Objective: It has been proposed that the early diastolic flow pattern
inside the left ventricle reflects ventricular diastolic function. The
flow pattern often divides into two phases towards the apex. The aim
of this study was to examine the hydrodynamic nature of this phenomeno
n. Methods: A rubber balloon representing the ventricle was connected
to a reservoir representing the atrium, both filled with anticoagulate
d blood. Rigid ''mitral'' orifices 20 mm long and 15-40 mm in diameter
were used. Surrounding the ventricle was a water filled chamber to wh
ich suction could be applied. The colour M mode Doppler sample beam co
incided with the ventricular long: axis. Colour M mode measures veloci
ty at multiple sites along the beam simultaneously. The timing and mag
nitude of the velocities were analysed digitally. The filling was also
qualitatively studied by ultrasonic two dimensional sector scanning.
Results: At the start of the filling, blood moved simultaneously at al
l levels, behaving as a fluid column. This was denoted ''phase I''. A
how wave then propagated from the mitral orifice towards the apex, cal
led ''phase II''. This was found to represent a ring vortex, with bloo
d velocities twice its propagation velocity [ratio 2.1(SEM 0.016)]. Th
e ratio between the velocity time integrals of phase II and phase I de
creased progressively from 51(15) to 0.41(0.14) when mitral orifice di
ameter was increased in 5 mm steps from 15 mm to 40 mm (p<0.001). The
propagation velocity correlated strongly with peak transmitral blood v
elocity, r=0.95, p<0.001. The flow patterns resembled patterns recorde
d in patients. Conclusions: The two phases of the filling pattern repr
esented the motion of a blood column and the propagation of a ring vor
tex, respectively. Mitral orifice size determined which phase dominate
d the flow pattern.