Multicolor particle shadow accelerometry.

This paper describes the extension of multicolor particle shadow velocimetry (CPSV) to the measurement of local acceleration in an Eulerian frame of reference. A validation experiment was conducted on a pendulous disk undergoing unsteady rigid body rotation. Angular velocity and acceleration profiles by CPSA are presented along with a comparison to recordings by an accelerometer mounted on the pendulum. CPSA is also demonstrated in a fully-developed turbulent pipe flow. Profiles of standard deviation of the local acceleration in the near wall region (0< y + <75) are compared to similar measurements by Christensen and Adrian. A favorable comparison is found between CPSA and particle image accelerometry (PIA). The effect of acceleration time delay, or the time between two velocity estimates, on local acceleration estimates is discussed.

Correcting for color crosstalk and chromatic aberration in multicolor particle shadow velocimetry.

Color crosstalk and chromatic aberration can bias estimates of fluid velocity measured by color particle shadow velocimetry (CPSV), using multicolor illumination and a color camera. This article describes corrections to remove these bias errors, and their evaluation. Color crosstalk removal is demonstrated with linear unmixing. It is also shown that chromatic aberrations may be removed using either scale calibration, or by processing an image illuminated by all colors simultaneously. CPSV measurements of a fully developed turbulent pipe flow of glycerin were conducted. Corrected velocity statistics from these measurements were compared to both single-color PSV and LDV measurements and showed excellent agreement to fourth-order, to well into the viscous sublayer. Recommendations for practical assessment and correction of color aberration and color crosstalk are discussed.

Improved in vitro quantification of the force exerted by the papillary muscle on the left ventricular wall: three-dimensional force vector measurement system.

Recent developments indicate that the forces acting on the papillary muscles can be a measure of the severity of mitral valve regurgitation. Pathological conditions, such as ischemic heart disease, cause changes in the geometry of the left ventricle and the mitral valve annulus, often resulting in displacement of the papillary muscles relative to the annulus. This can lead to increased tension in the chordae tendineae. This increased tension is transferred to the leaflets, and can disturb the coaptation pattern of the mitral valve. The force balance on the individual components governs the function of the mitral valve. The ability to measure changes in the force distribution from normal to pathological conditions may give insight into the mechanisms of mitral valve insufficiency. A unique in vitro model has been developed that allows quantification of the papillary muscle spatial position and quantification of the three-dimensional force vector applied to the left ventricular wall by the papillary muscles. This system allows for the quantification of the global force exerted on the posterior left ventricular wall from the papillary muscles during simulation of normal and diseased conditions.

Observation and quantification of gas bubble formation on a mechanical heart valve.

Clinical studies using transcranial Doppler ultrasonography in patients with mechanical heart valves (MHV) have detected gaseous emboli. The relationship of gaseous emboli release and cavitation on MHV has been a subject of debate in the literature. To study the influence of cavitation and gas content on the formation and growth of stable gas bubbles, a mock circulatory loop, which employed a Medtronic-Hall pyrolytic carbon disk valve in the mitral position, was used. A high-speed video camera allowed observation of cavitation and gas bubble release on the inflow valve surfaces as a function of cavitation intensity and carbon dioxide (CO2) concentration, while an ultrasonic monitoring system scanned the aortic outflow tract to quantify gas bubble production by calculating the gray scale levels of the images. In the absence of cavitation, no stable gas bubbles were formed. When gas bubbles were formed, they were first seen a few milliseconds after and in the vicinity of cavitation collapse. The volume of the gas bubbles detected in the aortic track increased with both increased CO2 and increased cavitation intensity. No correlation was observed between O2 concentration and bubble volume. We conclude that cavitation is an essential precursor to stable gas bubble formation, and CO2, the most soluble blood gas, is the major component of stable gas bubbles.

The osmotic swelling characteristics of cardiac valve prostheses.

Several types of mechanical cardiac prostheses have been constructed with Delrin occluders, a material that is subject to osmotic swelling. The leaftets are designed to expand to specific tolerances when immersed in blood. The synthetic blood analogs commonly used in vitro contain hydrophilic compounds that can alter the osmotic expansion of the Delrin occluders. A static leak test chamber was employed to illustrate the effects of various test fluids on the sustained regurgitation phase of Delrin valves.

Papillary muscle misalignment causes multiple mitral regurgitant jets: an ambiguous mechanism for functional mitral regurgitation.

BACKGROUND AND AIMS OF THE STUDY: The study aim was to test the hypothesis that asymmetric alignment (misalignment) of the papillary muscles is sufficient to cause incomplete mitral leaflet coaptation and functional mitral regurgitation (MR). METHODS: Different spatial relationships between the papillary muscles and the mitral annulus were investigated in isolated porcine mitral valves in vitro to assess the impact on mitral valve competence. The systolic occlusional leaflet area (OLA) needed to cover the mitral orifice and the anterolateral (ACOM) and posteromedial (PCOM) commissural portion (OLA(ACOM), OLA(PCOM)) were assessed by 2D echocardiography to quantitate incomplete mitral leaflet coaptation. The regurgitant fraction (RF) and MR jet location were assessed by a flow meter and color Doppler ultrasound. RESULTS: Posterolateral dislocation of the posteromedial papillary muscle impaired mitral leaflet coaptation at the corresponding half-portion of the mitral orifice (OLA(PCOM): 351-397 mm2 versus 296 mm2 (normal); p < 0.001) and modified the contralateral part (OLA(ACOM): 354-387 mm2 versus 304 mm2 (normal); p <0.001). The mitral leaflet coaptation line moved in apical and posterior directions, creating a commissural MR orifice at the PCOM side. At the ACOM side, anterior leaflet prolapse and restricted posterior leaflet mobility created an additional commissural regurgitant jet (RF = 0.11-0.13). Symmetrical papillary muscle misalignment restricted mitral leaflet mobility on both sides of the orifice in a synergistic manner (OLA(PCOM): 416-459 mm2 and OLA(ACOM): 427-489 mm2; both p <0.001 versus normal). The central MR jet orifice, which extended towards both commissures, caused more significant MR (RF = 0.15-0.26). CONCLUSIONS: Papillary muscle misalignment caused mitral regurgitant jet ambiguity with an anterior MR jet location following posteromedial papillary muscle displacement. These findings may improve understanding of the relation between myocardial lesion and mitral regurgitant jet location and thereby facilitate rational strategies for valvular interventions.

Chordal force distribution determines systolic mitral leaflet configuration and severity of functional mitral regurgitation.

OBJECTIVES: The purpose of this study was to investigate the impact of the chordae tendineae force distribution on systolic mitral leaflet geometry and mitral valve competence in vitro. BACKGROUND: Functional mitral regurgitation is caused by changes in several elements of the valve apparatus. Interaction among these have to comply with the chordal force distribution defined by the chordal coapting forces (F(c)) created by the transmitral pressure difference, which close the leaflets and the chordal tethering forces (FT) pulling the leaflets apart. METHODS: Porcine mitral valves (n = 5) were mounted in a left ventricular model where leading edge chordal forces measured by dedicated miniature force transducers were controlled by changing left ventricular pressure and papillary muscle position. Chordae geometry and occlusional leaflet area (OLA) needed to cover the leaflet orifice for a given leaflet configuration were determined by two-dimensional echo and reconstructed three-dimensionally. Occlusional leaflet area was used as expression for incomplete leaflet coaptation. Regurgitant fraction (RF) was measured with an electromagnetic flowmeter. RESULTS: Mixed procedure statistics revealed a linear correlation between the sum of the chordal net forces, sigma[Fc - FT]S, and OLA with regression coefficient (minimum - maximum) beta = -115 to -65 [mm2/N]; p < 0.001 and RF (beta = -0.06 to -0.01 [%/N]; p < 0.001). Increasing FT by papillary muscle malalignment restricted leaflet mobility, resulting in a tented leaflet configuration due to an apical and posterior shift of the coaptation line. Anterior leaflet coapting forces increased due to mitral leaflet remodeling, which generated a nonuniform regurgitant orifice area. CONCLUSIONS: Altered chordal force distribution caused functional mitral regurgitation based on tented leaflet configuration as observed clinically.

Integrated mechanism for functional mitral regurgitation: leaflet restriction versus coapting force: in vitro studies.

BACKGROUND: Functional mitral regurgitation in patients with ischemic or dilated ventricles has been related to competing factors: altered tension on the leaflets due to displacement of their papillary muscle and annular attachments, which restricts leaflet closure, versus global ventricular dysfunction with reduced transmitral pressure to close the leaflets. In vivo, however, geometric changes accompany dysfunction, making it difficult to study these factors independently. Functional mitral regurgitation also paradoxically decreases in midsystole, despite peak transmitral driving pressure, suggesting a change in the force balance acting to create a regurgitant orifice, with rising transmitral pressure counteracting forces that restrict leaflet closure. In vivo, this mechanism cannot be tested independently of annular contraction that could also reduce midsystolic regurgitation. METHODS AND RESULTS: An in vitro model was developed that allows independent variation of papillary muscle position, annular size, and transmitral pressure, with direct regurgitant flow rate measurement, to test the hypothesis that functional mitral regurgitation reflects an altered balance of forces acting on the leaflets. Hemodynamic and echocardiographic measurements of excised porcine valves were made under physiological pressures and flows. Apical and posterolateral papillary muscle displacement caused decreased leaflet mobility and apical leaflet tethering or tenting with regurgitation, as seen clinically. It reproduced the clinically observed midsystolic decrease in regurgitant flow and orifice area as transmitral pressure increased. Tethering delayed valve closure, increased the early systolic regurgitant volume before complete coaptation, and decreased the duration of coaptation. Annular dilatation increased regurgitation for any papillary muscle position, creating clinically important regurgitation; conversely, increased transmitral pressure decreased regurgitant orifice area for any geometric configuration. CONCLUSIONS: The clinically observed tented-leaflet configuration and dynamic regurgitant orifice area variation can be reproduced in vitro by altering the three-dimensional relationship of the annular and papillary muscle attachments of the valve so as to increase leaflet tension. Increased transmitral pressure acting to close the leaflets decreases the regurgitant orifice area. These results are consistent with a mechanism in which an altered balance of tethering versus coapting forces acting on the leaflets creates the regurgitant orifice.

An automated method for analysis and visualization of laser Doppler velocimetry data.

The analysis and visualization of large data sets collected by use of laser Doppler velocimetry has presented a challenge to researchers using this technique to investigate complex flow fields. This paper describes an automated procedure for analysis and animation of two- and three-dimensional laser Doppler velocimetry data. The procedure consists of a suite of FORTRAN programs for calculating phase window averages of velocity and the Reynolds stress tensor, calculating the principal normal stresses, maximum shear stresses, and preparation of data files for input into Plot-3D compatible data visualization software. An example application of these techniques to data collected from an in vitro investigation of the retrograde flow field associated with a bileaflet mechanical heart valve is also presented.

Velocity measurements and flow patterns within the hinge region of a Medtronic Parallel bileaflet mechanical valve with clear housing.

BACKGROUND AND AIMS OF THE STUDY: During recent clinical trials the Medtronic Parallel bileaflet mechanical heart valve was found to have an unacceptable number of valves with thrombus formation when implanted in the mitral position. Thrombi were observed in the hinge region and also in the upstream portion of the valve housing in the vicinity of the hinge. It was hypothesized that the flow conditions inside the hinge may have contributed to the thrombus formation. METHODS: In order to investigate the flow structures within the hinge, laser Doppler anemometry (LDA) measurements were conducted in both steady and pulsatile flow at approximately 70 predetermined sites within the hinge region of a 27 mm Medtronic Parallel mitral valve with transparent housing. The pulsatile flow velocity measurements were animated in time using a graphical software package to visualize the hinge flow field throughout the cardiac cycle. RESULTS: The LDA measurements revealed that mean forward flow velocities through the hinge region were on the order of 0.10-0.20 m/s. In the inflow channel, a large vortical structure was present during diastole. Upon valve closure, peak reverse velocity reached 3 m/s close to the housing wall in the inflow channel. This area also experienced high turbulent shear stresses (> 6000 dynes/cm2) during the leakage flow phase. A disturbed, vortical flow was again present in the inflow channel after valve closure, while slightly above the leaflet peg and relief the flow was essentially stagnant. The high turbulent stresses near the top of the inflow channel, combined with a persistent vortex, implicate the inflow channel of the hinge as a likely region of thrombus formation. CONCLUSIONS: This experimental investigation revealed zones of flow stagnation in the inflow region of the hinge throughout the cardiac cycle and elevated turbulent shear stress levels in the inflow region during the leakage flow phase. These fluid mechanic phenomena are most likely a direct result of the complex geometry of the hinge of this valve. Although the LDA measurements were conducted at only a limited number of sites within the hinge, these results suggest that the hinge design can significantly affect the washout capacity and thrombogenic potential of the Medtronic Parallel bileaflet mechanical heart valve. The use of LDA within the confines of the hinge region of a mechanical heart valve is a new application, made possible by recent advances in manufacturing technologies and a proprietary process developed by Medtronic that allowed the production of a transparent valve housing. Together, these modalities represent a new method by which future valve designs can be assessed before clinical trials are initiated.

An in vitro investigation of the retrograde flow fields of two bileaflet mechanical heart valves.

BACKGROUND AND AIM OF THE STUDY: Fluid stresses occurring in retrograde flow fields during valve closure may play a significant role in thrombogenesis. The squeeze flow and regurgitant jets can cause damage to formed blood elements due to high levels of turbulent shear stress. The aim of this study was to characterize in detail the spatial structure and temporal behavior of the retrograde flow fields of the St. Jude Medical and Medtronic Parallel bileaflet mechanical heart valves. METHODS: Three-component, coincident laser Doppler anemometry (LDA) velocity measurements were obtained facilitating the determination of the full Reynolds stress tensor and the principal stresses in the valve flow fields. The experiments were performed in the Georgia Tech aortic flow chamber under physiologic pulsatile flow conditions. Data were collected over several hundred cardiac cycles for subsequent phase window averaging and generation of mean velocity and turbulence statistics over 20 ms intervals. A region approximately 8 mm x 10 mm was mapped 1.0 mm upstream of one hinge of each valve with an incremental resolution of 0.13-0.25 mm. Animation of the data allowed the visualization of the flow fields and a quantitative display of mean velocity and turbulent stress values. RESULTS: In the St. Jude Medical squeeze flow, the peak turbulent shear stress was 800 dynes/cm2 and the peak reverse velocity was 0.60 m/s. In the Medtronic Parallel squeeze flow, the peak turbulent shear stress was 1,000 dynes/cm2 and the peak velocity 0.70 m/s. The leakage jet fields of the two valves were very different: in the case of the St. Jude Medical valve, turbulent shear stresses reached 1,800 dynes/cm2 and peak jet velocity was 0.80 m/s; in the case of the Medtronic Parallel valve, turbulent shear stresses reached 3,690 dynes/cm2 and the peak jet velocity was 1.9 m/s. CONCLUSIONS: The retrograde flow fields of these two bileaflet mechanical heart valves appear to be design-dependent. The elevated turbulent shear stresses generated by both valve designs may indicate a propensity for blood element damage during the reverse flow phase of the cardiac cycle, but the extent of flow disturbance was twice as high with the Medtronic Parallel than with the St. Jude Medical valve. This research should yield a better understanding of the significance of retrograde flow to the functionality and potential thrombogenicity of bileaflet mechanical heart valves and aid in the development of new designs.

Experimental analysis of fluid mechanical energy losses in aortic valve stenosis: importance of pressure recovery.

Current methods for assessing the severity of aortic stenosis depend primarily on measures of maximum systolic pressure drop at the aortic valve orifice and related calculations such as valve area. It is becoming increasingly obvious, however, that the impact of the obstruction on the left ventricle is equally important in assessing its severity and could potentially be influenced by geometric factors of the valve, causing variable degrees of downstream pressure recovery. The goal of this study was to develop a method for measuring fluid mechanical energy losses in aortic stenosis that could then be directly related to the hemodynamic load placed on the left ventricle. A control volume form of conservation of energy was theoretically analyzed and modified for application to aortic valve stenosis measurements. In vitro physiological pulsatile flow experiments were conducted with different types of aortic stenosis models, including a venturi meter, a nozzle, and 21-mm Medtronic-Hall tilting disc and St. Jude bileaflet mechanical valves. The energy loss created by each model was measured for a wide range of experimental conditions, simulating physiological variation. In all cases, there was more energy lost for the nozzle (mean = 0.27 J) than for any other model for a given stroke volume. The two prosthetic valves generated approximately the same energy losses (mean = 0.18 J), which were not statistically different, whereas the venturi meter had the lowest energy loss for all conditions (mean = 0.037 J). Energy loss correlated poorly with orifice pressure drop (r2 = 0.34) but correlated well with recovered pressure drop (r2 = 0.94). However, when the valves were considered separately, orifice and recovered pressure drop were both strongly correlated with energy loss (r2 = 0.99, 0.96). The results show that recovered pressure drop, not orifice pressure drop, is directly related to the energy loss that determines pump work and therefore is a more accurate measure of the hemodynamic significance of aortic stenosis.

Identification of peak stresses in cardiac prostheses. A comparison of two-dimensional versus three-dimensional principal stress analyses.

This study assessed the accuracy of using a two-dimensional principal stress analysis compared to a three-dimensional analysis in estimating peak turbulent stresses in complex three-dimensional flows associated with cardiac prostheses. Three-component, coincident laser Doppler anemometer measurements were obtained in steady flow downstream of three prosthetic valves: a St. Jude bileaflet, Bjork-Shiley monostrut tilting disc, and Starr-Edwards ball and cage. Two-dimensional and three-dimensional principal stress analyses were performed to identify local peak stresses. Valves with locally two-dimensional flows exhibited a 10-15% underestimation of the largest measured normal stresses compared to the three-dimensional principal stresses. In nearly all flows, measured shear stresses underestimated peak principal shear stresses by 10-100%. Differences between the two-dimensional and three-dimensional principal stress analysis were less than 10% in locally two-dimensional flows. In three-dimensional flows, the two-dimensional principal stresses typically underestimated three-dimensional values by nearly 20%. However, the agreement of the two-dimensional principal stress with the three-dimensional principal stresses was dependent upon the two velocity-components used in the two-dimensional analysis, and was observed to vary across the valve flow field because of flow structure variation. The use of a two-dimensional principal stress analysis with two-component velocity data obtained from measurements misaligned with the plane of maximum mean flow shear can underpredict maximum shear stresses by as much as 100%.

Haemodynamic and echocardiographic characteristics of a stentless allograft mitral prosthesis: an in vitro study.

Poor long-term durability and impaired haemodynamic performance are known disadvantages of bioprosthetic heart valves when compared to valve replacement using aortic allografts. A new stentless allograft mitral implant was developed and tested in vitro in a left ventricular model and pulsatile flow system to evaluate hydrodynamic function. Mitral valves were excised from sheep hearts and the mitral annulus reinforced by a strip of ovine pericardium. A patch of expanded polytetrafluoroethylene (ePTFE) was placed above the tips of the remaining papillary muscles. For in vitro evaluation of a total of five valves were investigated in a pulse duplicator. Transvalvular pressure gradients (delta P) were measured over a flow range corresponding to a cardiac output of 5l/min, at a heart rate of 70 beats/min, with a systole accounting for approximately 35% of the cardiac cycle. The systolic ejection period and diastolic filling period in this model were 350 and 510 ms, respectively, and aortic pressure was 120/80 mmHg. The effective orifice area was calculated from measurements of mean pressure drop and root mean square flow. Additionally, valve performance was evaluated by Doppler echocardiography. Results of in vitro studies of a 25 mm stentless allograft mitral implant, which is similar to the valves implanted in a chronic weaning sheep model, revealed a mean(s.d.) delta P of 2.0(1.6) mmHg (range 1.0-4.9 mmHg). The mean calculated effective orifice area was 3.38(0.52) cm(2) (range 2.5-3.8 cm(2)). Doppler echocardiography showed excellent performance of the mitral valve components and valve competence could be achieved. During the in vitro studies no failure caused by tissue rupture was detected. The results of the in vitro studies revealed data for delta P and effective orifice area superior to data obtained for standard 25 mm porcine bioprostheses.

In vitro assessment of prosthetic valve function in mitral valve replacement with chordal preservation techniques.

BACKGROUND AND AIM OF THE STUDY: The importance of chordal preservation techniques in maintaining improved left ventricular function after mitral valve replacement has been well documented clinically. Currently, the choice of prosthetic valve used in chordal preservation is dependent upon the surgeon's preference. However, the transvalvular flow characteristics of common, clinically used prosthetic valves may be influenced by the mitral subvalvular apparatus, and may result in degraded valve function. The goal of this study was to perform an in vitro evaluation of the influence of chordal preservation on the transvalvular and left ventricular flow patterns of common valve prostheses. METHODS: Tissue and mechanical valves have been evaluated under physiologic pulsatile flow with anterior and/or posterior chordal preservation. Flow patterns were assessed by 2-D planar flow visualization, pulsed wave Doppler velocity measurements, 2-D echocardiography, and selected color Doppler flow mapping. Based on changes in transvalvular and left ventricular flow patterns, favorable prosthetic valve/chordal preservation combinations were identified. Additionally, valve orientation was varied to determine optimal orientation. RESULTS: Baseline results without chordal preservation indicate that the anti-anatomic orientation is preferred for the bileaflet valve design while the tilting disc valve should be oriented with the major axis toward the posterior (free) wall of the ventricle, corroborating published conclusions by other investigators. Some form of flow restriction is observed in all test cases with chordal preservation due to the presence of the subvalvular tissue. In general, bioprostheses showed less flow restriction then the mechanical valves, particularly with lateral flow expansion. This flow restriction may influence pressure recovery downstream of the mechanical valves tested. Increased flow constriction is observed with anterior and posterior chordal preservation. CONCLUSIONS: This study favors the use of the St. Jude Medical bileaflet valve orientated in the anti-anatomic position, or the Carpentier-Edwards pericardial valve with chordal preservation.

The effect of aortic outflow on the quantification of mitral regurgitation by the flow convergence method.

The effect of aortic outflow on the quantification of mitral regurgitation by the flow convergence method was investigated by both in vitro experiments and computational simulations. Digital analysis of the color Doppler M-mode images was compared with results obtained with laser Doppler anemometry, an engineering gold standard, and three-dimensional computational simulations. Regurgitant orifices of 3.2 and 6.4 mm in diameter were used with instantaneous aortic flow rates from 0 to 500 ml/sec, corresponding to net cardiac outputs of 0 to 5 L/min. In the absence of aortic outflow, a clear plateau was observed in plots of the calculated flow rate as a function of the distance from the orifice, indicating that there was a zone in which the hemispheric assumption was valid. As the aortic outflow was increased, the length of this plateau region decreased and then disappeared at high aortic flow rates. Farther from the orifice, beyond the plateau zone, the flow rate was overestimated and this overestimation increased with increasing aortic flow rate. Results showed excellent agreement between in vitro experiments and computational stimulations. This study demonstrated that aortic outflow has a dramatic effect on the flow convergence region and therefore must be considered in flow rate calculations.

Comparison of magnetic resonance imaging and Laser Doppler Anemometry velocity measurements downstream of replacement heart valves: implications for in vivo assessment of prosthetic valve function.

BACKGROUND AND AIM OF THE STUDY: The non-invasive, in-vivo assessment of prosthetic valve function is compromised by the lack of accurate measurements of the transvalvular flow fields or hemodynamics by current techniques. Short echo time magnetic resonance imaging (MRI) may provide a method for the non-invasive, in vivo assessment of prosthetic valve function by accurately measuring changes in the transvalvular flow fields associated with normal and dysfunctional prosthetic valves. The objectives of these in vitro experiments were to investigate the potential for using MRI as a tool to measure the complex flow fields distal to replacement heart valves, and to assess the accuracy of MRI velocity measurements by comparison with Laser Doppler Anemometry (LDA), a gold standard. METHODS: The velocity fields downstream of tilting disc, bileaflet, ball and cage, and pericardial tissue valves were measured using both three-component LDA and MRI phase velocity encoding under a steady flow rate of 22.8 l/min, simulating peak systolic flow. The valves were tested under normal and stenotic conditions to assess the MRI capabilities under a wide range of local flow conditions, velocities and turbulence levels. A new short echo time MRI technique (FAcE), which allowed velocity measurements in stenotic jets with high turbulence, was tested. RESULTS: Good overall agreement was obtained between the MRI velocity measurements and the LDA data. The MRI velocity measurements adequately reproduced the spatial structure of the flow fields. In most cases peak velocities were accurately measured to within 15%. CONCLUSIONS: The results indicate that the FAcE MRI method has the potential to be used as a diagnostic tool to assess prosthetic valve function.

Hemodynamics of the Fontan connection: an in-vitro study.

The Fontan operation is one in which the right heart is bypassed leaving the left ventricle to drive the blood through both the capillaries and the lungs, making it important to design an operation which is hemodynamically efficient. The object here was to relate the pressure in Fontan connections to its geometry with the aim of increasing the hemodynamically efficiency. From CT or magnetic resonance images, glass models were made of realistic atrio-pulmonary (AP) and cavo-pulmonary (CP) connections in which the right atrium and/or ventricle are bypassed. The glass models were connected to a steady flow loop and flow visualization, pressure and 3 component LDA measurements made. In the AP model the large atrium and curvature of the conduit created swirling patterns, the magnitude of which was similar to the axial velocity. This led to an inefficient flow and a subsequent large pressure loss (780 Pa). In contrast, the CP connection with a small intra-atrial chamber had reduced swirling and a significantly smaller pressure loss (400 Pa at 8 l.min) and was therefore a more efficient connection. There were, however, still pressure losses and it was found that these occurred where there was a large bending of the flow, such as from the superior vena cava to the MPA and from the MPA to the right pulmonary artery.