Model-Based Blind System Identification Approach to Estimation of Central Aortic Blood Pressure Waveform From Noninvasive Diametric Circulatory Signals

Abstract

This paper presents a model-based blind system identification approach to estimation of central aortic blood pressure (BP) waveform from noninvasive diametric circulatory signals. First, we developed a mathematical model to reproduce the relationship between central aortic BP waveform and a class of noninvasive circulatory signals at diametric locations by combining models to represent wave propagation in the artery, arterial pressure–volume relationship, and mechanics of the measurement instrument. Second, we formulated the problem of estimating central aortic BP waveform from noninvasive diametric circulatory signals into a blind system identification problem. Third, we performed identifiability analysis to show that the mathematical model could be identified and its parameters determined up to an unknown scale. Finally, we illustrated the feasibility of the approach by applying it to estimate central aortic BP waveform from two diametric pulse volume recording (PVR) signals. Experimental results from ten human subjects showed that the proposed approach could estimate central aortic BP waveform accurately: the average root-mean-squared error (RMSE) associated with the central aortic BP waveform was 4.1 mm Hg (amounting to 4.5% of the underlying mean BP) while the average errors associated with central aortic systolic pressure (SP) and pulse pressure (PP) were 2.4 mm Hg and 2.0 mm Hg (amounting to 2.5% and 2.1% of the underlying mean BP). The proposed approach may contribute to the improved monitoring of cardiovascular (CV) health by enabling estimation of central aortic BP waveform from conveniently measurable diametric circulatory signals.

Estimation of Cardiovascular Risk Predictors from Non-Invasively Measured Diametric Pulse Volume Waveforms via Multiple Measurement Information Fusion

“…2 ): where s is the Laplace operator, τ 1 and τ 2 are the central-arm and central-ankle PTTs, respectively, η ij , j = 1, 2 are the polynomial parameters associated with the y j ( s ), while E 1 , E 2 , and η VE are the parameters characterizing the viscoelastic model associated with the brachial artery-tissue-arm cuff interface, the following correlation equation can be formed by canceling the unknown yet common central BP P 0 from Eq. ( 1 ): which can be solved to derive the unknown subject-specific parameters in the arterial line models G 1 ( s ) and G 2 ( s ) via, e.g., numerical optimization in the time domain 10 , 38 : where is the Laplace transform operator and is the vector of unknown arterial line model parameters. Note that the first, second, and third terms in Eq.…”

A review on low-dimensional physics-based models of systemic arteries: application to estimation of central aortic pressure

“…A radial [71, 73, 74], brachial [26] or femoral [72] pressure measurement is commonly used as model input. Moreover, some researchers [27, 90–92] chose two measured peripheral pressure waveforms as inflow conditions such as radial and femoral arteries. Because intravascular and extravascular pressure at radial artery is very close, the radial pressure waveform can be accurately recorded using an applanation tonometry.…”

“…Individualized transfer functions were built by Sugimachi et al [73] and Hahn et al [74] using a single tube model with a three element Windkessel load. Ghasemi et al [27, 90], Lee [91] and Kim et al [92] utilized a T-tube model with a three-element Windkessel load, to reconstruct central aortic pressure from two measured peripheral pressure.…”

“…In order to acquire an adaptive or individualized transfer function, most researchers measured the pulse transit time for each subject through some noninvasive approaches and calculated the remaining parameters by population averaging [26, 27, 71–74, 92, 93]. Blind system identification is another method of estimating model parameters, which can reconstruct the central aortic pressure waveform from two distinct peripheral pressure waveforms [90, 91]. The blind system identification method can obtain fully individualized parameters.…”

The Noninvasive Measurement of Central Aortic Blood Pressure Waveform

“…The blind system identification (BSI) method reconstructs the input from two or more outputs. In the estimation of central aortic pressure waveform, BSI reconstructs the central aortic pressure waveform based on two peripheral pressure waveforms [45][46][47][48][49][50]. This method is fully individualized, without the need of measuring or estimating pulse transit time.…”