1162 
MONITORING 
V 
c 
Figure 1. — Schematic representation of the hemi-an- 
nular averaging computation of the blood flow rate 
from the mean velocity profile. The velocity Vi at a 
distance ri from the vessel axis is assumed to be the 
average velocity on half a circle of radius ri, as rep- 
resented on a section of the vessel (above). 
Qo is the result of the hemiannular averaging 
computation of the mean velocity profile (Fig. 
1) . The degree of asymmetry of the mean veloc- 
ity profile is expressed in percent as the differ- 
ence between the computed flow for each half of 
the mean velocity profile divided by the total 
flow Qo. 
The experiment was carried out on normal 
dogs (mean body weight 23.5 ± SD 1.1 Kg), 
anesthetized, intubated and mechanically venti- 
lated. A laparotomy was performed. The ultra- 
sonic probes were placed on the hepatic artery 
and the portal vein. Snares were placed around 
the supra-hepatic veins. The hepatic arterial, 
portal, wedged supra-hepatic pressures were 
continuously monitored. Two groups of dogs 
were compared: normal dogs (n=15) and dogs 
with increased intra-hepatic resistances, ob- 
tained by the ligation of the right and the left 
supra-hepatic veins (n=15). The changes in 
pressures and flow due to the completion of a ' 
wide and distally located side to side porta- 
caval shunt were studied. ;i 
RESULTS i| 
I 
1. Flow pattern. The hepatic arterial flow | 
profile is never symmetrical in the different ^ 
experimental conditions. The degree of the 
asymmetry is related to changes in the direc- \ 
tion of the hepatic artery. The highest veloci- \ 
ties were located in the outer side of the ves- | 
sel's curvature. The increase of the arterial j 
velocity Vd following the ligation of the supra- j 
hepatic veins did not alter the velocity distri- 
bution in the hepatic artery. After side to side | 
portacaval shunt, in both normal dogs and dogs | 
with ligated supra-hepatic veins, the average j 
velocity Vd was increased but the shape of the 
velocity profile unchanged. 
In the portal vein, very different velocity pro- i 
files were observed in the various experimental I 
conditions. On the normal dog, in the plane de- 
fined by the portal vein and the gastro-duodenal 
vein two different aspects were observed with | 
an equal frequency: a curved profile (Fig. 2 A) | 
and a double humped profile (Fig. 3 left). This j 
latter shape was caused by the arrival of the 
gastroduodenal vein upstream from the ultra- \ 
sonic probe. It reveals a stream line effect in j 
the portal vein flow. The ligation of the supra- 
hepatic veins reduced the average velocity Vd | 
but did not alter the shape of the velocity profile I 
of the portal flow. The completion of the side | 
to side portacaval anastomosis decreased the 
average velocity Vd on both groups of dogs : in 
normal dogs, as Vd remains directed to the 
liver, the velocity profile showed a bidirectional ! 
flow pattern (Fig. 2D). Two different streams j 
of opposite direction, hepatopetal and hepato- | 
fugal, were juxtaposed in the vein. On dogs ' 
with ligated supra-hepatic veins, both average I 
velocity Vd and segmental velocities Vi were [ 
reversed, directed from the liver towards the 
shunt. The shape of the reversed mean veloc- 
ity profile was then flattened (Fig. 2C). 
2. Blood flotv rate. In the normal dog, the [ 
mean portal flow value Qo was 303 ±: SD 35 
ml/min and the hepatic arterial flow rate Qo 
was 107, ± 8 ml/min. In the group of dogs with 
