Non-invasive Blood Pressure Estimation Research Articles

Performance Evaluation of Deep Learning Architectures for Blood Pressure Estimation Using Photoplethysmography

High blood pressure (BP) monitoring Blood pressure (BP) is one of the common cardiovascular diseases and therefore the early high blood pressure (hypertension) detection, management, and prevention are mandatory. One promising method of continuous, non-invasive blood pressure estimation is photoplethysmography (PPG). In this study, a novel method was proposed to introduce the AlexNet framework into the time-frequency domain for classification of BP levels based on PPG signals. The study was conducted using the publicly available Figshare dataset which offers PPG signals, and the blood pressure labels against them. Data balancing techniques were used to alleviate class imbalances. Preprocessing and Feature Extraction of PPG Signals. The PPG signals were preprocessed with noise filtering and signals were then transformed from 1D-time to image to facilitate robust feature extraction. The proposed classification model, based on AlexNet showed the best result, with 98.89% accuracy, recall, and precision, and 99.44% specificity. This model outperformed alternative models (VGG16, DenseNet, ResNet50, GoogleNet) for classifying BP levels into the JNC 7 report standard categories normotension, prehypertension and hypertension. This study has two primary contributions. Initially, it demonstrates the efficacy of AlexNet model to extract meaningful features from PPG signals by its hierarchical convolutional and max-pooling layers thereby enabling accurate classification of BP levels. This study underscores the potential of deep learning and PPG signals for developing a highly accurate and truly non-invasive BP monitoring system. In the second aspect, the study offers a systematic assessment and comparison of the proposed over other well-known deep-learning networks, presenting the effectiveness of the AlexNet-based one. These results are of critical importance in the development of novel non-invasive BP monitoring modalities and optimization of cardiovascular health managements and personalized health cares.

A review of machine learning methods for non-invasive blood pressure estimation

A review of machine learning methods for non-invasive blood pressure estimation

Accuracy of the WatchBP Office Central as a Type 2 device for non-invasive estimation of central aortic blood pressure in children and adolescents.

High blood pressure (BP) in childhood is a recognised precursor of elevated cardiovascular risk in adulthood. Brachial BP is normally used for clinical decision making, but central BP may be a better marker of pressure load on the heart. There is a paucity of validated non-invasive, automated devices for estimating central BP in children and adolescents. In this study, we compared the WatchBP Office Central (a Type 2 central pressure estimation device) against a high-fidelity micromanometer in the ascending aorta of anaesthetised patients undergoing clinically-indicated catheterisation (n = 15, age 4-16 years). As a secondary aim, central systolic BP (cSBP) was also compared to two non-invasive estimation methods in 34 awake patients undergoing routine cardiac MRI (age 10-18 years). WatchBP substantially overestimated cSBP compared to the intra-arterial gold-standard reference (26.1 ± 7.4mmHg), and recruitment was terminated at n = 11 (included in the analysis) due to high statistical certainty that the device would not pass the validation criteria of 5±8 mmHg. WatchBP cSBP was also substantially higher than values obtained from a phase contrast MRI method (11.8 ± 7.9 mmHg) and the SphygmoCor XCEL (13.5 ± 8.9 mmHg) in the awake patient group, which translate to 21-23 mmHg on average after accounting for known/estimated biases in these non-invasive comparators. Compared with invasive central diastolic and systolic BPs, the brachial measures from WatchBP yielded errors of 0.1 ± 5.6 and 12.5 ± 6.0 mmHg respectively. We conclude that the WatchBP substantially overestimates cSBP in children and adolescents. These findings reinforce the need for central BP-measuring devices to be further developed and validated in this population.

UPR-BP: Unsupervised Photoplethysmography Representation Learning for Noninvasive Blood Pressure Estimation

UPR-BP: Unsupervised Photoplethysmography Representation Learning for Noninvasive Blood Pressure Estimation

Non-Invasive Continuous Blood Pressure Estimation from Single-Channel PPG Based on a Temporal Convolutional Network Integrated with an Attention Mechanism

Non-Invasive Continuous Blood Pressure Estimation from Single-Channel PPG Based on a Temporal Convolutional Network Integrated with an Attention Mechanism

An algorithm to detect dicrotic notch in arterial blood pressure and photoplethysmography waveforms using the iterative envelope mean method

Background and objectiveDetection of the dicrotic notch (DN) within a cardiac cycle is essential for assessment of cardiac output, calculation of pulse wave velocity, estimation of left ventricular ejection time, and supporting feature-based machine learning models for noninvasive blood pressure estimation, and hypotension, or hypertension prediction. In this study, we present a new algorithm based on the iterative envelope mean (IEM) method to detect automatically the DN in arterial blood pressure (ABP) and photoplethysmography (PPG) waveforms. MethodsThe algorithm was evaluated on both ABP and PPG waveforms from a large perioperative dataset (MLORD dataset) comprising 17,327 patients. The analysis involved a total of 1,171,288 cardiac cycles for ABP waveforms and 3,424,975 cardiac cycles for PPG waveforms. To evaluate the algorithm's performance, the systolic phase duration (SPD) was employed, which represents the duration from the onset of the systolic phase to the DN in the cardiac cycle. Correlation plots and regression analysis were used to compare the algorithm against marked DN detection, while box plots and Bland-Altman plots were used to compare its performance with both marked DN detection and an established DN detection technique (second derivative). The marking of the DN temporal location was carried out by an experienced researcher using the help of the 'find_peaks' function from the scipy Python package, serving as a reference for the evaluation. The marking was visually validated by both an engineer and an anesthesiologist. The robustness of the algorithm was evaluated as the DN was made less visually distinct across signal-to-noise ratios (SNRs) ranging from -30 dB to -5 dB in both ABP and PPG waveforms. ResultsThe correlation between SPD estimated by the algorithm and that marked by the researcher is strong for both ABP (R2(87,343) =0.99, p<.001) and PPG (R2(86,764) =0.98, p<.001) waveforms. The algorithm had a lower mean error of DN detection (s): 0.0047 (0.0029) for ABP waveforms and 0.0046 (0.0029) for PPG waveforms, compared to 0.0693 (0.0770) for ABP and 0.0968 (0.0909) for PPG waveforms for the established 2nd derivative method. The algorithm has high rate of detectability of DN detection for SNR of >= -9 dB for ABP waveforms and >= -12 dB for PPG waveforms indicating robust performance in detecting the DN when it is less visibly distinct. ConclusionOur proposed IEM- based algorithm can detect DN in both ABP and PPG waveforms with low computational cost, even in cases where it is not distinctly defined within a cardiac cycle of the waveform (‘DN-less signals’). The algorithm can potentially serve as a valuable, fast, and reliable tool for extracting features from ABP and PPG waveforms. It can be especially beneficial in medical applications where DN-based features, such as SPD, diastolic phase duration, and DN amplitude, play a significant role.

STP: Self-supervised transfer learning based on transformer for noninvasive blood pressure estimation using photoplethysmography

Non-invasive blood pressure (BP) monitoring plays a crucial role in cardiovascular disease prevention, but traditional cuff-based methods lack continuous monitoring capability. Photoplethysmography (PPG) offers a promising alternative by capturing blood volume changes optically. The challenge of effectively capturing fine-grained discriminative features of BP using limited paired signals persists, despite the benefits offered by deep models trained on extensive data. The objective of this study is to create a Transformer framework for continuous monitoring of noninvasive BP through self-supervised transfer learning and BP pattern adaptation. We developed a data preprocessing strategy that integrates eleven physiological signal transformations to perform unsupervised pseudo-labels for PPG signals. Then, we developed a self-supervised learning network based on the Transformer architecture to extract robust signal representations from transformed PPGs during the pretraining phase. Furthermore, we designed a transfer learning approach that incorporates BP pattern adaptation to derive discriminative features for accurate estimation of BP values. The proposed STP model was evaluated on multi-source datasets containing 1,213 subjects based on a subject-wise paradigm. The clinical standard was met with estimation errors of 0.85 ± 4.21 mmHg and 0.49 ± 2.76 mmHg for systolic BP and diastolic BP, respectively. These results indicate that the STP model performs competitively in terms of BP estimation when compared to current state-of-the-art approaches. Our study presents a novel STP approach for continuous monitoring of noninvasive BP using PPG signals. By integrating BP pattern adaptation, our approach achieves a competitive performance in estimating BP values, demonstrating its potential for clinical applications.

Current Developments in Cuff-Free Non-invasive Continuous Blood Pressure Estimation Using Photoplethysmography

Current Developments in Cuff-Free Non-invasive Continuous Blood Pressure Estimation Using Photoplethysmography

Utilization of Facial PPG Signals as a Novel Source for Blood Pressure Estimation

Hypertension, a prevalent global health concern, necessitates accurate and non-invasive blood pressure estimation techniques for effective monitoring and management. This paper proposes a novel machine learning approach utilizing Photo plethysmography (PPG) signals for precise blood pressure estimation. PPG signals, obtained conveniently through wearable devices, offer valuable physiological information related to cardiovascular activity. Leveraging advanced machine learning algorithms, including deep learning architectures and feature extraction methods, our proposed technique aims to establish a robust model for blood pressure estimation using facial image analysis. The methodology involves preprocessing PPG signals, extracting relevant features, and employing sophisticated machine learning models for regression analysis. The evaluation of this novel approach involves comprehensive experimentation with diverse datasets, ensuring its efficacy across various demographic groups and conditions. Results demonstrate promising accuracy and reliability in estimating blood pressure values, suggesting the potential for practical implementation in healthcare settings. The proposed technique showcases a promising avenue for non-invasive and accessible blood pressure monitoring, contributing significantly to personalized healthcare and continuous health monitoring systems.

Non-invasive estimation of sbp pressure using a single ppg sensor and self-calibration

This paper presents an efficient approach to the non-invasive estimation of Systolic Blood Pressure (SBP) using just a single photoplethysmography (PPG) sensor and self-calibration procedure. In this scheme, two features of the measured PPG signal, which are systolic upstroke time (SUT) and PPG inter-beat interval (IBI), are extracted to be used in the estimation model. Here, for each individual, a unique model is self-calibrated in order to provide better accuracy. The experimental results demonstrated the merits and promising potentials for future application of the proposed approach.

BrainZ-BP: A Noninvasive Cuff-Less Blood Pressure Estimation Approach Leveraging Brain Bio-Impedance and Electrocardiogram

BrainZ-BP: A Noninvasive Cuff-Less Blood Pressure Estimation Approach Leveraging Brain Bio-Impedance and Electrocardiogram

Meta-Analysis of Pulse Transition Features in Non-Invasive Blood Pressure Estimation Systems: Bridging Physiology and Engineering Perspectives.

The pulse transition features (PTFs), including pulse arrival time (PAT) and pulse transition time (PTT), hold significant importance in estimating non-invasive blood pressure (NIBP). However, the literature showcases considerable variations in terms of PTFs' correlation with blood pressure (BP), accuracy in NIBP estimation, and the comprehension of the relationship between PTFs and BP. This inconsistency is exemplified by the wide-ranging correlations reported across studies investigating the same feature. Furthermore, investigations comparing PAT and PTT have yielded conflicting outcomes. Additionally, PTFs have been derived from various bio-signals, capturing distinct characteristic points like the pulse's foot and peak. To address these inconsistencies, this study meticulously reviews a selection of such research endeavors while aligning them with the biological intricacies of blood pressure and the human cardiovascular system (CVS). Each study underwent evaluation, considering the specific signal acquisition locale and the corresponding recording procedure. Moreover, a comprehensive meta-analysis was conducted, yielding multiple conclusions that could significantly enhance the design and accuracy of NIBP systems. Grounded in these dual aspects, the study systematically examines PTFs in correlation with the specific study conditions and the underlying factors influencing the CVS. This approach serves as a valuable resource for researchers aiming to optimize the design of BP recording experiments, bio-signal acquisition systems, and the fine-tuning of feature engineering methodologies, ultimately advancing PTF-based NIBP estimation.

Blood Pressure Estimation from Speech Recordings: Exploring the Role of Voice-over Artists

Hypertension, a prevalent global health concern, is associated with cardiovascular diseases and significant morbidity and mortality. Accurate and prompt Blood Pressure monitoring is crucial for early detection and successful management. Traditional cuff-based methods can be inconvenient, leading to the exploration of non-invasive and continuous estimation methods. This research aims to bridge the gap between speech processing and health monitoring by investigating the relationship between speech recordings and Blood Pressure estimation. Speech recordings offer promise for non-invasive Blood Pressure estimation due to the potential link between vocal characteristics and physiological responses. In this study, we focus on the role of Voice-over Artists, known for their ability to convey emotions through voice. By exploring the expertise of Voice-over Artists in controlling speech and expressing emotions, we seek valuable insights into the potential correlation between speech characteristics and Blood Pressure. This research sheds light on presenting an innovative and convenient approach to health assessment. By unraveling the specific role of Voice-over Artists in this process, the study lays the foundation for future advancements in healthcare and human-robot interactions. Through the exploration of speech characteristics and emotional expression, this investigation offers valuable insights into the correlation between vocal features and Blood Pressure levels. By leveraging the expertise of Voice-over Artists in conveying emotions through voice, this study enriches our understanding of the intricate relationship between speech recordings and physiological responses, opening new avenues for the integration of voice-related factors in healthcare technologies.

Non-Invasive Estimation of Central Systolic Blood Pressure by Radial Tonometry: A Simplified Approach.

Central systolic blood pressure (cSBP) provides valuable clinical and physiological information. A recent invasive study showed that cSBP can be reliably estimated from mean (MBP) and diastolic (DBP) blood pressure. In this non-invasive study, we compared cSBP calculated using a Direct Central Blood Pressure estimation (DCBP = MBP2/DBP) with cSBP estimated by radial tonometry. Consecutive patients referred for cardiovascular assessment and prevention were prospectively included. Using applanation tonometry with SphygmoCor device, cSBP was estimated using an inbuilt generalized transfer function derived from radial pressure waveform, which was calibrated to oscillometric brachial SBP and DBP. The time-averaged MBP was calculated from the radial pulse waveform. The minimum acceptable error (DCBP-cSBP) was set at ≤5 (mean) and ≤8 mmHg (SD). We included 160 patients (58 years, 54%men). The cSBP was 123.1 ± 18.3 mmHg (range 86-181 mmHg). The (DCBP-cSBP) error was -1.4 ± 4.9 mmHg. There was a linear relationship between cSBP and DCBP (R2 = 0.93). Forty-seven patients (29%) had cSBP values ≥ 130 mmHg, and a DCBP value > 126 mmHg exhibited a sensitivity of 91.5% and specificity of 94.7% in discriminating this threshold (Youden index = 0.86; AUC = 0.965). Using the DCBP formula, radial tonometry allows for the robust estimation of cSBP without the need for a generalized transfer function. This finding may have implications for risk stratification.

Estimation of Physiologic Pressures: Invasive and Non-Invasive Techniques, AI Models, and Future Perspectives.

The measurement of physiologic pressure helps diagnose and prevent associated health complications. From typical conventional methods to more complicated modalities, such as the estimation of intracranial pressures, numerous invasive and noninvasive tools that provide us with insight into daily physiology and aid in understanding pathology are within our grasp. Currently, our standards for estimating vital pressures, including continuous BP measurements, pulmonary capillary wedge pressures, and hepatic portal gradients, involve the use of invasive modalities. As an emerging field in medical technology, artificial intelligence (AI) has been incorporated into analyzing and predicting patterns of physiologic pressures. AI has been used to construct models that have clinical applicability both in hospital settings and at-home settings for ease of use for patients. Studies applying AI to each of these compartmental pressures were searched and shortlisted for thorough assessment and review. There are several AI-based innovations in noninvasive blood pressure estimation based on imaging, auscultation, oscillometry and wearable technology employing biosignals. The purpose of this review is to provide an in-depth assessment of the involved physiologies, prevailing methodologies and emerging technologies incorporating AI in clinical practice for each type of compartmental pressure measurement. We also bring to the forefront AI-based noninvasive estimation techniques for physiologic pressure based on microwave systems that have promising potential for clinical practice.

Nonlinear features of photoplethysmography signals for Non-invasive blood pressure estimation

ObjectiveContinuous monitoring of blood pressure (BP) plays an essential role in the prognosis and prevention of hypertension and related cardiovascular diseases. Moreover, the ever-increasing demand for portable continuous health monitoring systems coupled with promising capabilities of photoplethysmography (PPG) sensors for developing easy-to-use, portable wearable devices have motivated many researchers toward applying PPG signals for non-invasive health monitoring. Nonlinear nature of BP and proven capability of the Poincaré plot for analyzing dynamic behavior of nonlinear systems motivated us to use this powerful tool for BP estimation. This study aims to explore the dynamical behavior of PPG signals to assist feature extraction for machine-learning (ML) algorithms to estimate BP. We proposed a Poincaré-based feature extraction method for BP estimation with no need to extract precise local points on the PPG signal. MethodsFirst, a Poincaré plot of 10-s segments of the PPG signal was prepared. Then, three distinct time series were retrieved from Poincaré mapping of PPG signals; following this, numerous indices were extracted from the three time series. The F-Test feature selection method was applied to pick the most effective features. Finally, the picked features were fed into different ML algorithms for the estimation of BP. ResultsThe proposed method was validated using a subset of the Medical Information Mart for Intensive Care II (MIMIC-II) database containing ambulatory blood pressure (ABP) and PPG records. The performance evaluation was carried out in terms of mean absolute error (MAE), standard deviation (STD), and Pearson’s correlation coefficient. According to the results, application of the Gaussian process regression (GPR) led to the best performance for both systolic and diastolic BP (0.79 ± 3.08 and 1.38 ± 4.53 mmHg, respectively). Satisfying the criteria for Advancement of Medical Instrumentation (AAMI), it was rated as Grade A for systolic and diastolic BP measurements under terms of British Hypertension Society (BHS) standards, which confirms the efficiency of the Poincaré-based features for BP estimation. ConclusionThe results justify the usefulness of the proposed Poincaré-based features as a powerful tool in BP estimation scenarios.

Towards continuous non-invasive blood pressure measurements-interpretation of the vasculature response to cuff inflation.

Blood pressure (BP) surrogates, such as pulse transit time (PTT) or pulse arrival time (PAT), have been intensively explored with the goal of achieving cuffless, continuous, and accurate BP inference. In order to estimate BP, a one-point calibration strategy between PAT and BP is typically used. Recent research focuses on advanced calibration procedures exploiting the cuff inflation process to improve calibration robustness by active and controlled modulation of peripheral PAT, as measured via plethysmograph (PPG) and electrocardiogram (ECG) combination. Such methods require a detailed understanding of the mechanisms behind the vasculature's response to cuff inflation; for this, a model has recently been developed to infer the PAT-BP calibration from measured cuff-induced vasculature changes. The model, while promising, is still preliminary and only partially validated; in-depth analysis and further developments are still needed. Therefore, this work aims to improve our understanding of the cuff-vasculature interaction in this model; we seek to define potential opportunities and to highlight which aspects may require further study. We compare model behaviors with clinical data samples based on a set of observable characteristics relevant for BP inference and calibration. It is found that the observed behaviors are qualitatively well represented with the current simulation model and complexity, with limitations regarding the prediction of the onset of the distal arm dynamics and behavior changes at high cuff pressures. Additionally, a sensitivity analysis of the model's parameter space is conducted to show the factors that influence the characteristics of its observable outputs. It was revealed that easily controllable experimental variables, such as lateral cuff length and inflation rate, have a significant impact on cuff-induced vasculature changes. An interesting dependency between systemic BP and cuff-induced distal PTT change is also found, revealing opportunities for improved methods for BP surrogate calibration. However, validation via patient data shows that this relation does not hold for all patients, indicating required model improvements to be validated in follow up studies. These results provide promising directions to improve the calibration process featuring cuff inflation towards accurate and robust non-invasive blood pressure estimation.

A Multi-Parameter Fusion Method for Cuffless Continuous Blood Pressure Estimation Based on Electrocardiogram and Photoplethysmogram

Blood pressure (BP) is an essential physiological indicator to identify and determine health status. Compared with the isolated BP measurement conducted by traditional cuff approaches, cuffless BP monitoring can reflect the dynamic changes in BP values and is more helpful to evaluate the effectiveness of BP control. In this paper, we designed a wearable device for continuous physiological signal acquisition. Based on the collected electrocardiogram (ECG) and photoplethysmogram (PPG), we proposed a multi-parameter fusion method for noninvasive BP estimation. An amount of 25 features were extracted from processed waveforms and Gaussian copula mutual information (MI) was introduced to reduce feature redundancy. After feature selection, random forest (RF) was trained to realize systolic BP (SBP) and diastolic BP (DBP) estimation. Moreover, we used the records in public MIMIC-III as the training set and private data as the testing set to avoid data leakage. The mean absolute error (MAE) and standard deviation (STD) for SBP and DBP were reduced from 9.12 ± 9.83 mmHg and 8.31 ± 9.23 mmHg to 7.93 ± 9.12 mmHg and 7.63 ± 8.61 mmHg by feature selection. After calibration, the MAE was further reduced to 5.21 mmHg and 4.15 mmHg. The result showed that MI has great potential in feature selection during BP prediction and the proposed multi-parameter fusion method can be used for long-term BP monitoring.

A benchmark for machine-learning based non-invasive blood pressure estimation using photoplethysmogram

Blood Pressure (BP) is an important cardiovascular health indicator. BP is usually monitored non-invasively with a cuff-based device, which can be bulky and inconvenient. Thus, continuous and portable BP monitoring devices, such as those based on a photoplethysmography (PPG) waveform, are desirable. In particular, Machine Learning (ML) based BP estimation approaches have gained considerable attention as they have the potential to estimate intermittent or continuous BP with only a single PPG measurement. Over the last few years, many ML-based BP estimation approaches have been proposed with no agreement on their modeling methodology. To ease the model comparison, we designed a benchmark with four open datasets with shared preprocessing, the right validation strategy avoiding information shift and leak, and standard evaluation metrics. We also adapted Mean Absolute Scaled Error (MASE) to improve the interpretability of model evaluation, especially across different BP datasets. The proposed benchmark comes with open datasets and codes. We showcase its effectiveness by comparing 11 ML-based approaches of three different categories.

A Machine Learning Approach to the Non-Invasive Estimation of Continuous Blood Pressure Using Photoplethysmography

Blood pressure is an important vital sign that sometimes requires continuous measurement. The current methods include cuff measurements (manual auscultation and oscillometric techniques) for non-continuous measurement and invasive arterial cannulation for continuous measurement. The use of photoplethysmography as a cuffless, non-invasive, and continuous blood pressure measurement system is investigated through the use of four neural networks. These predict the systolic blood pressure, diastolic blood pressure, mean arterial blood pressure, and waveform shape. The models are trained on 890 h of data from 1669 patients in the MIMIC-III database. Feature-trained artificial neural networks predict the systolic blood pressure to 5.26 ± 6.53 mmHg (mean error ± standard deviation), the diastolic blood pressure to 2.96 ± 3.31 mmHg, and the mean arterial pressure to 3.27 ± 3.55 mmHg. These are used to shift and scale the predicted waveform, allowing the waveform prediction neural network to optimise for the wave shape rather than the amplitude. The waveform prediction has 86.4% correlation with the actual arterial blood pressure waveform. All results meet international clinical blood pressure measurement standards and could potentially change how blood pressure is measured in both clinical and research settings. However, more data from healthy individuals and analysis of the models’ biases based on clinical features is required.