Functional respiratory imaging as a tool to assess upper airway patency in children with obstructive sleep apnea.

OBJECTIVE: We aim to investigate if anatomical and functional properties of the upper airway using computerized 3D models derived from computed tomography (CT) scans better predict obstructive sleep apnea (OSA) severity than standard clinical markers. METHODS: Consecutive children with suspected OSA underwent polysomnography, clinical assessment of upper airway patency, and a CT scan while awake. A three-dimensional (3D) reconstruction of the pharyngeal airway was built from these images, and computational fluid dynamics modeling of low inspiratory flow was performed using open-source software. RESULTS: Thirty-three children were included (23 boys; mean age, was 6.0±3.2y). OSA was diagnosed in 23 patients. Children with OSA had a significantly lower volume of the overlap region between tonsils and the adenoids (median volume, 1408 mm compared to 2173 mm; p=0.04), a lower mean cross-sectional area at this location (median volume, 69.3mm(2) compared to 114.3mm2; p=0.04), and a lower minimal cross-sectional area (median volume, 17.9 mm2 compared to 25.9 mm2; p=0.05). Various significant correlations were found between several imaging parameters and the severity of OSA, most pronounced for upper airway conductance (r=-0.46) (p<0.01) for correlation between upper airway conductance and the apnea-hypopnea index. No differences or significant correlations were observed with clinical parameters of upper airway patency. Preliminary data after treatment showed that none of the patients with residual OSA had their smallest cross-sectional area located in segment 3, and this frequency was significantly lower than in their peers whose sleep study normalized (64%; p=0.05). CONCLUSION: Functional imaging parameters are highly correlated with OSA severity and are a more powerful correlate than clinical scores of upper airway patency. Preliminary data also showed that we could identify differences in the upper airway of those subjects who did not benefit from a local upper airway treatment.

Determining carotid artery pressure from scaled diameter waveforms: comparison and validation of calibration techniques in 2026 subjects.

Calibrated diameter distension waveforms could provide an alternative for local arterial pressure assessment more widely applicable than applanation tonometry. We compared linearly and exponentially calibrated carotid diameter waveforms to tonometry readings. Local carotid pressures measured by tonometry and diameter waveforms measured by ultrasound were obtained in 2026 subjects participating in the Asklepios study protocol. Diameter waveforms were calibrated using a linear and an exponential calibration scheme and compared to measured tonometry waveforms by examining the mean root-mean-squared error (RMSE), carotid systolic blood pressure (SBPcar) and augmentation index (AIx) of calibrated and measured pressures. Mean RMSE was 5.2(3.3) mmHg (mean(stdev)) for linear and 4.6(3.6) mmHg for exponential calibration. Linear calibration yielded an underestimation of SBPcar by 6.4(4.1) mmHg which was strongly correlated to values of brachial pulse pressure (PPbra) (R = 0.4, P < 0.05). Exponential calibration underestimated true SBPcar by 1.9(3.9) mmHg, independent of PPbra. AIx was overestimated by linear calibration by 1.9(10.1)%, the difference significantly increasing with increasing AIx (R = 0.25, P < 0.001) and by exponential calibration by 5.4(10.6)%, independently of the value of AIx. Properly calibrated diameter waveforms offer a viable alternative for local pressure estimation at the carotid artery. Compared to linear calibration, exponential calibration significantly improves the pressure estimation.

Three- and four-element Windkessel models: assessment of their fitting performance in a large cohort of healthy middle-aged individuals.

Lumped-parameter models are used to estimate the global arterial properties by fitting the model to measured (aortic) pressure and flow. Different model configurations coexist, and it is still an open question as to which model optimally reflects the arterial tree and leads to correct estimates of arterial properties. An assessment was made of the performance of (a) the three-element Windkessel model (WK3) consisting of vascular resistance R, total arterial compliance C, and characteristic impedance Zc; (b) a four-element model with an inertance element L placed in parallel with Zc (WK4-p); and (c) a four-element model with L placed in series with Zc (WK4-s). Models were fitted to data measured non-invasively in 2404 healthy subjects, aged between 35 and 55 years. It was found that model performance segregated into two groups. In a group containing 20 per cent of the dataset (characterized by low blood pressure and wave reflection) the WK4-p model outperformed the other models, with model behaviour as envisioned by its promoters. In these cases, the WK3 and WK4-s models led to increased overestimation of total arterial compliance and underestimation of characteristic impedance. However, in about 80 per cent of the cases, the WK4-p model showed a behaviour that was very similar to that of the WK3 and WK4-s models. Here, the WK4-s model yielded the best quality of fit, although model parameters reached physically impossible values for L in about 12 per cent of all cases. The debate about which lumped-parameter model is the better approximation of the arterial tree is therefore still not fully resolved.

Noninvasive assessment of left ventricular and myocardial contractility in middle-aged men and women: disparate evolution above the age of 50?

End-systolic elastance (E(es)) is a frequently used index of left ventricular (LV) contractility. However, because of its inherent dependence on LV geometry, E(es) cannot be used to compare myocardial contractile state between ventricles with different geometries, which is the case in any cross-sectional study. Various normalization methods for E(es) have been proposed in the literature, but a standardized method is still lacking. In this study, we introduced a novel alternative normalization technique and compared it with three previously suggested methods. We tested all normalization methods to assess the age- and sex-related differences in myocardial contractility in a large population sample of 2,184 middle-aged (ages, 35-55 yr) untreated subjects free from overt cardiovascular disease. Ventricular contractility E(es) was determined using a previously validated noninvasive single-beat method, based on two-dimensional echocardiographic and brachial blood pressure measurements. Myocardial contractility was estimated as 1) E(es).end-diastolic volume (EDV); 2) E(es).LV mass (LVM); 3) 0.433.E(es).LVM/relative wall thickness (RWT), based on a theoretical LV model; and 4) 0.0941.E(es).LVM(0.455).RWT(-0.159), a novel semiempirical expression derived in this study. Because of the difference in their underlying assumptions, the various myocardial contractility indexes do not provide consistent information with respect to sex differences. Despite these discrepancies, it was found that myocardial contractility in women appears to be better preserved after the age of 50 yr compared with that in men. The physiological mechanisms behind this potentially clinically important phenomenon at population level require further investigation.

Aortic reflection coefficients and their association with global indexes of wave reflection in healthy controls and patients with Marfan's syndrome.

Early return of reflected pressure waves increases the load on central arteries and may increase the risk of aortic rupture in patients with Marfan's syndrome (MFS). To assess whether wave reflection is elevated in MFS, we used ultrasound and MRI to measure central pressure and flow waveforms in 26 patients (13-54 yr of age) and 26 age- and gender-matched controls. Aortic systolic and diastolic cross-sectional areas were measured at the ascending and descending aorta (AA and DA), diaphragm (DIA), and lower abdominal aorta (AB). From these measurements, local characteristic impedance (Z(0-xx)) and local reflection coefficients (Gamma(xx-yy)) were calculated. Calculated global wave reflection indexes were the augmentation index (AIx) and the ratio of backward to forward pressure wave (P(b)/P(f)). The aorta was wider in MFS patients at AA (P < 0.01) and DA (P < 0.01). Aortic pulse wave velocity was 42 cm/s higher in MFS patients (P < 0.05). Z(0-xx) was not different between groups, except at DA, where it was lower in MFS patients. In controls, Gamma(AA-DA) was 0.31 +/- 0.08, Gamma(DA-DIA) was 0.00 +/- 0.11, and Gamma(DIA-AB) was 0.31 +/- 0.16. Mean values of Gamma(xx-yy) were not different between MFS patients and controls. In controls, aging diminished Gamma(AA-DA) but increased Gamma(DIA-AB). Clear age-related patterns were absent in MFS patients. AIx or P(b)/P(f) was not higher in MFS patients than in controls. There were indications for enhanced wave reflection in young MFS patients. Our data demonstrated that the major determinants of AIx were pulse wave velocity and the effective length of the arterial system and, to a lesser degree, HR and P(b)/P(f).

Nonlinear isochrones in murine left ventricular pressure-volume loops: how well does the time-varying elastance concept hold?

The linear time-varying elastance theory is frequently used to describe the change in ventricular stiffness during the cardiac cycle. The concept assumes that all isochrones (i.e., curves that connect pressure-volume data occurring at the same time) are linear and have a common volume intercept. Of specific interest is the steepest isochrone, the end-systolic pressure-volume relationship (ESPVR), of which the slope serves as an index for cardiac contractile function. Pressure-volume measurements, achieved with a combined pressure-conductance catheter in the left ventricle of 13 open-chest anesthetized mice, showed a marked curvilinearity of the isochrones. We therefore analyzed the shape of the isochrones by using six regression algorithms (two linear, two quadratic, and two logarithmic, each with a fixed or time-varying intercept) and discussed the consequences for the elastance concept. Our main observations were 1) the volume intercept varies considerably with time; 2) isochrones are equally well described by using quadratic or logarithmic regression; 3) linear regression with a fixed intercept shows poor correlation (R(2) < 0.75) during isovolumic relaxation and early filling; and 4) logarithmic regression is superior in estimating the fixed volume intercept of the ESPVR. In conclusion, the linear time-varying elastance fails to provide a sufficiently robust model to account for changes in pressure and volume during the cardiac cycle in the mouse ventricle. A new framework accounting for the nonlinear shape of the isochrones needs to be developed.

Wave intensity analysis of left ventricular filling.

Wave intensity analysis (WIA) is a powerful technique to study pressure and flow velocity waves in the time domain in vascular networks. The method is based on the analysis of energy transported by the wave through computation of the wave intensity dI = dPdU, where dP and dU denote pressure and flow velocity changes per time interval, respectively. In this study we propose an analytical modification to the WIA so that it can be used to study waves in conditions of time varying elastic properties, such as the left ventricle (LV) during diastole. The approach is first analytically elaborated for a one-dimensional elastic tube-model of the left ventricle with a time-dependent pressure-area relationship. Data obtained with a validated quasi-three dimensional axi-symmetrical model of the left ventricle are employed to demonstrate this new approach. Along the base-apex axis close to the base wave intensity curves are obtained, both using the standard method and the newly proposed modified method. The main difference between the standard and modified wave intensity pattern occurs immediately after the opening of the mitral valve. Where the standard WIA shows a backward expansion wave, the modified analysis shows a forward compression wave. The proposed modification needs to be taken into account when studying left ventricular relaxation, as it affects the wave type.

Validation of a fluid-structure interaction model of a heart valve using the dynamic mesh method in fluent.

Simulations of coupled problems such as fluid-structure interaction (FSI) are becoming more and more important for engineering purposes. This is particularly true when modeling the aortic valve, where the FSI between the blood and the valve determines the valve movement and the valvular hemodynamics. Nevertheless only a few studies are focusing on the opening and closing behavior during the ejection phase (systole). In this paper, we present the validation of a FSI model using the dynamic mesh method of Fluent for the two-dimensional (2D) simulation of mechanical heart valves during the ejection phase of the cardiac cycle. The FSI model is successfully validated by comparing simulation results to experimental data obtained from in vitro studies using a CCD camera.

Influence of head position on carotid hemodynamics in young adults.

Studies in adults have shown marked changes in geometry and relative positions of the carotid arteries when rotating the head. The aim of this study was to quantify the change in geometry and analyze its effect on carotid hemodynamics as a result of head rotation. The right carotid arteries of nine young adult subjects were investigated in supine position with straight and left turned head positions, respectively. The three-dimensional (3D) carotid geometry was reconstructed by using 3D ultrasound (3D US), and the carotid hemodynamics were calculated by combining 3D US with computational fluid dynamics. It was observed that cross-sectional areas and shapes did not change markedly with head rotation, but carotid vessel center lines altered with planarification of the common carotid artery as a main feature (P < 0.05). Measured common carotid flow rates changed significantly at the individual level when the head was turned, but on the average, the change in mean common carotid flow rate was relatively small (0.37 +/- 1.11 ml/s). The effect of the altered center lines and flow rates on the atherogenic nature of the carotid bifurcation was evaluated by using calculated hemodynamic wall parameters, such as wall shear stress (WSS) and oscillatory shear index (OSI). It was found that WSS and OSI patterns changed significantly with head rotation, but the variations were very subject dependent and could not have been predicted without assessing the altered geometry and flow of the carotid bifurcation for individual cases. This study suggests that there is a need for standardization of the choice of head position in the 3D US scan protocol, and that carotid stents and emboli diverters should be studied in different head positions.

The integration of medical imaging and computational fluid dynamics for measuring wall shear stress in carotid arteries.

The link between atherosclerosis and wall shear stress (WSS) has lead to considerable interest in the in vivo estimation of WSS. Both magnetic resonance imaging (MRI) and three-dimensional ultrasound (3DUS) are capable of providing the anatomical and flow data required for subject-specific computational fluid dynamics (CFD) simulations. This study compares, for the first time, predicted 3D flow patterns based on black blood MRI and 3DUS. Velocity fields in the carotid arteries of nine subjects have been reconstructed, and the haemodynamic wall parameters WSS, oscillatory shear index (OSI), WSS gradients (WSSG) and angle gradients (WSSAG) were computed and compared. There was a good qualitative agreement between results derived from MRI and 3DUS, embodied by a strong linear correlation between the patched representations of the haemodynamic wall parameters. The root-mean-square error between haemodynamic wall parameters was comparable to the range of the expected variability of each imaging technique (WSS: 0.411 N/m; OSI: 0.048; temporal WSSG: 2.29 N/(s.m/sup 2/); spatial WSSG: 150 N/m/sup 3/; WSSAG: 87.6 rad/m). In conclusion, MRI and 3DUS are comparable techniques for combining with CFD in the carotid artery. The relatively high cost of MRI favour 3DUS to MRI for future haemodynamic studies of superficial arteries.

Image-based carotid flow reconstruction: a comparison between MRI and ultrasound.

Atherosclerosis is a major cause of morbidity and mortality. Its apparent link with wall shear stress (WSS) has led to considerable interest in the in vivo estimation of WSS. Determining WSS by combining medical images with computational fluid dynamics (CFD) simulations can be performed both with magnetic resonance imaging (MRI) and three-dimensional ultrasound (3DUS). This study compares predicted 3D flow patterns based on black blood MRI and 3DUS. Velocity fields in the carotid arteries of nine subjects have been reconstructed, and the haemodynamic wall parameters WSS, oscillatory shear index (OSI), WSS gradients (WSSG) and angle gradients (WSSAG) were compared between the two imaging techniques. There was a good qualitative agreement between results derived from MRI and 3DUS (average correlation strength above 0.60). The root mean square error between haemodynamic wall parameters was comparable to the range of the expected variability of each imaging technique (WSS: 0.411 N m(-2); OSI: 0.048; temporal WSSG: 150 N s(-1) m(-2); spatial WSSG: 2.29 N m(-3); WSSAG: 87.6 rad m(-1)). In conclusion, MRI and 3DUS are capable of providing haemodynamic parameters when combined with CFD, and the predictions are in most cases qualitatively and quantitatively similar. The relatively high cost of MRI and continuing improvement in ultrasound favour US to MRI for future haemodynamic studies of superficial arteries.

Reproducibility study of magnetic resonance image-based computational fluid dynamics prediction of carotid bifurcation flow.

The importance of shear stress in the initiation and progression of atherosclerosis has been recognized for some time. A novel way to quantify wall shear stress under physiologically realistic conditions is to combine magnetic resonance imaging (MRI) and computational fluid dynamics. The present study aims to investigate the reproducibility of the simulated flow by using this combined approach. The right carotid bifurcations of eight healthy subjects were scanned twice with MRI within a few weeks. Three-dimensional geometries of the vessels were reconstructed for each scan and each subject. Pulsatile flows through these models were calculated to assess errors associated with the predicted flow parameters. This was done by comparing various wall shear stress indices, including the time-averaged wall shear stress (WSS), oscillating shear index (OSI), WSS Gradients (WSSG) and WSS Angle Deviation (WSSAD). Qualitatively, all the wall shear parameters proved to be highly reproducible. Quantitatively, the reproducibility was over 90% for OSI and WSSAD, but less impressive (60%) for other parameters. Our results indicated that WSS and WSSG values were extremely sensitive to subtle variations in local geometry and mesh design, particularly in regions around the bifurcation apex where WSS values were high and least reproducible.

Carotid geometry reconstruction: a comparison between MRI and ultrasound.

Image-based Computational Fluid Dynamics (CFD) has become a popular tool for the prediction of in vivo flow profiles and hemodynamic wall parameters. Currently, Magnetic Resonance Imaging (MRI) is most widely used for in vivo geometry acquisition. For superficial arteries such as the carotids and the femoral artery, three-dimensional (3-D) extravascular ultrasound (3-DUS) could be a cost-effective alternative to MRI. In this study, nine healthy subjects were scanned both with MRI and 3-DUS. The reconstructed carotid artery geometries for each subject were compared by evaluating cross-sectional areas, centerlines, and carotid nonplanarity. Lumen areas agreed very well between the two different acquisition techniques, whereas centerlines and nonplanarity parameters showed measurable disagreement, possibly due to the different neck and head positions adopted for 3-DUS versus MRI. With the current level of agreement achieved, 3-DUS has the potential to become an inexpensive and fast alternative to MRI for image-based CFD modeling of superficial arteries.

Mass transfer characteristics of artificial lungs.

An artificial lung is used during cardiopulmonary bypass to oxygenate blood and control blood temperature. The oxygen transfer rate-flow rate characteristics of three hollow fiber membrane artificial lungs (Sarns Turbo 440, Cobe Optima, Dideco Compactflo) were determined in vitro to characterize design features. Results are presented as a unique dimensionless relationship between Sherwood number, NSh (ratio of convective to diffusive mass transfer), Schmidt number, NSc (ratio of momentum to diffusive transport), and Reynolds number, NRe (ratio of inertial to viscous forces). This relationship is a function of device porosity, epsilon, and characteristic device length, xi, defined as the ratio of the mean blood path and manifold length: Nsh/NSc(1/3) x xi(1/2) = phi x (epsilon(1/m) x NRe)(m) where phi = 0.26 and m = 1.00 for NPe < 3,200 and phi = 0.47 and m = 0.64 for NPe > 3,200 where NPe is the dimensionless Péclet number defined as NRe x NSc. We found good correspondence between the model predictions and in vitro blood oxygen transfer rates. We conclude that this dimensionless approach allows us (1) to compare artificial lungs independently, (2) to relate water tests to blood, and (3) to predict the oxygen transfer rate of a new artificial lung design.

In vitro study of FFR, QCA, and IVUS for the assessment of optimal stent deployment.

We tested whether fractional flow reserve (FFR) discriminates between suboptimally and optimally deployed stents. Latex tubes (diameter solidus in circle = 4 mm) with diameter stenosis 40% (n = 3), 50% (n = 3) and 60% (n = 3) were tested in a pulsatile flow system, using water. Measurements were done at baseline (n = 9; FFR/QCA) and after suboptimal (SOD; 3-mm balloon at 8 atm) and optimal (OD; 4 mm balloon at 16 atm) deployment of a 35-mm stent (n = 6; FFR/QCA/IVUS). Varying Q from 150 to 50 ml/min increased FFR by 2-7%. Conversely, at 100 ml/min, FFR increased by only 0.8% from SOD to OD (P < 0.05). Extrapolating data to blood flow, the gain in FFR from SOD to OD is less than 5% for Q = 100 ml/min, while FFR may increase by 15-20% by changes in blood flow from 50 to 150 ml/min. We conclude that IVUS and QCA are more appropriate for the assessment of optimal stent deployment.

Two-dimensional finite element model for oxygen transfer in cross-flow hollow fiber membrane artificial lungs.

A predictive, two-dimensional model with good absolute accuracy for flow and mass transfer in cross-flow hollow fiber membrane artificial lungs is developed. The proposed model is able to predict the gas transfer to water flowing outside and perpendicular to hollow fibers in the artificial lung. The model uses a finite element technique to solve the Navier-Stokes equations and the convection-diffusion equation on the computational domain of a unit fiber cell. Subsequent stream-wise and cross-wise unit fiber cells are then coupled/assembled to the relationship between the oxygen transfer rate and flow rate of a cross-flow hollow fiber membrane artificial lung. The model is compared to experimental water data obtained by perfusing three commercial artificial lungs with water.

Assessment of the time constant of relaxation: insights from simulations and hemodynamic measurements.

The objective of this study was to use high-fidelity animal data and numerical simulations to gain more insight into the reliability of the estimated relaxation constant derived from left ventricular pressure decays, assuming a monoexponential model with either a fixed zero or free moving pressure asymptote. Comparison of the experimental data with the results of the simulations demonstrated a trade off between the fixed zero and the free moving asymptote approach. The latter method more closely fits the pressure curves and has the advantage of producing an extra coefficient with potential diagnostic information. On the other hand, this method suffers from larger standard errors on the estimated coefficients. The method with fixed zero asymptote produces values of the time constant of isovolumetric relaxation (tau) within a narrow confidence interval. However, if the pressure curve is actually decaying to a nonzero pressure asymptote, this method results in an inferior fit of the pressure curve and a biased estimation of tau.

Blood flow around hollow fibers.

In an artificial lung, blood is oxygenated and flows around a bundle of hollow fibers while gas flows inside the fiber. The objective of this study is to understand the hydrodynamics of three different fiber banks (inline square IS, staggered square SS and equilateral triangle ET) and to investigate the influence of both a Newtonian and non-Newtonian Casson viscosity model on the flow field. A two-dimensional finite element model for permanent, isothermal, laminar blood flow perpendicular to hollow fibers is used. All fibers are assumed identical, straight and parallel. Porosity ranges from 0.4 to 0.6 and Reynolds number varies from 1 to 60. For a given Re, ET generates less resistance than SS, the latter being comparable with IS. A lower porosity increases resistance. Non-Newtonian viscosity affects velocity patterns only at low Re (<0.5) and higher porosity (>0.5). Resistance at low Re is significantly elevated in the fiber banks due to an overall increase in viscosity. This model makes it possible to study the influence of geometry and viscosity on hydrodynamics in fiber banks and may aid in the optimization of hollow fiber artificial lung design.

Hydrodynamic characteristics of artificial lungs.

An artificial lung is used during cardiopulmonary bypass to oxygenate blood and to control blood temperature. The pressure drop-flow rate characteristics of the membrane compartment in three hollow fiber membrane oxygenators were determined in vitro to characterize design features. Results are presented in a unique dimensionless relationship between Euler number, N(Eu) (ratio of pressure drop to kinetic energy), and Reynolds number, N(Re) (ratio of inertial to viscous forces), and are a function of the device porosity, epsilon, and a characteristic device length, xi, defined as the ratio of the mean blood path and manifold length: [equation in text]. This dimensionless approach allows us (1) to compare oxygenators independently, and (2) to relate water tests to blood.