, this system accurately extracts relevant features of PPG and ECG signals and predicts cuff-less blood pressure using the BP Transformer-XL model. The main control chip, ECG detection module, Bluetooth module, and PPG detection module are integrated, and band-pass filtering and data feature extraction techniques are combined. The study found that the average errors in systolic and diastolic readings were 2.09 mmHg and 1.57 mmHg, respectively, compared to those measured by an electronic sphygmomanometer. Therefore, this study successfully combined hydrogel electrodes with a cuffless blood pressure monitoring system to provide technical support for the development of personalized training programs for future athletes and the guarantee of long-term competitive performance.

A paralleled CNN and Transformer network for PPG-based cuff-less blood pressure estimation

Cuff-less blood pressure measurement plays a pivotal part in hypertension diagnosis and prevention. By employing deep learning algorithms, blood pressure estimation applying photoplethysmography (PPG) signals has achieved remarkable improvements and attracted growing attention. However, most existing methods utilize the PPG signal emphasized on localized feature learning, often neglecting the global features in the temporal data. To tackle the issue, we propose a novel architecture. It leverages a parallel processing approach by using Convolutional Neural Network for extracting local features and simultaneously employing Transformer to capture global features. The feature fusion block integrates the features using spatial and channel attention. The blood pressure values are regressed by two fully connected layers. To verify the performance of our method, we conduct evaluations on the Medical Information Mart for Intensive Care (MIMIC) database. On the MIMIC database, our algorithm demonstrates exceptional performance. For Diastolic Pressure, Mean Arterial Pressure, and Systolic Pressure, the mean absolute errors are 2.36 mmHg, 2.03 mmHg, and 4.44 mmHg, respectively. This performance aligns with the rigorous criteria required by an Association for the Advancement of Medical Instrumentation (AAMI) and receives a Grade A rating in accordance with the British Hypertension Society (BHS) standard.

DNN-BP: a novel framework for cuffless blood pressure measurement from optimal PPG features using deep learning model.

Continuous blood pressure (BP) provides essential information for monitoring one's health condition. However, BP is currently monitored using uncomfortable cuff-based devices, which does not support continuous BP monitoring. This paper aims to introduce a blood pressure monitoring algorithm based on only photoplethysmography (PPG) signals using the deep neural network (DNN). The PPG signals are obtained from 125 unique subjects with 218 records and filtered using signal processing algorithms to reduce the effects of noise, such as baseline wandering, and motion artifacts. The proposed algorithm is based on pulse wave analysis of PPG signals, extracted various domain features from PPG signals, and mapped them to BP values. Four feature selection methods are applied and yielded four feature subsets. Therefore, an ensemble feature selection technique is proposed to obtain the optimal feature set based on major voting scores from four feature subsets. DNN models, along with the ensemble feature selection technique, outperformed in estimating the systolic blood pressure (SBP) and diastolic blood pressure (DBP) compared to previously reported approaches that rely only on the PPG signal. The coefficient of determination ( ) and mean absolute error (MAE) of the proposed algorithm are 0.962 and 2.480 mmHg, respectively, for SBP and 0.955 and 1.499 mmHg, respectively, for DBP. The proposed approach meets the Advancement of Medical Instrumentation standard for SBP and DBP estimations. Additionally, according to the British Hypertension Society standard, the results attained Grade A for both SBP and DBP estimations. It concludes that BP can be estimated more accurately using the optimal feature set and DNN models. The proposed algorithm has the potential ability to facilitate mobile healthcare devices to monitor continuous BP.

HGCTNet: Handcrafted Feature-Guided CNN and Transformer Network for Wearable Cuffless Blood Pressure Measurement.

Biosignals collected by wearable devices, such as electrocardiogram and photoplethysmogram, exhibit redundancy and global temporal dependencies, posing a challenge in extracting discriminative features for blood pressure (BP) estimation. To address this challenge, we propose HGCTNet, a handcrafted feature-guided CNN and transformer network for cuffless BP measurement based on wearable devices. By leveraging convolutional operations and self-attention mechanisms, we design a CNN-Transformer hybrid architecture to learn features from biosignals that capture both local information and global temporal dependencies. Then, we introduce a handcrafted feature-guided attention module that utilizes handcrafted features extracted from biosignals as query vectors to eliminate redundant information within the learned features. Finally, we design a feature fusion module that integrates the learned features, handcrafted features, and demographics to enhance model performance. We validate our approach using two large wearable BP datasets: the CAS-BP dataset and the Aurora-BP dataset. Experimental results demonstrate that HGCTNet achieves an estimation error of 0.9 ± 6.5 mmHg for diastolic BP (DBP) and 0.7 ± 8.3 mmHg for systolic BP (SBP) on the CAS-BP dataset. On the Aurora-BP dataset, the corresponding errors are -0.4 ± 7.0 mmHg for DBP and -0.4 ± 8.6 mmHg for SBP. Compared to the current state-of-the-art approaches, HGCTNet reduces the mean absolute error of SBP estimation by 10.68% on the CAS-BP dataset and 9.84% on the Aurora-BP dataset. These results highlight the potential of HGCTNet in improving the performance of wearable cuffless BP measurements.

A Wireless Noninvasive Blood Pressure Measurement System Using MAX30102 and Random Forest Regressor for Photoplethysmography Signals

Hypertension, often termed “the silent killer”, is associated with cardiovascular risk and requires regular blood pressure (BP) monitoring. However, existing methods are cumbersome and require medical expertise, which is worsened by the need for physical contact, particularly during situations such as the coronavirus pandemic that started in 2019 (COVID-19). This study aimed to develop a cuffless, continuous, and accurate BP measurement system using a photoplethysmography (PPG) sensor and a microcontroller via PPG signals. The system utilizes a MAX30102 sensor and ESP-WROOM-32 microcontroller to capture PPG signals that undergo noise reduction during preprocessing. Peak detection and feature extraction algorithms were introduced, and their output data were used to train a machine learning model for BP prediction. Tuning the model resulted in identifying the best-performing model when using a dataset from six subjects with a total of 114 records, thereby achieving a coefficient of determination of 0.37/0.46 and a mean absolute error value of 4.38/4.49 using the random forest algorithm. Integrating this model into a web-based graphical user interface enables its implementation. One probable limitation arises from the small sample size (six participants) of healthy young individuals under seated conditions, thereby potentially hindering the proposed model’s ability to learn and generalize patterns effectively. Increasing the number of participants with diverse ages and medical histories can enhance the accuracy of the proposed model. Nevertheless, this innovative device successfully addresses the need for convenient, remote BP monitoring, particularly during situations like the COVID-19 pandemic, thus making it a promising tool for cardiovascular health management.

Improving the accuracy of blood pressure estimation using a photoplethysmogram due to additional training of an artificial neural network

The paper presents a neural network algorithm for cuffless blood pressure measurement based on computer analysis of the photoplethysmogram signal. An approach is proposed that allows you to significantly reduce the requirements for the size of the training sample required to build an effective machine learning model. The essence of the approach is the procedure for retraining a pre-trained neural network model using a limited amount of training data. At the same time, it is possible to achieve such values of the accuracy of estimating blood pressure values as in full-fledged training, but at the same time, the practical use of the method is significantly simplified.

COMPARISON OF CALIBRATION METHODS IN THE PRECISION OF A RING-TYPE CUFFLESS BLOOD PRESSURE MEASUREMENT DEVICE

Objective: This study aimed to compare the efficacy of two calibration methods for a ring-type cuffless blood pressure (BP) measurement device, CART-I Plus, to assess if regular calibration with a traditional brachial arm cuff device enhances its precision. Design and method: The study involved 40 patients, with 28 completing the study. Participants underwent calibration in three different positions (sitting, standing, supine). Prior to the initial calibration, they were seated quietly for 3–5 minutes. Calibration was executed using a conventional brachial arm cuff device on the same arm. In each position, calibration was performed twice, separated by a minimum interval of one minute. A rest period exceeding two minutes was allowed between position changes for stabilization. BP estimates from the CART-I Plus, derived from photoplethysmography signals, were then compared with corresponding conventional 24-hour ambulatory BP monitoring (ABPM) measurements. Results: The mean differences between the CART-I Plus and ABPM, when calibrated using the sitting position alone (twice), were -2.12±8.53/2.36±7.82 mmHg (awake/sleep SBP) and -4.11±7.73/0.27±7.99 mmHg (awake/sleep DBP), respectively. When calibration included all three positions (six total calibrations), the mean differences remained similar; -2.14±8.95/2.17±8.53 mmHg (awake/sleep SBP) and -4.10±7.91/0.25±7.73 mmHg (awake/sleep DBP) (p > 0.5). Neither the standing nor the supine position, nor a combination of sitting and supine positions, significantly improved precision. In both standing and supine positions, the devices demonstrated a strong correlation for systolic blood pressure (SBP) (r=0.86 and r=0.90, respectively) and diastolic blood pressure (DBP) (r=0.90 for both), with all correlations being statistically significant (p < 0.001). Conclusions: Provided that patient stabilization is adequately ensured, and precise brachial BP measurements are taken by medical personnel, a single sitting position calibration appears to be sufficient. The addition of dynamic three-point calibration involving standing and supine positions did not significantly enhance precision.

FEASIBILITY OF WATCH-TYPE CUFF-LESS BLOOD PRESSURE MEASUREMENT DEVICE IN DAILY BLOOD PRESSURE MONITORING

Objective: Cuff-less blood pressure (BP) measurement devices integrated into smartwatches have gained prominence for daily monitoring, yet limited studies provide the analysis of real-world BP data collected by such devices. In this study, we evaluated the feasibility of photoplethysmographic BP measuring smartwatch by analyzing data collected in real-world. Design and method: The campaign, “Smartwatch Blood Pressure Monitoring Challenge with Korean Hypertension Society”, was held in June 1-14, 2023 and 896 participants reported 35592 BP values measured by smartwatch (Samsung Galaxy Watch 6, approved in Korean Ministry of Food and Drug Safety). Participants were instructed to measure BP daily, in the morning (5AM-8AM) and evening (6PM-9PM) for two weeks, with initial calibration and re-calibration after the first week. We evaluated 1) BP difference between before and after re-calibration, and 2) difference in morning and evening BP. We also evaluated the determining factors of the BP differences. Results: Morning and evening BP values showed significant difference, higher in the evening by 1.42±5.25 mmHg (p<0.05). However, the diurnal variation was much smaller than reported in other studies using conventional BP monitoring devices. ANOVA identified significant variables for evening-morning differences, including basal metabolic rate, skeletal muscle mass, total body water, average systolic BP (SBP), diastolic BP (DBP), and heart rate in the morning (p < 0.01 in both ITT and PP). Participants who have evening-morning difference higher than the third quartile, where evening SBP was significantly higher than morning SBP, exhibited smaller variable sizes. Turkey’s Honest Significant Difference showed that this relationship is statistically significant between Q1-Q4 and Q2-Q4 in all variables. The calibration stability was assessed by the difference in average BP before and after calibration, resulting in an SBP difference of 4.64±4.73 mmHg and DBP of 3.66±3.62 mmHg, smaller than a 2021 campaign (6.8±5.6mmHg). When the difference of average was measured between same period of time (morning versus morning and evening versus evening), SBP of 5.39±5.22 mmHg and DBP of 4.3±3.99 mmHg. Conclusions: In conclusion, watch-based devices may not detect clinical-level BP variability, but refining calibration protocols offers improvement potential.

Non-invasive pulse arrival time as a surrogate for oscillometric systolic blood pressure changes during non-pharmacological intervention

Background. Non-invasive continuous blood pressure (BP) monitoring is of longstanding interest in various cardiovascular scenarios. In this context, pulse arrival time (PAT), i.e., a surrogate parameter for systolic BP (change), became very popular recently, especially in the context of cuffless BP measurement and dedicated lifestyle interventions. Nevertheless, there is also understandable doubt on its reliability in uncontrolled and mobile settings. Objective. The aim of this work is therefore the investigation whether PAT follows oscillometric systolic BP readings during moderate interventions by physical or mental activity using a medical grade handheld device for non-invasive PAT assessment. Approach. A study was conducted featuring an experimental group performing a physical and a mental task, and a control group. Oscillometric BP and PAT were assessed at baseline and after each intervention. Interventions were selected randomly but then performed sequentially in a counterbalanced order. Multivariate analyses of variance were used to test within-subject and between-subject effects for the dependent variables, followed by univariate analyses for post-hoc testing. Furthermore, correlation analysis was performed to assess the association of intervention effects between BP and PAT. Main results. The study included 51 subjects (31 females). Multivariate analysis of variances showed that effects in BP, heart rate, PAT and pulse wave parameters were consistent and significantly different between experimental and control groups. After physical activity, heart rate and systolic BP increased significantly whereas PAT decreased significantly. Mental activity leads to a decrease in systolic BP at stable heart rate. Pulse wave parameters follow accordingly by an increase of PAT and mainly unchanged pulse wave analysis features due to constant heart rate. Finally, also the control group behaviour was accurately registered by the PAT method compared to oscillometric cuff. Correlation analyses revealed significant negative associations between changes of systolic BP and changes of PAT from baseline to the physical task (−0.33 [−0.63, 0.01], p < 0.048), and from physical to mental task (−0.51 [−0.77, −0.14], p = 0.001), but not for baseline to mental task (−0.12 [−0,43,0,20], p = 0.50) in the experimental group. Significance. PAT and the used digital, handheld device proved to register changes in BP and heart rate reliably compared to oscillometric measurements during intervention. Therefore, it might add benefit to future mobile health solutions to support BP management by tracking relative, not absolute, BP changes during non-pharmacological interventions.

Using Data Augmentation to Improve the Accuracy of Blood Pressure Measurement Based on Photoplethysmography

Convenient and accurate blood pressure (BP) measurement is of great importance in both clinical and daily life. Although deep learning (DL) can achieve cuff-less BP measurement based on Photoplethysmography (PPG), the performance of DL is affected by few-shot data. Data augmentation becomes an effective way to enhance the size of the training data. In this paper, we use cropping, flipping, DTW barycentric averaging (DBA), generative adversarial network (GAN) and variational auto-encoder (VAE) for the data augmentation of PPG. Furthermore, a PE–CNN–GRU model is designed for cuff-less BP measurement applying position encoding (PE), convolutional neural networks (CNNs) and gated recurrent unit (GRU) networks. Experiment results based on real-life datasets show that VAE is the most suitable method for PPG data augmentation, which can reduce the mean absolute error (MAE) of PE–CNN–GRU when measuring systolic blood pressure (SBP) and diastolic blood pressure (DBP) by 18.80% and 19.84%. After the data augmentation of PPG, PE–CNN–GRU achieves accurate and cuff-less BP measurement, thus providing convenient support for preventing cardiovascular diseases.

A continuous cuffless blood pressure measurement from optimal PPG characteristic features using machine learning algorithms

Background and objectiveHypertension is a potentially dangerous health condition that can be detected by measuring blood pressure (BP). Blood pressure monitoring and measurement are essential for preventing and treating cardiovascular diseases. Cuff-based devices, on the other hand, are uncomfortable and prevent continuous BP measurement. MethodsIn this study, a new non-invasive and cuff-less method for estimating Systolic Blood Pressure (SBP), Mean Arterial Pressure (MAP), and Diastolic Blood Pressure (DBP) has been proposed using characteristic features of photoplethysmogram (PPG) signals and nonlinear regression algorithms. PPG signals were collected from 219 participants, which were then subjected to preprocessing and feature extraction steps. Analyzing PPG and its derivative signals, a total of 46 time, frequency, and time-frequency domain features were extracted. In addition, the age and gender of each subject were also included as features. Further, correlation-based feature selection (CFS) and Relief F feature selection (ReliefF) techniques were used to select the relevant features and reduce the possibility of over-fitting the models. Finally, support vector regression (SVR), K-nearest neighbour regression (KNR), decision tree regression (DTR), and random forest regression (RFR) were established to develop the BP estimation model. Regression models were trained and evaluated on all features as well as selected features. The best regression models for SBP, MAP, and DBP estimations were selected separately. ResultsThe SVR model, along with the ReliefF-based feature selection algorithm, outperforms other algorithms in estimating the SBP, MAP, and DBP with the mean absolute error of 2.49, 1.62 and 1.43 mmHg, respectively. The proposed method meets the Advancement of Medical Instrumentation standard for BP estimations. Based on the British Hypertension Society standard, the results also fall within Grade A for SBP, MAP, and DBP. ConclusionThe findings show that the method can be used to estimate blood pressure non-invasively, without using a cuff or calibration, and only by utilizing the PPG signal characteristic features.

A multi-channel photoplethysmography array with contact-force regulation for tonoarteriographic imaging

Aim: Photoplethysmography (PPG) has gained widespread popularity as a non-invasive method for potential cuff-less blood pressure (BP) measurement in smart devices. However, the accuracy of PPG-based devices is often hindered by motion artifacts, site variability, and inconsistent contact force (CF). This study aims to investigate the influence of CF variations on PPG signals. Methods: To address these challenges, we present a novel approach involving a multi-channel PPG array integrated with CF regulation in the form of a wearable wristband. This platform enables the visualization of regional PPG/BP distribution while simultaneously monitoring CF. Moreover, our research explored the relationship between PPG waveform characteristics and CF during wrist extension. Results: The results of this study reveal that the PPG amplitude (PPGA) and the b/a ratios, computed from the second derivative peaks of the PPG AC pulse wave, exhibit inconsistency in reaction to CF variations. Notably, a shape correlation coefficient of 0.65069 between normalized PPG and flipped CF sheds light on how changes in posture affect PPG measurements. Conclusions: The proposed platform shows promise in mitigating the effects of CF and spatial positioning on PPG, thereby improving measurement precision and offering a novel approach to image tonoarteriographic (TAG) activities for continuous hypertension management.

Ability of a 24-h ambulatory cuffless blood pressure monitoring device to track blood pressure changes in clinical practice.

There is an increasing number of cuffless blood pressure (BP) measurement (BPM) devices. Despite promising results when comparing single measurements, the ability of these devices to track changes in BP levels over 24 h related to an initial calibration BP (CalibBP) is unknown. Our aim was to analyse this ability in a cuffless device using pulse transit time. We prospectively enrolled 166 participants for simultaneously performed cuffless (Somnotouch-NIBP) and cuff-based (Spacelabs 90217A/IEM Mobil-O-graph) 24 h BPM. As CalibBP for the cuffless device, first cuff-based BP was used. As surrogate for changes in BP levels after the CalibBP, we used the difference between the CalibBP and mean 24 h, awake and asleep BP measured by the two devices. In addition, we analysed the relationship between the difference of the CalibBP and the cuff-based BPM versus the difference between the cuff-based and the cuffless BPM devices. Mean(SD) difference between the CalibBP and mean 24hBP by the cuff-based or cuffless BP device were 7.4 (13.2) versus 1.8 (8.3) mmHg for systolic ( P < 0.0001) and 6.6 (6.8) versus 1.6 (5.8) mmHg for diastolic ( P < 0.0001). A near linear relationship was seen among the difference between the CalibBP and the cuff-based BPM values and the difference between the cuff-based and cuffless BPM device. Our data indicate a lower ability of the cuffless BPM device to track changes of BP levels after CalibBP. In addition, cuffless device accuracy was associated with the changes in BP levels after the initial CalibBP - the larger the BP level change, the larger the difference between the devices. https://www.clinicaltrials.gov ; Unique identifier: NCT03054688; NCT03975582.

Feasibility and Effectiveness of a Ring-Type Blood Pressure Measurement Device Compared With 24-Hour Ambulatory Blood Pressure Monitoring Device.

This study aimed to evaluate the applicability and precision of a ring-type cuffless blood pressure (BP) measurement device, CART-I Plus, compared to conventional 24-hour ambulatory BP monitoring (ABPM). Forty patients were recruited, and 33 participants were included in the final analysis. Each participant wore both CART-I Plus and ABPM devices on the same arm for approximately 24 hours. BP estimation from CART-I Plus, derived from photoplethysmography (PPG) signals, were compared with the corresponding ABPM measurements. The CART-I Plus recorded systolic blood pressure (SBP)/diastolic blood pressure (DBP) values of 131.4±14.1/81.1±12.0, 132.7±13.9/81.9±11.9, and 128.7±14.6/79.3±12.2 mmHg for 24-hour, daytime, and nighttime periods respectively, compared to ABPM values of 129.7±11.7/84.4±11.2, 131.9±11.6/86.3±11.1, and 124.5±13.6/80.0±12.2 mmHg. Mean differences in SBP/DBP between the two devices were 1.74±6.69/-3.24±6.51 mmHg, 0.75±7.44/-4.41±7.42 mmHg, and 4.15±6.15/-0.67±5.23 mmHg for 24-hour, daytime, and nighttime periods respectively. Strong correlations were also observed between the devices, with r=0.725 and r=0.750 for transitions in SBP and DBP from daytime to nighttime, respectively (both p<0.001). The CART-I Plus device, with its unique ring-type design, shows promising accuracy in BP estimation and offers a potential avenue for continuous BP monitoring in clinical practice. ClinicalTrials.gov Identifier: NCT06084065.

Robust Feature Selection for BP Estimation in Multiple Populations: Towards Cuffless Ambulatory BP Monitoring.

Current blood pressure (BP) estimation methods have not achieved an accurate and adaptable approach for ambulatory diagnosis and monitoring applications of populations at risk of cardiovascular disease, generally due to a limited sample size. This paper introduces an algorithm for BP estimation solely reliant on photoplethysmography (PPG) signals and demographic features. It automatically obtains signal features and employs the Markov Blanket (MB) feature selection to discern informative and transmissible features, achieving a robust space adaptable to the population shift. This approach was validated with the Aurora-BP database, compromising ambulatory wearable cuffless BP measurements for over 500 individuals. After evaluating several machine-learning regression methods, Gradient Boosting emerged as the most effective. According to the MB feature selection, temporal, frequency, and demographic features ranked highest in importance, while statistical ones were deemed non-significant. A comparative assessment of a generic model (trained on unclassified BP data) and specialized models (tailored to each distinct BP population), demonstrated a consistent superiority of our proposed MB feature space with a mean absolute error of [Formula: see text] for systolic BP and [Formula: see text] for diastolic BP on the whole dataset. Moreover, we present a first comparison of in-clinic vs. ambulatory models, with performance significantly lower for the latter with a drop of [Formula: see text] in systolic ( ) and [Formula: see text] for diastolic ( ) estimation errors. This work contributes to the resilient understanding of BP estimation algorithms from PPG signals, providing causal features in the signal and quantifying the disparities between ambulatory and in-clinic measurements.

First-in-Human Study for Evaluating the Accuracy of Smart Ring Based Cuffless Blood Pressure Measurement.

Recently, a ring-type cuffless blood pressure (BP) measuring device has been developed. This study was a prospective, single arm, first-in-human pivotal trial to evaluate accuracy of BP measurement by the new device. The ring-type smart wearable monitoring device measures photoplethysmography signals from the proximal phalanx and transmits the data wirelessly to a connected smartphone. For the BP comparison, a cuff was worn on the arm to check the reference BP by auscultatory method, while the test device was worn on the finger of the opposite arm to measure BP simultaneously. Measurements were repeated for up to three sets each on the left and right arms. The primary outcome measure was mean difference and standard deviation of BP differences between the test device and the reference readings. We obtained 526 sets of systolic BP (SBP) and 513 sets of diastolic BP (DBP) from 89 subjects, with ranges of 80 to 175 mmHg and 43 to 122 mmHg for SBP and DBP, respectively. In sample-wise comparison, the mean difference between the test device and the reference was 0.16 ± 5.90 mmHg (95% limits of agreement [LOA], -11.41, 11.72) in SBP and -0.07 ± 4.68 (95% LOA, -9.26, 9.10) in DBP. The test device showed a strong correlation with the reference for SBP (r = 0.94, P < 0.001) and DBP (r = 0.95, P < 0.001). There were consistent results in subject-wise comparison. The new ring-type BP measuring device showed a good correlation for SBP and DBP with minimal bias compared with an auscultatory method.

Performance of wearable watch-type home blood pressure measurement devices in a real-world clinical sample.

Independent testing of home blood pressure (BP) measurement (HBPM) devices is often lacking, particularly among older and multi-morbid patients. We studied the Bpro G2 (using tonometry), Omron HeartGuide (using occlusive oscillometric technology), and Heartisans (using photoplethysmography) wrist watch HBPM devicesagainst a gold standard brachial sphygmomanometer. To test device performance, we used the ISO81060-2 protocol (though this protocol cannot formally validate cuffless devices). We also used linear mixed models to compare adjusted longitudinal BP measurements between devices. Finally, as a surrogate for usability, we recorded instances of device failure where no BP measurement was returned. We enrolled 128 participants (median [Q1-Q3] age 53 [40-65] years, 51% male, 46% on antihypertensive drugs), of whom 100 were suitable for the primaryanalysis. All three devices had mean BP values within 5mmHg of sphygmomanometry. However, due to insufficient reliability (e.g., wider than accepted standard deviations of mean BP), none of the three devices passed all criteria required by the ISO81060-2 protocol. In adjusted longitudinal analyses, the Omron device also systematically underestimated systolic and diastolic BP (-8.46mmHg; 95% CI 6.07, 10.86; p < 0.001; and -2.53mmHg; 95% CI -4.03, -1.03; p = 0.001; respectively). Nevertheless, compared to the Omron device, BPro and Heartisans devices had increased odds of failure (BPro: odds ratio [OR] 5.24; p < 0.0001; Heartisans: OR 5.61; p < 0.001). While we could not formally validate the cuffless devices, our results show that wearable technologies will require improvements to offer reliable BP assessment. This study also highlights the need for validation protocols specifically designed for cuffless BP measurement technologies.

Practical Application of a New Cuffless Blood Pressure Measurement Method.

This study involved 167 patients aged 31 to 88 years (mean 64.2 ± 7.8 years) with normal blood pressure, high blood pressure, and compensated high blood pressure. During each session, three routine blood pressure measurements with intervals of 30 s were taken using a sphygmomanometer with an appropriate cuff size, and the mean value was selected for comparison. The measurements were carried out by two observers trained at the same time with a reference sphygmomanometer using a Y-shaped connector. In the minute following the last cuff-based measurements, an electrocardiogram (ECG) with an I-lead and a photoplethysmocardiogram were recorded simultaneously for 3 min with the CardioQVARK device. We compared the systolic and diastolic BP obtained from a cuff-based mercury sphygmomanometer and smartphone-case-based BP device: the CardioQVARK monitor. A statistical analysis plan was developed using the IEEE Standard for Wearable Cuffless Blood Pressure Devices. Bland-Altman plots were used to estimate the precision of cuffless measurements. The mean difference between the values defined by CardioQVARK and the cuff-based sphygmomanometer for systolic blood pressure (SBP) was 0.31 ± 3.61, while that for diastolic blood pressure (DBP) was 0.44 ± 3.76. The mean absolute difference (MAD) for SBP was 3.44 ± 2.5 mm Hg, and that for DBP was 3.21 ± 2.82 mm Hg. In the subgroups, the smallest error (less than 3 mm Hg) was observed in the prehypertension group, with a slightly larger error (up to 4 mm Hg) found among patients with a normal blood pressure and stage 1 hypertension. The largest error was found in the stage 2 hypertension group (4-5.5 mm Hg). The largest error was 4.2 mm Hg in the high blood pressure group. We, therefore, did not record an error in excess of 7 mmHg, the upper boundary considered acceptable in the IEEE recommendations. We also did not reach a mean error of 5 mmHg, the upper boundary considered acceptable according to the very recent ESH recommendations. At the same time, in all groups of patients, the systolic blood pressure was determined with an error of less than 5 mm Hg in more than 80% of patients. While this study shows that the CardioQVARK device meets the standards of IEEE, the Bland-Altman analysis indicates that the cuffless measurement of diastolic blood pressure has significant bias. The difference was very small and unlikely to be of clinical relevance for the individual patient, but it may well have epidemiological relevance on a population level. Therefore, the CardioQVARK device, while being worthwhile for monitoring patients over time, may not be suitable for screening purposes. Cuffless blood pressure measurement devices are emerging as a convenient and tolerable alternative to cuff-based devices. However, there are several limitations to cuffless blood pressure measurement devices that should be considered. For instance, this study showed a high proportion of measurements with a measurement error of <5 mmHg, while detecting a small, although statistically significant, bias in the measurement of diastolic blood pressure. This suggests that this device may not be suitable for screening purposes. However, its value for monitoring BP over time is confirmed. Furthermore, and most importantly, the easy measurement method and the device portability (integrated in a smartphone) may increase the self-awareness of hypertensive patients and, potentially, lead to an improved adherence to their treatment. The cuffless blood pressure technology developed in this study was tested in accordance with the IEEE protocol and showed great precision in patient groups with different blood pressure ranges. This approach, therefore, has the potential to be applied in clinical practice.

Photoplethysmography-based cuffless blood pressure estimation: an image encoding and fusion approach

Objective. Photoplethysmography (PPG) is a promising wearable technology that detects volumetric changes in microcirculation using a light source and a sensor on the skin’s surface. PPG has been shown to be useful for non-invasive blood pressure (BP) measurement. Deep learning-based BP measurements are now gaining popularity. However, almost all methods focus on 1D PPG. We aimed to design an end-to-end approach for estimating BP using image encodings from a 2D perspective. Approach. In this paper, we present a BP estimation approach based on an image encoding and fusion (BP-IEF) technique. We convert the PPG into five image encodings and use them as input. The proposed BP-IEF consists of two parts: an encoder and a decoder. In addition, three kinds of well-known neural networks are taken as the fundamental architecture of the encoder. The decoder is a hybrid architecture that consists of convolutional and fully connected layers, which are used to fuse features from the encoder. Main results. The performance of the proposed BP-IEF is evaluated on the UCI database in both non-mixed and mixed manners. On the non-mixed dataset, the root mean square error and mean absolute error for systolic BP (SBP) are 13.031 mmHg and 9.187 mmHg respectively, while for diastolic BP (DBP) they are 5.049 mmHg and 3.810 mmHg. On the mixed dataset, the corresponding values for SBP are 4.623 mmHg and 3.058 mmHg, while for DBP the values are 2.350 mmHg and 1.608 mmHg. In addition, both SBP and DBP estimation on the mixed dataset achieved grade A compared to the British Hypertension Society standard. The DBP estimation on the non-mixed dataset also achieved grade A. Significance. The results indicate that the proposed approach has the potential to improve on the current mobile healthcare for cuffless BP measurement.