Continuous Blood Pressure Monitoring Research Articles

Low Heart Rate Is Associated with Cerebral Pulsatility after TIA or Minor Stroke.

Beta-blockers are beneficial in coronary artery disease but less so in stroke prevention and dementia, potentially due to reduced heart rate (HR). Cerebral pulsatility is strongly associated with cerebral small vessel disease (SVD) and may be increased by lower diastolic pressures resulting from longer cardiac cycles. Patients 4-6 weeks after TIA or non-disabling stroke (Oxford Vascular Study) underwent 5 minutes continuous monitoring of blood pressure (BP), electrocardiogram (ECG), and middle cerebral artery flow velocity (transcranial ultrasound). Beat-to-beat relationships between HR, blood pressure and Gosling's pulsatility index (MCA-PI) are reported as beta-coefficients from general linear models for each individual. Across 759 patients, average MCA-PI during monitoring was associated with lower HR and diastolic BP (DBP) and greater systolic BP (SBP) (∆MCA-PI per 10 bpm/mmHg: -0.02, -0.04, 0.03, all p < 0.001), with HR particularly associated with low end-diastolic cerebral velocity (0.86, p=0.014). Beat-to-beat HR was strongly associated with concurrent low DBP and high SBP, potentially mediating the association with greater beat-to-beat cerebral pulsatility (average ∆MCA-PI vs HR/DBP/SBP unadjusted: -0.062, -0.052, 0.0092; adjusted for concurrent BP: -0.039, -0.11, 0.041). The beat-to-beat association between HR and MCA-PI increased with age, beta-blockers, arterial stiffness, low HR (age > 70 + HR < 65 vs age < 70 + HR > 65: -0.081 vs -0.024, interaction p < 0.001), and severe SVD on MRI (age > 70 + severe vs age < 70 + none: -0.087 vs -0.047, interaction p=0.03), with interactions between age, severe SVD, and low HR synergistically increasing MCA-PI. Low HR is associated with greater cerebral pulsatility in patients with SVD, potentially mediated by lower diastolic blood flow and representing a novel potential treatment target. ANN NEUROL 2022;92:909-920.

Recent Advances in Non-Invasive Blood Pressure Monitoring and Prediction Using a Machine Learning Approach.

Blood pressure (BP) monitoring can be performed either invasively via arterial catheterization or non-invasively through a cuff sphygmomanometer. However, for conscious individuals, traditional cuff-based BP monitoring devices are often uncomfortable, intermittent, and impractical for frequent measurements. Continuous and non-invasive BP (NIBP) monitoring is currently gaining attention in the human health monitoring area due to its promising potentials in assessing the health status of an individual, enabled by machine learning (ML), for various purposes such as early prediction of disease and intervention treatment. This review presents the development of a non-invasive BP measuring tool called sphygmomanometer in brief, summarizes state-of-the-art NIBP sensors, and identifies extended works on continuous NIBP monitoring using commercial devices. Moreover, the NIBP predictive techniques including pulse arrival time, pulse transit time, pulse wave velocity, and ML are elaborated on the basis of bio-signals acquisition from these sensors. Additionally, the different BP values (systolic BP, diastolic BP, mean arterial pressure) of the various ML models adopted in several reported studies are compared in terms of the international validation standards developed by the Advancement of Medical Instrumentation (AAMI) and the British Hypertension Society (BHS) for clinically-approved BP monitors. Finally, several challenges and possible solutions for the implementation and realization of continuous NIBP technology are addressed.

Remote multimodality monitoring of maternal physiology from the first trimester to postpartum period: study results.

Objectives:Current antenatal care largely relies on widely spaced appointments, hence only a fraction of the pregnancy period is subject to monitoring. Continuous monitoring of physiological parameters could represent a paradigm shift in obstetric care. Here, we analyse the data from daily home monitoring in pregnancy and consider the implications of this approach for tracking pregnancy health.Methods:Prospective feasibility study of continuous home monitoring of blood pressure, weight, heart rate, sleep and activity patterns from the first trimester to 6 weeks postpartum.Results:Fourteen out of 24 women completed the study (58%). Compared to early pregnancy [week 13, median heart rate (HR) 72/min, interquartile range (IQR) 12.8], heart rate increased by week 35 (HR 78/min, IQR 16.6; P = 0.041) and fell postpartum (HR 66/min, IQR 11.5, P = 0.021). Both systolic and diastolic blood pressure were lower at mid-gestation (week 20: SBP 103 mmHg, IQR 6.6; DPB 63 mmHg, IQR 5.3 P = 0.005 and P = 0.045, respectively) compared to early pregnancy (week 13, SBP 107 mmHg, IQR 12.4; DPB 67 mmHg, IQR 7.1). Weight increased during pregnancy between each time period analyzed, starting from week 15. Smartwatch recordings indicated that activity increased in the prepartum period, while deep sleep declined as pregnancy progressed.Conclusion:Home monitoring tracks individual physiological responses to pregnancy in high resolution that routine clinic visits cannot. Changes in the study protocol suggested by the study participants may improve compliance for future studies, which was particularly low in the postpartum period. Future work will investigate whether distinct adaptative patterns predate obstetric complications, or can predict long-term maternal cardiovascular health.

Plasma Biomarkers of Evolving Encephalopathy and Brain Injury in Neonates with Hypoxic-Ischemic Encephalopathy

The objective of the study was to evaluate the relationship between a panel of candidate plasma biomarkers and (1) death or severe brain injury on magnetic resonance imaging (MRI) and (2) dysfunctional cerebral pressure autoregulation as a measure of evolving encephalopathy. Neonates with moderate-to-severe hypoxic-ischemic encephalopathy (HIE) at 2 level IV neonatal intensive care units were enrolled into this observational study. Patients were treated with therapeutic hypothermia (TH) and monitored with continuous blood pressure monitoring and near-infrared spectroscopy. Cerebral pressure autoregulation was measured by the hemoglobin volume phase (HVP) index; a higher HVP index indicates poorer autoregulation. Serial blood samples were collected during TH and assayed for Tau, glial fibrillary acidic protein, and neurogranin. MRIs were assessed using National Institutes of Child Health and Human Development scores. The relationships between the candidate biomarkers and (1) death or severe brain injury on MRI (defined as a National Institutes of Child Health and Human Development score of ≥ 2B) and (2) autoregulation were evaluated using bivariate and adjusted logistic regression models. Sixty-two patients were included. Elevated Tau levels on days 2-3 of TH were associated with death or severe injury on MRI (aOR: 1.06, 95% CI: 1.03-1.09; aOR: 1.04, 95% CI: 1.01-1.06, respectively). Higher Tau was also associated with poorer autoregulation (higher HVP index) on the same day (P=.022). Elevated plasma levels of Tau are associated with death or severe brain injury by MRI and dysfunctional cerebral autoregulation in neonates with HIE. Larger-scale validation of Tau as a biomarker of brain injury in neonates with HIE is warranted.

Feasibility of Blood Pressure Measurement With a Wearable (Watch-Type) Monitor During Impending Syncopal Episodes.

BackgroundWe assessed the reliability and feasibility of blood pressure (BP) measurements by means of a new wearable watch‐type BP monitor (HeartGuide) in detecting episodes of hypotensive (pre)syncope induced by tilt table test.Methods and ResultsAn intrapatient comparison between systolic BP (SBP) measured by means of the HeartGuide device and noninvasive finger beat‐to‐beat BP monitoring was undertaken both at baseline in supine position and repeatedly during tilt table test in patients evaluated for reflex syncope. Intrapatient fall of systolic BP from baseline was measured. Eighty‐one patients (mean age, 61±19 years; 46 women) were included. Overall, HeartGuide was able to yield BP values at the time of BP nadir in 58 (72%) patients (average HeartGuide SBP 102±18 mm Hg, versus finger SBP 101±19 mm Hg). Compared with baseline, the maximum SBP decrease was on average −28.5±27.8 and −30.3±33.9 mm Hg respectively (Lin's concordance correlation coefficient=0.78, r=0.79, P=0.001). In the subgroup of 38 patients with tilt table test induced (pre)syncope, the average HeartGuide SBP during symptoms was 97±16 mm Hg, and the finger SBP was 94±18 mm Hg. Compared with baseline, the maximum SBP decrease was on average −35.2±29.3 and −43.3±31.8 mm Hg, respectively (Lin's concordance correlation coefficient=0.83, r=0.87, P=0.001).ConclusionsOur data indicate that the HeartGuide BP monitor can detect low BP during presyncope and that its measure of SBP change is consistent with that simultaneously obtained through continuous BP monitoring, despite some intrapatient variability. Thus, this device might be useful in determining the hypotensive nature of spontaneous (pre)syncopal symptoms, a possibility that should be verified by field studies.

A Computational Modeling and Simulation Workflow to Investigate the Impact of Patient-Specific and Device Factors on Hemodynamic Measurements from Non-Invasive Photoplethysmography.

Cardiovascular disease is the leading cause of death globally. To provide continuous monitoring of blood pressure (BP), a parameter which has shown to improve health outcomes when monitored closely, many groups are trying to measure blood pressure via noninvasive photoplethysmography (PPG). However, the PPG waveform is subject to variation as a function of patient-specific and device factors and thus a platform to enable the evaluation of these factors on the PPG waveform and subsequent hemodynamic parameter prediction would enable device development. Here, we present a computational workflow that combines Monte Carlo modeling (MC), gaussian combination, and additive noise to create synthetic dataset of volar fingertip PPG waveforms representative of a diverse cohort. First, MC is used to determine PPG amplitude across age, skin tone, and device wavelength. Then, gaussian combination generates accurate PPG waveforms, and signal processing enables data filtration and feature extraction. We improve the limitations of current synthetic PPG frameworks by enabling inclusion of physiological and anatomical effects from body site, skin tone, and age. We then show how the datasets can be used to examine effects of device characteristics such as wavelength, analog to digital converter specifications, filtering method, and feature extraction. Lastly, we demonstrate the use of this framework to show the insensitivity of a support vector machine predictive algorithm compared to a neural network and bagged trees algorithm.

Comparative Study of Prediction of IDBP values for Hemodialysis using Deep Learning

Abstract: The field of medicine is expanding rapidly as new diseases appear regularly, necessitating the development of appropriate treatment options. A precise and efficient method of operation is necessary for correct diagnosis and treatment. Blood pressure (BP) is a vital sign that provides basic information about patients' health. During the clinical operation of hemodialysis, blood pressure (BP) variability affects significant global risks and secondary complications associated with adverse mortality. In patients with hypertension, continuous BP monitoring is important. If the scheme is automated, it can be very useful. Consequently, the implementation of an effective automatic medical diagnostic scheme could be very beneficial for all stratifications involved in this process.

Noninvasive, Wearable Multi Biosensors for Continuous, Long-term Monitoring of Blood Pressure via Internet of Things Applications

Continuous blood pressure (BP) monitoring through various physical stress tests is important for the prediction of hypertension and cardiovascular mortality. This study proposes a multimodal wearable biosensor device for continuous BP monitoring through various tests during exercise and internet of things (IoT) applications. We evaluated three pulse transit time (PTT)-BP estimation models to identify the best mode for predicting BP response to exercise. Moreover, the proposed BP monitoring system was designed to connect to the cloud server for health data processing and management. The efficacy of the proposed device was validated by monitoring the natural variability of BP in various cycling tests. Thus, BP estimation based on the impedance plethysmography waveform and arterial impedance offers superiority compared to other conventional PTT estimation methods. These results suggest the feasibility of the multimodal biosensor device for continuous, long-term BP monitoring in remote diagnosis and management with IoT-based healthcare applications.

Oscillometric Waveform Evaluation for Blood Pressure Devices

Most of the noninvasive blood pressure (NIBP) device algorithms have been derived from the clinical trials at various hospitals and human races in developed countries. It is urgent to develop an algorithm for NIBP devices for Indian citizens. In this regard, we have developed a flexible pressure sensor to record the human body's oscillometric waveform (OMW). As we know, assessing human blood pressure (BP) by noninvasive techniques is virtually a revolutionary step in biomedical science and engineering. High blood pressure is one of the most critical risk factors for heart disease and death; therefore, continuous monitoring of BP is necessary. We have obtained the BP pulse waveform by the flexible capacitive pressure sensor and further estimated the systolic blood pressure (SBP), diastolic blood pressure (DBP), and mean blood pressure (MBP) using international guidelines. In this work, the BP simulation was performed on 64 waveforms, combinations of SBP (80–180 mmHg) and DBP (60–170 mmHg) to compute the BP in different health conditions such as hypotension, normal, and hypertension. In addition, an oscillometric waveform was also obtained from the brachial artery of a healthy person and analyzed for its utility as an NIBP waveform recorder. The SBP, DBP, and MBP were calculated from the oscillometric waveform envelope (OMWE) using the maximum amplitude algorithm (MAA) technique. In-house developed capacitive pressure sensors give comparable results when tested against the NIBP simulator and the commercial BP monitor. The pulse rate is also computed using the same OMW. The mean and standard deviation of the error in SBP, DBP, and MBP are within the universal standards' validation protocol for BP devices. The results indicate that the flexible capacitive pressure sensor can measure the BP with acceptable error and replace the commercial BP monitor and recorder sensor. The developed sensors show potential for large-scale clinical trials for blood pressure and pulse rate measurements.

The Effect of Intermittent versus Continuous Non-Invasive Blood Pressure Monitoring on the Detection of Intraoperative Hypotension, a Sub-Study.

Intraoperative hypotension is associated with postoperative complications. However, in the majority of surgical patients, blood pressure (BP) is measured intermittently with a non-invasive cuff around the upper arm (NIBP-arm). We hypothesized that NIBP-arm, compared with a non-invasive continuous alternative, would result in missed events and in delayed recognition of hypotensive events. This was a sub-study of a previously published cohort study in adult patients undergoing surgery. The detection of hypotension (mean arterial pressure below 65 mmHg) was compared using two non-invasive methods; intermittent oscillometric NIBP-arm versus continuous NIBP measured with a finger cuff (cNIBP-finger) (Nexfin, Edwards Lifesciences). cNIBP-finger was used as the reference standard. Out of 350 patients, 268 patients (77%) had one or more hypotensive events during surgery. Out of the 286 patients, 72 (27%) had one or more missed hypotensive events. The majority of hypotensive events (92%) were detected with NIBP-arm, but were recognized at a median of 1.2 (0.6–2.2) minutes later. Intermittent BP monitoring resulted in missed hypotensive events and the hypotensive events that were detected were recognized with a delay. This study highlights the advantage of continuous monitoring. Future studies are needed to understand the effect on patient outcomes.

A Novel CNN-LSTM Model Based Non-Invasive Cuff-Less Blood Pressure Estimation System.

PPG (Photoplethysmography) and ECG (Electro-cardiogram) physiological signals have been known to have certain indicators for establishing blood pressure (BP) levels. Continuous monitoring of blood pressure (BP) is highly valuable for cardiovascular patients; however the existing non-invasive cuff-based blood pressure monitoring system is discreet and applies artificial pressure on patients' arms that is uncomfortable. The other invasive method is highly interventional in nature and is highly disturbing when the patient is recuperating in the hospital wards or elsewhere. A non-invasive cuff-less, non-disturbing, and continuous BP measurement system targeted toward surgical, clinical, and domestic usage are proposed in this work. A convolutional neural network (CNN) followed by a long short-term memory layer (LSTM) was designed and applied to ECG and PPG signals to present accurate systolic blood pressure (SBP), and diastolic blood pressure (DBP). For developing the CNN-LSTM layers, a novel and open-source dataset was compiled that consisted of PPG and ECG signals from 30 healthy participants and is made publicly available for further usage to the research community. The novel CNN-LSTM based cuff-less blood pressure evaluation system presented a mean absolute error (MAE) of 2.57 mmHg and 3.44 mmHg for SBP and DBP respectively with similar standard-deviation (SD) metrics. The characterized error metrics of the proposed system are the lowest to date when compared to other prior work. Clinical Relevance- A cuff-less BP estimation system allows patients to have easy access to blood pressure evaluation as well as aid in determining unsafe health ailments like hypertension. Ready access to such system will not only allow practitioners to continuously monitor BP in hospitals but also help patients to regularly monitor BP more frequently at their convenience.

Novel Blood Pressure Waveform Reconstruction from Photoplethysmography using Cycle Generative Adversarial Networks.

Continuous monitoring of blood pressure (BP) can help individuals manage their chronic diseases such as hypertension, requiring non-invasive measurement methods in free-living conditions. Recent approaches fuse Photoplethys-mograph (PPG) and electrocardiographic (ECG) signals using different machine and deep learning approaches to non-invasively estimate BP; however, they fail to reconstruct the complete signal, leading to less accurate models. In this paper, we propose a cycle generative adversarial network (CycleGAN) based approach to extract a BP signal known as ambulatory blood pressure (ABP) from a clean PPG signal. Our approach uses a cycle generative adversarial network that extends the GAN architecture for domain translation, and outperforms state-of-the-art approaches by up to 2× in BP estimation.

Waveform Morphology Comparison in Wearable Blood Pressure Sensors.

Wearable devices for continuous non-invasive blood pressure monitoring must be capable of providing a continuous waveform representative of arterial blood pressure. This paper establishes the distinctions in waveform morphology between wearable sensor modalities, specifically millimeter-wave radar and photoplethysmography, when compared to a reference continuous non-invasive blood pressure monitor. An analysis of a 115-subject dataset was conducted to assess waveform suitability. Millimeter-wave radar waveform morphology was found to more closely resemble continuous non-invasive blood pressure than photoplethysmography. Clinical Relevance- This paper compares the waveform morphology and content of signals from wearable sensors in the context of continuous non-invasive blood pressure monitoring.

Using Bayesian Optimization and Wavelet Decomposition in GPU for Arterial Blood Pressure Estimation.

Continuous monitoring of arterial blood pressure (ABP) of patients in hospital is currently carried out in an invasive way, which could represent a risk for them. In this paper, a noninvasive methodology to optimize ABP estimators using electrocardiogram and photoplethysmography signals is proposed. For this, the XGBoost machine learning model, optimized with Bayesian techniques, is executed in a Graphics Processing Unit, which drastically reduces execution time. The methodology is evaluated using the MIMIC-III Waveform Database. Systolic and diastolic pressures are estimated with mean absolute error values of 15.85 and 11.59 mmHg, respectively, similar to those of the state of the art. The main advantage of the proposed methodology with respect to others of the current state of the art is that it allows the optimization of the estimator model to be performed automatically and more efficiently at the computational level for the data available. Clinical Relevance- This approach has the advantage of using noninvasive methods to continuously monitor patient's arterial blood pressure, reducing the risk for patients.

Blood Pressure Estimation using a Single Channel Bio-Impedance Ring Sensor.

The demand for non-obtrusive, accurate, and continuous blood pressure (BP) monitoring systems is becoming more prevalent with the realization of its significance in preventable cardiovascular disease (CVD) globally. Current cuff-based standards are bulky, uncomfortable, and are limited to discrete recording periods. Wearable sensor technologies such as those using optical photoplethysmography (PPG) have been used to develop blood pressure estimation models through a variety of methods. However, this technology falls short as optical based systems have bias favoring lighter skin tones and lower body fat compositions. Bioimpedance (Bio-Z) is a capable modality of sensing arterial blood flow without implicit inadvertent bias towards individuals. In this paper we propose a ring-based bioimpedance system to capture arterial blood flow from the digital artery of the finger. The ring design provides a more compact wearable device utilizing only a single Bio-Z channel, making it a familiar fit to individuals. Post-processing the acquired Bio-Z signals, we extracted 9 frequency domain features from windowed beat cycles to train subject specific regression models. Results indicate the average mean absolute errors for systolic/diastolic BP to be 4.38/3.63mmHg, consistent with AAMI standards.

Cuff-less and continuous blood pressure measurement based on pulse transit time from carotid and toe photoplethysmograms

Blood pressure (BP) is a vital health parameter that varies throughout the day. As a single reading of high BP may not indicate hypertension, continuous monitoring of BP is usually recommended by medical doctors to confirm the diagnosis of hypertension. In the last few decades, researchers have investigated cuff-less and continuous BP measurements based on pulse transit time (PTT). The main purpose of this research is to develop an autoregressive (ARX) system identification (SI)-based PTT calculation model using two PPG signals acquired from carotid and toe. The signals were recorded from 65 subjects with an age range between 20 and 60 years. The results of the study have been validated in two stages. The first validation comprised the estimated BP from PTT using SI compared to the measured BP using the cuff-based method for all subjects. The results of the estimated BP using the proposed method compared to the measured BP obtained using the standard BP cuff measurement method are highly correlated to both systolic blood pressure (R 2 = 0.8132) and diastolic blood pressure (R 2 = 0.8357). The second validation consisted of comparing PTT values using system identification to the results of the PTT derived from the ECG–PPG method. The results showed that both methods are highly correlated (R 2 = 0.7808), and there is no significant difference between them (p < 0.05) with a slightly better PTT estimation related to DBP in the proposed method. Our results have proven that the PTT obtained from the carotid PPG and toe PPG using the system identification approach yielded SBP and DBP estimations that are consistent with the values of the conventional BP cuff method. The newly proposed method has the advantage of being cuff-less and able to provide continuous BP measurements.

Non-invasive continuous blood pressure monitoring using the ClearSight system for pregnant women at high risks of post-partum hemorrhage: comparison with invasive blood pressure monitoring during cesarean section.

ObjectiveThis study aimed to investigate the accuracy and precision of continuous, non-invasive blood pressure obtained using the ClearSight system by comparing it with invasive arterial blood pressure, and to assess the hemodynamic changes using invasive methods and the ClearSight system in patients undergoing cesarean section.MethodsArterial pressure was measured invasively with an intra-arterial catheter and non-invasively using the ClearSight system during cesarean section in patients with placenta previa or placenta accreta. Blood pressure measurements obtained using these two means were then compared.ResultsTotal 1,277 blood pressure measurement pairs were collected from 21 patients. Under Bland-Altman analysis, the ClearSight system demonstrated an acceptable accuracy with a bias and standard deviation of 8.8±13.4 mmHg for systolic blood pressure, −6.3±7.1 mmHg for diastolic blood pressure, and −2.7±8.0 mmHg for median blood pressure. Cardiac index levels were significantly elevated during fetal delivery and 5 minutes after placental removal, and systemic vascular resistance index levels were significantly decreased during fetal delivery and 40 minutes after placental removal.ConclusionIn patients undergoing cesarean section, the ClearSight system showed excellent accuracy and precision compared to that of the currently used invasive monitoring system.

Continuous cuffless monitoring of arterial blood pressure via graphene bioimpedance tattoos.

Continuous monitoring of arterial blood pressure (BP) in non-clinical (ambulatory) settings is essential for understanding numerous health conditions, including cardiovascular diseases. Besides their importance in medical diagnosis, ambulatory BP monitoring platforms can advance disease correlation with individual behaviour, daily habits and lifestyle, potentially enabling analysis of root causes, prognosis and disease prevention. Although conventional ambulatory BP devices exist, they are uncomfortable, bulky and intrusive. Here we introduce a wearable continuous BP monitoring platform that is based on electrical bioimpedance and leverages atomically thin, self-adhesive, lightweight and unobtrusive graphene electronic tattoos as human bioelectronic interfaces. The graphene electronic tattoos are used to monitor arterial BP for >300 min, a period tenfold longer than reported in previous studies. The BP is recorded continuously and non-invasively, with an accuracy of 0.2 ± 4.5 mm Hg for diastolic pressures and 0.2 ± 5.8 mm Hg for systolic pressures, a performance equivalent to Grade A classification.

PPG-based blood pressure estimation can benefit from scalable multi-scale fusion neural networks and multi-task learning

The continuous measurement of blood pressure (BP) plays an important role in preventing cardiovascular diseases. However, common cuff-based devices with cumbersome operations are difficult to realize continuous BP monitoring. We propose a continuous BP estimation method based on the end-to-end multi-scale fusion and multi-task learning neural network (MSF-MTLNet). It can use the 8-second PPG segment and its derivatives as input and automatically learn their multi-scale features to improve BP accuracy. The proposed method implements feature extraction by adaptively fusing multiple convolution kernels with different sizes through the attention mechanism. Simultaneously, multi-task learning was introduced to discover task-relevant features. A total of 277,600 segments from 1825 patients in the University of California, Irvine database are used for experiments. They are divided into the independent training set, validation set, and test set according to the ratio of 8:1:1 in an inter-subject manner. The mean and standard deviation of estimation error for systolic blood pressure and diastolic blood pressure are 0.97 ± 8.87 mmHg and 0.55 ± 4.23 mmHg, respectively, which outperforms the current cutting-edge method. Our approach almost meets the Association for the Advancement of Medical Instrumentation standard. Considering the British Hypertension Society standards, it achieves an 'A' grade for DBP estimation and a 'B' grade for SBP estimation. Experimental results show that MSF-MTLNet can obtain promising precision in cuffless and continuous BP estimation.

Formula omitted]: Cuff-less and non-invasive blood pressure estimation via a generic deep convolutional architecture

Objective:The paper focuses on development of robust and accurate processing solutions for continuous and cuff-less blood pressure (BP) monitoring. In this regard, a robust deep learning-based framework is proposed for computation of low latency and continuous upper and lower bounds on the systolic and diastolic BP. Methods:Referred to as the BP-Net, the proposed framework is a novel convolutional architecture that provides longer effective memory while achieving superior performance due to incorporation of casual dialated convolutions and residual connections. To utilize the real potential of deep learning in extraction of intrinsic features (deep features) and enhance the long-term robustness, the BP-Net uses raw Electrocardiograph (ECG) and Photoplethysmograph (PPG) signals without extraction of any form of hand-crafted features as it is common in existing solutions. Results:By capitalizing on the fact that datasets used in recent literature are not unified and properly defined, a benchmark dataset is constructed from the MIMIC-I and MIMIC-III databases obtained from PhysioNet. The proposed BP-Net is evaluated based on this benchmark dataset demonstrating promising performance and shows superior generalizable capacity. Conclusion:The proposed BP-Net architecture is more accurate than canonical recurrent networks and enhances the long-term robustness of the BP estimation task. Significance:The proposed BP-Net architecture addresses key drawbacks of existing BP estimation solutions, i.e., relying heavily on extraction of hand-crafted features, such as pulse arrival time (PAT), and; Lack of robustness. Finally, the constructed BP-Net dataset provides a unified base for evaluation and comparison of deep learning-based BP estimation algorithms.