SUMMARYFor a better understanding of the characteristics of blood flow in the pulmonary artery, we constructed three-dimensional images of velocity profiles of blood flow in the pulmonary artery from pulsed Doppler ultrasound recordings in 14 lambs aged 28-40 days. In 8 lambs, pulmonary hypertension was created by the central venous injection of monocrotaline pyrrole. Six lambs served as unaltered controls. The velocity data were sampled in 2mm increments along both an anterior-posterior axis and a right-left orthogonal axis in the main pulmonary artery. Using a computer-generated cross-sectional velocity matrix consisting of 0.25 mm square grids, the velocity of blood flow was estimated at each intersection. The cross-sectional velocity matrices were generated at 5 msec intervals during the entire cardiac cycle. In all animals, significant velocity reversal was detected near the posterior wall. In 7 of 14 animals, the peak forward velocity was located near the posterior wall. Three of 8 hypertensive models showed reacceleration during the mid-systolic phase at the center of the velocity waveform, but one reacceleration disappeared at a point only 2mm away from the center of the vessel toward the posterior wall. Acceleration time correlated well with the mean pulmonary arterial pressure (PAP) (r=-0.85) and the log10 PAP (r=-0.86).Corrected acceleration time (acceleration time divided by the square root of the cardiac cycle length) also correlated with PAP (r=-0.78) and the log10 PAP (r=-0.81).

Ultrasonic measurement of blood flow within the main pulmonary artery (MPA) requires a precise knowledge of the mean blood velocity within this vessel and the cross-sectional area (CSA). Small conformational changes in the elliptical shape of the MPA have substantial effects on the calculation of CSA and, subsequently, flow. We examined the extent of these changes by measuring the pulsatile and mean elliptical dimensions of the MPA in nine anaesthetised, open-chested, mechanically ventilated mongrel dogs using two pairs of 10 MHz ultrasonic, piezoelectric crystals. These custom-made devices were sutured to the PA adventitia along the long and short cross-sectional axes 2 cm distal to the pulmonary valve. Axial dimensions were collected during normal, elevated (via noradrenaline and fluid additions) and reduced (via exsanguination) PA pressures. We confirmed the linear pressure/diameter response in 15/18 axial data sets (r greater than 0.80). Further, the linear axial responses of the long and short diameters were parallel (7/9, p less than 0.05) and have different zero-pressure intercepts (7/9, p less than 0.0001). A mathematical consequence of this parallelism is predictable, although non-constant, eccentricity. Finally, error analysis of multi-axial measurement techniques were shown to improve CSA accuracy by as much as 50% when compared with uni-axial determinations.

No abstract

No abstract

No abstract