With the rise in prevalence of cardiovascular (CV) disease, risk stratification is becoming increasingly important. Accurate characterization of the CV system is required, for which central aortic blood pressure (BP) parameters form an integral part. However, invasive measurement of central aortic BP parameters (aP) is difficult. Therefore, non-invasive methods to estimate aP from the radial pressure pulse (rPulse) have been proposed. To analyze accuracy of estimated aP (aPhat) and applicability in risk stratification and diagnosis, this study presents: (1) a novel representation of the rPulse with minimal loss of information, (2) a framework for strict definition and statistical analysis of aPhat, and (3) a dynamic analysis of effects of mean BP (MP) and heart rate (HR) in the rPulse shape. Methods: (1) 2671 rPulse s measured by applanation tonometry were represented using the first eight principal components (PC) scores after standard PC transformation. rPulse shapes were compared in three subpopulations. (2) The concept of &quotestimation option&quot (EO) for aP estimation was presented. A framework for strict definition of aPhat and the comparison of EOs was proposed, and 7 different EOs compared. (3) A sequence of rPulse s was analyzed during soft exhalation maneuver (SEM) %, a mild Valsalva type maneuver, in eight healthy subjects. Radial BP and respiration pressure were continuously measured. The effects of MP and HR in the rPulse parameters were analyzed by standard linear regression for each subject. Results: (1) PC representation of the rPulse improves accuracy of the estimation of aPhat compared with the simple use of rPulse parameters. Subpopulations have distinctive rPulse shapes. (2) No single EO was better for the estimation of all aPhat. Inclusion of MP improves estimation accuracy. Despite further improvement when rPulse is included, the general transfer function EO is a biased estimator. (3) The dynamic analysis of the rPulse provides information of the effects of MP and HR in the rPulse not available in static analysis. The effects were specific for each individual and different from the results obtained from a general population. Conclusions: For accurate CV risk stratification, future studies should include a dynamic measurement of calibrated radial pressure pulse during SEM maneuver. Risk analysis and diagnosis should be based on representations of the rPulse with minimum loss of information. aPhat should be used for better understanding of the underlying physiological principles.

The central aortic blood pressure signal is an important source of information that contains cues about the cardiovascular system condition. Measuring this pulse wave clinically is burdensome as it can be only measured invasively with a catheter. As a result, many mathematical tools have been proposed in the past few decades to reconstruct the aortic pressure signal from the peripheral pressure signals that are usually easier to obtain noninvasively. At the distal level, the blood pressure signal is not directly useful since factors, such as length and stiffness of the arteries, play roles in changing the shape of the pressure signal significantly. In this thesis, multi-channel blind system identification techniques are proposed to estimate the central pressure waveform which vary in their accuracy and complex- ity. First, a simple linear method is applied by approximating the nonlinear arterial system as a linear time-invariant system and applying the cross-relation approach. Next, a more complicated nonlinear Wiener system is proposed to model the nonlinear arterial tree. Along with the channel’s coefficients, the nonlinear functions are estimated using cross-relation and kernel methods. Data-driven machine learning methods are tested to estimate the aortic pressure signals. In many cases, they suffer from underfitting problems. As a remedy, a hybrid machine learning and cross-relation approach is also proposed to add more robustness to the machine learning models. This hybrid approach is implemented by combining the cross-relation with any machine learning method, including deep learning approaches. The various methods are tested using pre-validated virtual databases. The results show that the linear method produces root mean squared errors between 3.40 mmHg and 6.24 mmHg depending on the cross-relation constraint and the equalization tech- nique. On the other hand, the root mean squared errors associated with the nonlinear methods are between 3.76 mmHg and 4.22 mmHg and hence more stable. For the hybrid machine learning and cross-relation approach, applying the cross-relation and the dictionary learning reduce the root mean squared errors up o 67% comparing with the pure machine learning models.

abstract: Background. Despite research aimed at understanding the mechanisms of essential hypertension, instances of this condition continue to rise. Recent findings indicate that the administration of dietary nitrates, in the form of beetroot juice and other nitrate-rich vegetables, may offer anti-hypertensive effects in various study populations. Objective. This randomized, placebo-controlled, crossover trial sought to compare the effects of high-nitrate vegetable salads to the effects of low-nitrate canned vegetables on plasma nitrate/nitrite concentration, peripheral and central-aortic systolic and diastolic blood pressures, pulse wave velocity, and flow-mediated dilation. Methods. Healthy, post-menopausal women (n=5; 80% Caucasian; 52.6 ± 5.7 years) with mildly elevated blood pressure (mean blood pressure ≥ 115/70 mm Hg and < 140/80 mm Hg) were randomly assigned to ingest a fresh, high-nitrate vegetable salad or a low-nitrate vegetable medley, twice per day, for a total of 10 consecutive days. Given the crossover design of the trial, participants observed a two to three week washout period followed by reassignment to the opposite condition. Findings were considered significant at a p-value < 0.05, and Wilcoxon Signed-Rank tests compared mean differences between conditions. Results. Plasma nitrate/nitrite concentration was significantly higher following consumption of the high-nitrate versus the low-nitrate condition (p = 0.043). Conversely, the differences in peripheral systolic and diastolic blood pressures were not statistically significant (p = 0.345 and p = 0.684 for systolic and diastolic pressures, respectively) nor were the differences in central-aortic systolic and diastolic blood pressures statistically significant (p = 0.225 and p = 0.465 for systolic and diastolic pressures, respectively). Similarly, when comparing the effects of the high-nitrate condition to the low-nitrate condition, the difference in pulse wave velocity was not statistically significant (p = 0.465). Finally, flow-mediated dilation tended to improve following consumption of the high nitrate condition (p = 0.080). Conclusion. Twice daily consumption of a fresh, high-nitrate vegetable salad significantly increased plasma nitrate/nitrite concentration. Although the trial was underpowered, there was a trend for improved flow-mediated dilation. Finally, twice daily consumption of a fresh, high-nitrate vegetable salad did not significantly lower peripheral or central-aortic systolic or diastolic blood pressures or pulse wave velocity. / Dissertation/Thesis / Masters Thesis Nutrition 2017