Search

Abstract Translating vascular aging research from bench to bedside presents both significant opportunities and challenges. This paper summarizes insights from a roundtable discussion at the Artery 23 conference, featuring perspectives from basic science, clinical trials, regulation, and industry. The main conclusions of the discussion are as follows: basic science research must align with clinical relevance, using appropriate animal models and standardized measurement techniques. Pragmatic and registry-based clinical trials offer viable alternatives to traditional randomized controlled trials, facilitating real-world applicability. The regulatory landscape, particularly for software medical devices, must evolve to keep pace with technological advancements like artificial intelligence. Industry efforts focus on developing devices or solutions for vascular aging assessment and treatment strategies, yet face hurdles in large-scale adoption and reimbursement. Despite significant progress, the development of pharmacological interventions to mitigate vascular aging remains a critical need. This discussion underscores the importance of interdisciplinary collaboration to overcome barriers and translate scientific discoveries into clinical practice effectively.

Arterial compliance (AC) plays a crucial role in vascular aging and cardiovascular disease. The ability to continuously estimate aortic AC or its surrogate, pulse pressure (PP), through wearable devices is highly desirable, given its strong association with daily activities. While the single-site photoplethysmography (PPG)-derived arterial stiffness indices show reasonable correlations with AC, they are susceptible to noise interference, limiting their practical use. To overcome this challenge, our study introduces a noise-resistant indicator of AC: Katz’s fractal dimension (KFD) of PPG signals. We showed that KFD integrated the signal complexity arising from compliance changes across a cardiac cycle and vascular structural complexity, thereby decreasing its dependence on individual characteristic points. To assess its capability in measuring AC, we conducted a comprehensive evaluation using both in silico studies with 4374 virtual human data and real-world measurements. In the virtual human studies, KFD demonstrated a strong correlation with AC (r = 0.75), which only experienced a slight decrease to 0.66 at a signal-to-noise ratio of 15dB, surpassing the best PPG-morphology-derived AC measure (r = 0.41) under the same noise condition. In addition, we observed that KFD’s sensitivity to AC varied based on the individual’s hemodynamic status, which may further enhance the accuracy of AC estimations. These in silico findings were supported by real-world measurements encompassing diverse health conditions. In conclusion, our study suggests that PPG-derived KFD has the potential to continuously and reliably monitor arterial compliance, enabling unobtrusive and wearable assessment of cardiovascular health.

The Fontan circulation, designed for managing patients with a single functional ventricle, presents challenges in long-term outcomes. Computational methods offer potential solutions, yet their application in cardiology practice remains largely unexplored. Our aim was to assess the ability of a patient-specific, closed-loop, reduced-order blood flow model to simulate pulsatile blood flow in the Fontan circulation. Using one-dimensional models, we simulated the aorta, superior and inferior venae cavae, and right and left pulmonary arteries, while lumping heart chambers and remaining vessels into zero-dimensional models. The model was calibrated with patient-specific haemodynamic data from combined cardiac catheterisation and magnetic resonance exams, using a novel physics-based stepwise methodology involving simpler open-loop models. Testing on a 10-year-old, anesthetised patient, demonstrated the model's capability to replicate pulsatile pressure and flow in the larger vessels and ventricular pressure. Average relative errors in mean pressure and flow were 2.9 % and 3.6 %, with average relative point-to-point errors (RPPE) in pressure and flow at 5.2 % and 16.0 %. Comparing simulation results to measurements, mean aortic pressure and flow values were 50.7 vs. 50.4 mmHg and 41.6 vs. 41.9 ml/s, respectively, while ventricular pressure values were 28.7 vs. 27.4 mmHg. The model accurately described time-varying ventricular volume with a RPPE of 2.9 %, with mean, minimum, and maximum ventricular volume values for simulation results vs. measurements at 59.2 vs. 58.2 ml, 38.0 vs. 37.6 ml, and 76.0 vs. 74.4 ml, respectively. It provided physiologically realistic predictions of haemodynamic changes from pulmonary vasodilation and atrial fenestration opening. The new model and calibration methodology are freely available, offering a platform to virtually investigate the Fontan circulation's response to clinical interventions and explore potential mechanisms of Fontan failure. Future efforts will concentrate on broadening the model's applicability to a wider range of patient populations and clinical scenarios, as well as testing its effectiveness.

BackgroundArterial stiffening and peripheral wave reflections have been considered the major determinants of raised pulse pressure (PP) and isolated systolic hypertension, but the importance of cardiac contractility and ventricular ejection dynamics is also recognised.MethodsWe examined the contributions of arterial compliance and ventricular contractility to variations in aortic flow and increased central (cPP) and peripheral (pPP) pulse pressure, and PP amplification (PPa) in normotensive subjects during pharmacological modulation of physiology, in hypertensive subjects, and in silico using a cardiovascular model accounting for ventricular–aortic coupling. Reflections at the aortic root and from downstream vessels were quantified using emission and reflection coefficients, respectively.ResultscPP was strongly associated with contractility and compliance, whereas pPP and PPa were strongly associated with contractility. Increased contractility by inotropic stimulation increased peak aortic flow (323.9 ± 52.8 vs. 389.1 ± 65.1 ml/s), and the rate of increase (3193.6 ± 793.0 vs. 4848.3 ± 450.4 ml/s2) in aortic flow, leading to larger cPP (36.1 ± 8.8 vs. 59.0 ± 10.8 mmHg), pPP (56.9 ± 13.1 vs. 93.0 ± 17.0 mmHg) and PPa (20.8 ± 4.8 vs. 34.0 ± 7.3 mmHg). Increased compliance by vasodilation decreased cPP (62.2 ± 20.2 vs. 45.2 ± 17.8 mmHg) without altering dP/dt, pPP or PPa. The emission coefficient changed with increasing cPP, but the reflection coefficient did not. These results agreed with in silico data obtained by independently changing contractility/compliance over the range observed in vivo.ConclusionsVentricular contractility plays a key role in raising and amplifying PP, by altering aortic flow wave morphology.

Magnetic resonance imaging and computed tomography allow the characterization of arterial state and function with high confidence and thus play a key role in the understanding of arterial aging and its translation into the clinic. Decades of research into the development of innovative imaging sequences and image analysis techniques have led to the identification of a large number of potential biomarkers, some bringing improvement in basic science, others in clinical practice. Nonetheless, the complexity of some of these biomarkers and the image analysis techniques required for their computation hamper their widespread use. In this narrative review, current biomarkers related to aging of the aorta, their founding principles, the sequence, and postprocessing required, and their predictive values for cardiovascular events are summarized. For each biomarker a summary of reference values and reproducibility studies and limitations is provided. The present review, developed in the COST Action VascAgeNet, aims to guide clinicians and technical researchers in the critical understanding of the possibilities offered by these advanced imaging modalities for studying the state and function of the aorta, and their possible clinically relevant relationships with aging.

Several studies suggest that central (aortic) blood pressure (cBP) is a better marker of cardiovascular disease risk than peripheral blood pressure (pBP). The morphology of the pBP wave, usually assessed non-invasively in the arm, differs significantly from the cBP wave, whose direct measurement is highly invasive. In particular, pulse pressure, PP (the amplitude of the pressure wave), increases from central to peripheral arteries, leading to the so-called pulse pressure amplification (ΔPP). The main purpose of this study was to develop a methodology for estimating central PP (cPP) from non-invasive measurements of aortic flow and peripheral PP. Our novel approach is based on a comprehensive understanding of the main cardiovascular properties that determine ΔPP along the aortic-brachial arterial path, namely brachial flow wave morphology in late systole, and vessel radius and distance along this arterial path. This understanding was achieved by using a blood flow model which allows for workable analytical solutions in the frequency domain that can be decoupled and simplified for each arterial segment. Results show the ability of our methodology to (i) capture changes in cPP and ΔPP produced by variations in cardiovascular properties and (ii) estimate cPP with mean differences smaller than 3.3 ± 2.8 mmHg on in silico data for different age groups (25–75 years old) and 5.1 ± 6.9 mmHg on in vivo data for normotensive and hypertensive subjects. Our approach could improve cardiovascular function assessment in clinical cohorts for which aortic flow wave data is available.

One of the European gold standard measurement of vascular ageing, a risk factor for cardiovascular disease, is the carotid-femoral pulse wave velocity (cfPWV), which requires an experienced operator to measure pulse waves at two sites. In this work, two machine learning pipelines were proposed to estimate cfPWV from the peripheral pulse wave measured at a single site, the radial pressure wave measured by applanation tonometry. The study populations were the Twins UK cohort containing 3,082 subjects aged from 18 to 110 years, and a database containing 4,374 virtual subjects aged from 25 to 75 years. The first pipeline uses Gaussian process regression to estimate cfPWV from features extracted from the radial pressure wave using pulse wave analysis. The mean difference and upper and lower limits of agreement (LOA) of the estimation on the 924 hold-out test subjects from the Twins UK cohort were 0.2 m/s, and 3.75 m/s & -3.34 m/s, respectively. The second pipeline uses a recurrent neural network (RNN) to estimate cfPWV from the entire radial pressure wave. The mean difference and upper and lower LOA of the estimation on the 924 hold-out test subjects from the Twins UK cohort were 0.05 m/s, and 3.21 m/s & -3.11m/s, respectively. The percentage error of the RNN estimates on the virtual subjects increased by less than 2% when adding 20% of random noise to the pressure waveform. These results show the possibility of assessing the vascular ageing using a single peripheral pulse wave (e.g. the radial pressure wave), instead of cfPWV. The proposed code for the machine learning pipelines is available from the following online depository (https://github.com/WeiweiJin/Estimate-Cardiovascular-Risk-from-Pulse-Wave-Signal).

Objective and Motivation: Pulse wave velocity (PWV) is known to be associated with vascular ageing, a risk factor for cardiovascular disease (CVD) [1]. The European gold standard measurement of PWV requires an experienced operator to measure pulse waveforms at multiple sites, sometimes together with an electrocardiogram [2,3]. This study aims to estimate PWV from the radial pulse waveform using machine learning. Methods: Radial pulse waveforms and carotid-femoral PWVs were acquired in 3,082 unselected twins (https://twinsuk.ac.uk). 14 fiducial points on each pulse waveform were extracted using an in-house algorithm [4]. LASSO regression and principal component analysis (PCA) were used to identify the key features (timing and magnitude of the fiducial points) associated with PWV and exclude outliers. Finally, Gaussian process regression was used to estimate the PWV based on those key features only. Results: Results show that PWV can be estimated from the radial pulse waveform only with an overall root mean squared error (RMSE) of 1.82 m/s (Figure A). Most of the measured PWV values were within the 95% confidence interval range of the estimated PWV. The difference between measured and estimated PWV values increased with the increasing PWV. PWV estimation on a subgroup of twins with a healthy range of blood pressure and PWV values [5] was achieved with a RMSE of 1.38 m/s (Figure B). Conclusion: In this proof-of-concept study we have shown the possibility of estimating PWV from the radial pulse waveform using machine learning. This approach could make CVD detection more accessible to the wider population.

Background: Abdominal aortic aneurysm (AAA) is usually asymptomatic and has an extremely high mortality if rupture occurs. Therefore, early detection and intervention are important. However, AAA is most often detected as an accidental finding during clinical imaging for other purposes [1]. Considering that AAA has a systemic impact on the biophysical properties of the cardiovascular system, pulse-wave-based diagnosis of AAA may be a potential approach for effective detection. There have been only a few studies of pulse wave propagation with AAA by using computational modelling or hydraulic simulators [2]. This study aims to provide some basic insights for pulse-wave-based diagnosis of AAA using computational modelling. Methods: We simulated blood flow in the larger systemic arteries using the 65-year-old baseline model from the pulse wave database in Charlton et al. [3] (see Figure, left). The influence on pulse waveforms of AAA morphology (including shape, maximum diameter and length) was simulated by adjusting the geometry of the abdominal aorta IV. In addition, the influence of the local AAA stiffness and global arterial stiffness (represented by carotid-femoral pulse wave velocity) was investigated by varying the relevant model parameters. Results: Results show that maximum diameter is the dominant morphological factor in reshaping the pulse waveform, introducing considerable changes in pulse wave indices. Moreover, both local and global arterial stiffness have a considerable impact on wave morphology in the presence of AAA (Figure, right). Conclusion: AAA produces considerable changes in pulse wave morphology that could be used for AAA detection by using pulse wave analysis.

Abstract Aortic tapering is a known characteristic of the arterial tree affecting the development of pressure in the aorta. With tapering, the cross-sectional area of vessels decreases moving towards the periphery causing reflections to travel back to the heart. The reflection waves present in the aorta are an amalgamation of reflections from tapering, bifurcations, and mismatched mechanical properties. The difference between the contribution of tapering and of bifurcations or mechanical properties in the overall reflection is not well understood. The aim of this study is to evaluate the impact of tapering, solely, on pressure and wave intensity. We simulated the blood flow using a one-dimensional model which solves the conservation of mass and momentum equations in the time domain. The model consists of the thoracic aorta connected to a symmetrical iliac bifurcation and three-element Windkessel models as the terminal boundary conditions [1]. Different tapering angles (θ) in the physiological range, up to 1.5 degrees [2], are implemented by changing the diameter at the outlet of the aortic vessel. The area ratio between the parent and the daughter vessels and the mechanical properties are kept constant which allows for investigating the effect of tapering. As shown in Figure 1 (middle), increased tapering angle is associated with an increase in mean blood pressure due to the higher resistance in the segment, and an increase in pulse pressure due to the reduced arterial compliance. Both forward and backward compression wave amplitudes rise with the tapering angle due to higher resistance of the arterial vessel. Figure 1 Schematic representation of the thoracic aorta and iliac bifurcation model (left), pressure waveforms at the inlet of the thoracic aorta (middle), and wave intensity analysis at the inlet of the thoracic aorta (right) with different tapering angles.

The influence of arterial and ventricular parameters on the main fiducial pressure points and index during systole has been investigated using a mix of in silico and in vivo data. Notably, an index, QIx, based entirely on ventricular ejection patterns has been developed and its potential in describing the augmentation pressure index, AIx, has been investigated and compared against the reflection coefficient Г. In a group of hypertensive patients (n = 156) and in healthy volunteers whose physiology was modulated via administration of vasoactive drugs (n = 13), QIx was found to be more correlated to AIx than Г (coefficient of determination, R2 = 0.71 vs R2 = 0.53 in the normotensive group; R2 = 0.52 vs R2 = 0.37 in the hypertensive group). This trend was confirmed by a LASSO analysis on in vivo data (standardised coefficient, β = 0.42 vs β = 0.22, p < 0.001 in both cases) and also observed in an in silicosensitivity analysis on a reduced 7-artery model of the upper aortic arch (for variations between –25% and +25% of QIx and Г from baseline values, AIx varied between −245% and +193%, −194% and +191% respectively) (Figure 1). All analyses were also concordant for the main determinants of the first systolic shoulder P1 and augmentation pressure AP: pulse wave velocity (β = 0.74, p < 0.001) and peak flow velocity (β = 0.41, p < 0.001) are the main determinants of P1 while reflection waves (β = 0.34, p < 0.001) and stroke volume (β = 0.37, p < 0.001) mainly define AP (Figure 1). These results further strengthen the case that ventricular dynamics is at least as important as arterial stiffening in raised pulse pressure. Figure 1

Purpose: Vascular age may be a better marker of cardiovascular risk than age. It has been proposed that vascular age could be assessed easily and inexpensively from the photoplethysmogram (PPG) pulse wave, which is measured by pulse oximeters and fitness bands. Our aim was to assess the performance of existing PPG-derived indices and provide directions for future research. Methods: Two datasets were used. The Airwave dataset [1] contains finger PPG pulse waves alongside brachial-femoral pulse wave velocities (PWVs) from police employees: high quality data from 4,378 subjects aged 20 to 66 were used in this study. The arterial pulse wave database [2] contains simulated PPG pulse waves for 3,837 virtual subjects aged 25 to 75 exhibiting healthy ageing and normal variation in cardiovascular properties. Pulse wave features were derived, and the following indices were assessed: the ageing index (AGI) [3], modified AGI [4], informal AGI [5], and a novel AGI obtained by fitting a linear regression model to the five features most strongly correlated with age. Results: See Figure 1. Existing indices correlated weakly with age. The new AGI was more strongly correlated with age on both datasets. It was weakly correlated with PWV and BP on Airwave, but strongly correlated with PWV on simulated data. Figure 1 Conclusion: The improved performance of the new AGI indicates that there is scope for improving PPG-derived indices of vascular ageing. Indices appear to be correlated with age, PWV and BP, indicating that with further research they could be useful markers of cardiovascular risk.

Objective and Motivation: Endothelial dysfunction is thought to underpin atherosclerotic cardiovascular diseases. However, the widely used flow mediated dilation (FMD) in-vivotest of endothelial function relates poorly to clinical outcomes. In this study, we investigated potential confounding factors affecting the FMD test that could prevent FMD from evaluating true endothelial cell function. Methods: Haemodynamics in the larger 116 systemic arteries was simulated using one-dimensional modelling [1] coupled to an endothelial function model relating changes in wall shear stress to changes in local vascular stiffness. Haemodynamics during cuff inflation and deflation were simulated by prescribing time-varying i) transmural pressure changes in the right radial and ulnar 1-D model arterial segments, and ii) peripheral resistances in all right-hand segments. Simulated results were qualitatively tested by comparison against in-vivo data [2]. Results: Our main findings were two. Firstly, by turning off the endothelial function in the model we showed that vasoconstriction immediately after cuff deflation is entirely due to a change in transmural pressure, which offsets shear-mediated vasodilation and leads to a peak dilator response occurring after peak blood flow. Secondly, for the same endothelial function, FMD test values dropped with decreasing peripheral vasodilation, increasing arterial stiffness, and decreasing central blood pressure. Conclusion: The influence of confounding factors on FMD test results suggests that the current FMD test does not assess true endothelial cell function.

It has been recognised for nearly 200 years that the human pressure waveform changes in shape with ageing and disease. The shape of the pressure waveform has been explained in terms of two fundamental models: the Windkessel (reservoir) and wave theory. In its simplest form the Windkessel model satisfactorily explains the pressure waveform in diastole but cannot model pressure changes in systole. Wave theory satisfactorily models the pressure waveform but predicts the existence of ‘self-cancelling’ forward and backward waves in diastole which are difficult to explain in biological terms. We propose that a hybrid reservoir–wave model better describes the pressure waveform and may enable assessment of aortic function from pressure measurements made at any large systemic artery.

Mental stress is a major burden for our society. Invasive and non-invasive methods have been proposed to monitor and quantify it using various sensors on and off body. In this Letter, the authors investigated the use of the arm photoplethysmogram (PPG) to assess mental stress in laboratory conditions. Results were in correspondence with their previous in-silico study which guided the present study. Three wave shape parameters were identified for stress assessment from the PPG signal: (i) the time from dicrotic notch to end diastole; (ii) the time from pulse onset to systolic peak; and (iii) the ratio of diastolic to systolic area. The proposed in-vivo results showed that the two first parameters responded significantly to increased mental stress and to a breathing relaxation procedure, complementing heart rate, heart rate variability, and pulse transit time as indices of stress.

Purpose: Aortic stiffness is predictive of cardiovascular morbidity and mortality. However, the gold standard method for assessing aortic stiffness, carotid-femoral pulse wave velocity, is time-consuming and requires a trained operator. An alternative approach could be to derive an arterial stiffness index (ASI) from the easily measured finger photoplethysmogram (PPG). Our aim was to investigate the performance of PPG-derived ASIs for assessing aortic stiffness. Methods: An insilico dataset of arterial pulse waves (PWs) was generated using a model of pulse wave propagation (1). PWs were generated for virtual healthy subjects aged 25 to 75. Several simulations were run for each age decade to mimic population-level variation in cardiac and vascular properties. PPG PWs were simulated from blood pressure PWs (2). The dataset was used to design an algorithm to calculate over 30 ASIs described in the literature from the finger PPG. In vivo testing was performed using 6,506 subjects from the Airwave dataset (3) who had triplicate PPG and reference PWV measurements. Results: In silico and in vivo performances of ASIs, including the stiffness index (SI) and reflection index, varied greatly. The SI performed well in vivo, showing strong correlation with reference PWVs. However, in silico assessment demonstrated that the SI and other ASIs were affected by other cardiac and vascular properties as well as aortic stiffness. Conclusions: This study identified the best-performing ASIs in both in silico and in vivo datasets. In the future multiple ASIs should be combined to improve performance, since different ASIs have different physiological determinants.

We examined the ability of a simple reduced model comprising a proximal characteristic impedance linked to a Windkessel element to accurately predict central Pulse Pressure (PP) from aortic blood flow, verified that parameters of the model corresponded to physical properties, and applied the model to examine PP dependence on cardiac and vascular properties. PP obtained from the reduced model was compared with theoretical values obtained in silico and measured values in vivo. Theoretical values were obtained using a distributed multisegment model in a population of virtual (computed) subjects in which cardiovascular properties were varied over the pathophysiological range (n = 3,095). In vivo measurements were in normotensive subjects during modulation of physiology with vasoactive drugs (n = 13) and in hypertensive subjects (n = 156). Central PP derived from the reduced model agreed with theoretical values (mean difference ±SD, −0.09 ± 1.96 mmHg) and with measured values (means differences −1.95 ± 3.74 and −1.18 ± 3.67 mmHg for normotensive and hypertensive subjects, respectively). Parameters extracted from the reduced model agreed closely with theoretical and measured physical properties. Central PP was seen to be determined mainly by total arterial compliance (inversely associated with central arterial stiffness) and ventricular dynamics: the blood volume ejected by the ventricle into the aorta up to time of peak pressure and blood flow into the aorta (corresponding to the rate of ventricular ejection) up to this point. Increased flow and volume accounted for 20.1 mmHg (52%) of the 39.0 mmHg difference in PP between the upper and lower tertiles of the hypertensive subjects 1.

Purpose: Central Blood Pressure (CBP) is a better cardiovascular risk indicator than brachial pressure [1]. However, gold standard CBP measurements require an invasive catheter. We propose an approach to estimate CBP non-invasively from Magnetic Resonance Imaging (MRI) data coupled with a non-invasive brachial pressure measurement, using reduced-order (0-D/1-D) computational models. Our objectives were: identifying optimum model parameter estimation methods and comparing the performance of 0-D/1-D models for estimating CBP. Methods: Populations of virtual (simulated) healthy subjects were generated based on [2]. Pressure and flow waveforms from these populations were used to develop and test Methods: for estimating model parameters. Two common clinical scenarios were considered: when a brachial pressure waveform is available; and when only systolic and diastolic blood pressures are available. Optimal parameter estimation Methods: were identified for each scenario and used with two 0-D Windkessel models and a 1-D aortic model. Results were compared with invasive CBP in a post-coarctation repair population (n = 10). Results: Model parameters were best estimated by: fitting an exponential to the pressure waveform to estimate compliance and outflow pressure; using the least-squares method to estimate pulse wave velocity from flow data; assuming characteristic impedance was 5% of arterial resistance; and estimating end-systolic time from the second derivative of the pressure waveform. Average pulse and systolic CBP errors were

Objective and motivation: Endothelial dysfunction is associated with cardiovascular diseases. Flow mediated dilation (FMD), assesses the endothelial function by measuring the brachial artery vasodilation following deflation of a sphygmomanometer cuff around the forearm. Vasodilation is assumed to be due to an increase in wall shear stress (WSS) only. However, there is evidence that the vasodilation may be affected by other confounding factors1. We aim to investigate the effects of confounding factors on the results of FMD. Methods: A dynamic simulation of FMD was carried out using a one-dimensional haemodynamic solver of blood flow in the arm arterial vasculature (Fig. 1a) 2. Haemodynamics during cuff deflation was simulated by prescribing a decrease in peripheral resistance (Fig. 1b) in a novel mathematical model which dynamically couples increasing WSS (Fig. 1c) to decreasing arterial wall Young’s modulus (Fig. 1d), taking into account endothelial function. Results: Our results show that the initial increase in flow velocity (Fig. 1e) is caused by the prescribed decrease in peripheral resistance and leads to an initial pressure drop affecting the FMD value. WSS induces a drop in Young’s modulus leading to vasodilation (Fig. 1f). In addition, for the same prescribed endothelial function (relating WSS to Young’s modulus variation) and decreased peripheral resistance, FMD increases with decreasing arterial stiffness (3.17% vs 5.31% vs 8.56% (Fig. 1f)). ConclusionOur numerical model successfully described FMD haemodynamics and highlighted one of the important confounding factors of FMD values: arterial stiffness. We are currently investigating other factors and ways of correcting those factors.

An abdominal aortic aneurysm (AAA) is usually asymptomatic until rupture, which is associated with extremely high mortality. Consequently, the early detection of AAAs is of paramount importance in reducing mortality; however, most AAAs are detected by medical imaging only incidentally. The aim of this study was to investigate the feasibility of machine learning-based pulse wave (PW) analysis for the early detection of AAAs using a database of in silico PWs. PWs in the large systemic arteries were simulated using one-dimensional blood flow modelling. A database of in silico PWs representative of subjects (aged 55, 65 and 75 years) with different AAA sizes was created by varying the AAA-related parameters with major impacts on PWs—identified by parameter sensitivity analysis—in an existing database of in silico PWs representative of subjects without AAAs. Then, a machine learning architecture for AAA detection was trained and tested using the new in silico PW database. The parameter sensitivity analysis revealed that the AAA maximum diameter and stiffness of the large systemic arteries were the dominant AAA-related biophysical properties considerably influencing the PWs. However, AAA detection by PW indexes was compromised by other non-AAA related cardiovascular parameters. The proposed machine learning model produced a sensitivity of 86.8 % and a specificity of 86.3 % in early detection of AAA from the photoplethysmogram PW signal measured in the digital artery with added random noise. The number of false positive and negative results increased with increasing age and decreasing AAA size, respectively. These findings suggest that machine learning-based PW analysis is a promising approach for AAA screening using PW signals acquired by wearable devices.

In this work we present a one-dimensional (1D) mathematical model of the coronary circulation and use it to study the effects of arrhythmias on coronary blood flow (CBF). Hydrodynamical models are rarely used to study arrhythmias’ effects on CBF. Our model accounts for action potential duration, which updates the length of systole depending on the heart rate. It also includes dependency of stroke volume on heart rate, which is based on clinical data. We apply the new methodology to the computational evaluation of CBF during interventricular asynchrony due to cardiac pacing and some types of arrhythmias including tachycardia, bradycardia, long QT syndrome and premature ventricular contraction (bigeminy, trigeminy, quadrigeminy). We find that CBF can be significantly affected by arrhythmias. CBF at rest (60 bpm) is 26% lower in LCA and 22% lower in RCA for long QT syndrome. During bigeminy, trigeminy and quadrigeminy, respectively, CBF decreases by 28%, 19% and 14% with respect to a healthy case.

Purpose: Hypertension, the single biggest killer worldwide1, arises mainly as a result of an increase in central pulse pressure (PP)2, yet haemodynamic basis of that increase is still disputed. We examined the ability of a simple “reduced” model comprising a proximal characteristic impedance linked to a Windkessel element to accurately predict PP from aortic blood flow and applied the model to examine PP dependence on cardiac and vascular properties. Method: PP obtained from the model was compared with theoretical values obtained in silico and in vivo. Theoretical values were obtained using a distributed multi-segment model in a population of “virtual” subjects (n = 3,095) in which cardiovascular properties were varied over the pathophysiological range. In vivo measurements were in normotensive subject (n = 13) during modulation of physiology with vasoactive drugs with divergent actions on cardiac and cascular properties and in hypertensive subjects (n = 156). Results: PP derived from the model agreed with theoretical values (mean difference SD, −0.09±1.96 mmHg) and with measured values (−1.95±3.74 and −1.18±3.67 mmHg for normotensive and hypertensive subjects respectively). Parameters extracted from the model agreed closely with theoretical and measured physical properties. PP was seen to be determined mainly by total arterial compliance (inversely associated with arterial stiffness) and ventricular dynamics: the volume of blood ejected up to time of pulse pressure and the rate of ventricular ejection up to this point. Conclusion: Increased flow and/or volume accounted for 20.1 mmHg (52%) of the 39.0 mmHg difference in pulse pressure between the upper and lower tertiles of the hypertensive subjects.

Background and objectives: Clinical evidence shows that central (aortic) blood pressure (CBP) is a better marker of cardiovascular risk than brachial pressure [1]. However, CBP can only be accurately measured invasively, through catheterisation. We propose a novel approach to estimate CBP non-invasively from aortic MRI data and a non-invasive peripheral (brachial) pressure measurement, using a one-dimensional (1-D) model of aortic blood flow. Methods: We created a population of virtual (computed) subjects, each with distinctive arterial pulse waveforms available at multiple arterial locations, to assess our approach. This was achieved by varying cardiac (stroke volume, cardiac period, time of systole) and arterial (pulse wave velocity, peripheral vascular resistance) parameters of a distributed 1-D model of the larger systemic arteries [2] within a wide range of physiologically plausible values. After optimising our algorithm for the aortic 1-D model in silico, we tested its accuracy in a clinical population of 8 post-coarctation repair patients. Results: Results from our in silico study, after varying cardiac and arterial parameters by ±30%, showed maximum relative errors for systolic, mean and diastolic CBP of 4.5%, 3.6% and 4.2%, respectively. Average relative errors for systolic, mean and diastolic CBP were 2.7%, 0.9% and 1.2%, respectively. Corresponding average relative errors from our clinical study were 5.4%, 1.5% and 8.0%. Figure 1CBP estimation using the aortic 1-D model for a given virtual patient. Figure 2Systolic CBP estimated using the aortic 1-D model against reference systolic CBP values from in silico and in vivo data. Conclusions: We have provided a proof of concept for the non-invasive estimation of patient-specific central blood pressure using computational aortic blood flow modelling in combination with MRI data and a non-invasive peripheral pressure measurement.

The rate at which an individual recovers from exercise is known to be indicative of cardiovascular risk. It has been widely shown that the reduction in heart rate immediately after exercise is predictive of mortality. However, little research has been conducted into whether the time taken for the blood vessels to return to normal is also indicative of risk. In this study, we present a novel approach to assess vascular recovery rate (VRR) using the photoplethysmogram (PPG) signal, which is monitored by smart wearables. The VORTAL dataset (http://peterhcharlton.github.io/RRest/) was used for this study, containing PPG signals from 39 healthy subjects before (baseline) and after exercise. 31 VRR indices were extracted from the PPG pulse wave shape, as well as heart rate for comparison. The rate at which indices returned to baseline after exercise was quantified, and the consistency of changes between subjects was assessed statistically. Many VRR indices exhibited changes after exercise which were consistent between subjects. Indices derived from the timings and second derivative of pulse waves were identified as candidates for future work. The rate at which the indices returned to baseline differed between indices and subjects, indicating that they may provide additional information beyond that of heart rate, and that they may be useful for stratifying subjects. This study demonstrated the feasibility of assessing VRR after exercise from the PPG. Future studies should investigate whether VRR indices are associated with cardiovascular fitness, and the potential utility of incorporating the indices into wearable sensors.

Background: A recent study has shown that the central pressure waveform could be determined by a very small set of parameters accounting for the physical properties of the heart and the arteries [1]. Particularly, main pressure features like first systolic shoulder (P1) and systolic (P2) pressures were estimated accurately. Methods: By combining a numerical virtual population (n=3,325) similar to [2] and experimental data acquired from a pressure/Doppler flow velocity transducer place in the ascending aorta in 18 patients (meanSD: age 63±11 yr, aortic BP 136±23/73±13 mmHg) at the time of cardiac catheterization, we assessed the accuracy of those predictions for magP1 (P1-DBP) and P2 using respectively a water hammer [3] and a 3-element Windkessel models [4]. Contributions of the heart and arterial properties to these estimates though respectively blood velocity, volume and pulse wave velocity, compliance, resistance were then derived from the theoretical models used. Results: P1 and P2 estimates agreed well with theoretical pressure both in the numerical dataset (mean+//0-SD difference, 1.1±3.2 mmHg and -1.6±3.1 mmHg respectively) and the clinical cohort (mean+//0-SD difference, -2.4±5.5 mmHg and 1.9±6.5 mmHg respectively). The ratio arterial-to-heart contribution has been shown to be fairly constant as magP1 was increasing. Conclusions: Arteries and heart contribute as much to rise in P1. More clinical data are being collected to quantify the contributions of the heart and arteries to P2.

Many physiological indexes and algorithms based on pulse wave analysis have been suggested in order to better understand the physiology of arterial hemodynamics (e.g. pulse wave velocity, transfer functions for central blood pressure derivation). Because these tools are most often computed from hemodynamic measurements, their validation is time-consuming and biased by measurement errors. We present a new methodology to assess theoretically these computed tools. We create a database of virtual healthy subjects using a numerical 1D-0D model of the arterial hemodynamics, which parameters are varied to cover a physiological healthy range. The generated set of simulations encloses a wide selection of possible cases that could be encountered in a clinical study. We illustrate this new concept by assessing the efficiency of indexes estimating aortic stiffness, such as central and peripheral foot-to-foot pulse wave velocities computed with different methods (foot-to-foot, sum of squares), the stiffness index and the augmentation index. We also apply our methodology to a new algorithm that estimates the central aortic pressure from peripheral measurements. We show that the results of our analysis confirm clinical observations. Our database of virtual subjects could become a new tool for the clinician: it provides insight into the physical mechanisms that are important when designing large cohort clinical studies, analyzing their results, and explaining the correlations observed in clinical practice.

Backward traveling waves, an important determinant of central haemodynamics, are usually regarded as being due to reflections from discontinuities in the arterial tree. However, consideration of a single tube model of the arterial with a single site of reflection shows that a backward pressure wave may be generated by elastic recoil of large arteries, in which case the magnitude of the backward wave is proportional to that of the forward wave. A 55-segment 1-D model of the arterial which allows reflection as a continuum along the arterial tree and, for a given prescribed aortic flow, generates physiological aortic pulse waveforms was used to examine the relation of the backward to forward pressure waves in 4107 “virtual subjects” with arterial parameters spanning the physiological range. Backward pressure wave was closely correlated with the forward wave (R = 0.931, P < 0.001). Clinical data was obtained by carotid tonometry and aortic Doppler sonography during modulation of cardiovascular function in healthy volunteers (n = 13, age 46.5 ± 10.1 years with inotropic, vasopressor and vasodilator drugs (dobutamine, norepinephrine phentolamine and nitroglycerin). The magnitude of backward pressure was highly correlated with forward pressure over a range 5–15 mmHg (R = 0.824, P < 0.001) with a constant ratio of backward to forward wave magnitude except during treatment with nitroglycerin, a vasodilator known to be highly selective for large muscular arteries. These numerical and experimental data suggest that backward pressure waves can be generated by elastic recoil of large arteries independent of pressure wave reflection and this effect dominates in human physiology.

Despite central pressure’s predictive power of cardiovascular risk, brachial pressure is the clinical standard. However, amplified brachial systolic pressure varies significantly with age, and during therapy. Our aim was to modulate individual arterial and ventricular parameters in an experimental model of the cardiovascular system, to quantify each parameter’s contribution to arterial pressure and its amplification. A piston driven ventricle provided computer-controlled flow waveforms into various silicone arterial trees. Silicone tubes diameters (20, 15, 10mm), wall thicknesses (0.5, 0.7, 1.0, 1.5mm), lengths (30–400cm), taper (20mm inlet to 20, 15, 10 and 5mm outlets), were each applied with various ventricular stroke profiles (sawtooth to sinewave). Intravascular pressure-tip wires and ultrasonic flow probes measured pressure and flow. MAP, flow and HR were maintained between tests for comparison. Ventricular stroke profile independently augmented pressure amplification from 16% to 82% between sinewave and sawtooth ejections profiles. As expected for any arterial model, the transfer function from central to distal pressure measurement sites remained constant. Decreasing taper, wall thickness, and length, and increasing diameter each increased amplification by shifting the peak of the amplifying transfer function towards the more prominent lower frequencies, (1–3Hz). However, the amplification variation between all vascular parameters was

Estimation of aortic and left ventricular (LV) pressure usually requires measurements that are difficult to acquire during the imaging required to obtain concurrent LV dimensions essential for determination of LV mechanical properties. We describe a novel method for deriving aortic pressure from the aortic flow velocity. The target pressure waveform is divided into an early systolic upstroke and a diastolic decay, interposed by a late systolic portion described by a second-order polynomial. Pulse wave velocity (PWV), mean arterial pressure, diastolic pressure and diastolic decay are required inputs for the algorithm. The algorithm was tested using a) pressure data derived theoretically from pre-specified flow waveforms and properties of the arterial tree using a single-tube 1-D model of the arterial tree and b) experimental data acquired from a pressure/Doppler flow velocity transducer placed in the ascending aorta (n=18, mean+/−SD, age: 63+/−11 years, aortic BP: 136+/−23 / 73+/−13 mmHg) at the time of cardiac catheterisation. For experimental data, PWV was calculated from measured pressures/flows and mean, diastolic pressures and diastolic decay were taken from measured pressure. Pressure reconstructed from measured flow agreed well with theoretical pressure: mean+/−SD root mean square (RMS) error 0.7+/−0.1 mmHg. Similarly, for experimental data, pressure reconstructed from measured flow agreed well with measured pressure (mean RMS error 2.4+/−1.0 mmHg). First systolic shoulder and systolic peak pressures were also accurately rendered (mean+/−SD difference 1.4+/−2.0 mmHg for peak systolic pressure). This is the first non-invasive derivation of aortic pressure based on fluid dynamics (flow and wave speed) in the aorta itself.