JOURNAL
OF
Vol. 41, No.
A PPLIED
PHYSIOLOC
6, December
Y
1976.
Printed
in U.S.A.
Venous ultrasound catheter-tip of arterial hemodynamics
technique
R. C. NEALEIGH, C. W. MILLER, AND F. D. McLEOD, Collaborative Radiological Health Laboratory and Department Colorado State University, Fort Collins, Colorado 80523
NEALEIGH, R. C., C. W. MILLER, AND F. D. MCLEOD, kvenous ultrasound catheter-tip technique for evaluation of arterial hemodynamics. J. Appl. Physiol. 41(6): 946-952. 1976. -A variety of devices has been used for measuring flow properties of deep-lying arteries, but many have limitations. This paper describes a relatively nontraumatic intravenous approach which uses a catheter in connection with a pulsed ultrasonic Doppler velocity meter (PUDVM) and an ultrasound echo track. The venous ultrasound catheter (VUC) has permitted measurements of local instantaneous blood velocity, flow, and wall motion in the abdominal aorta and iliac arteries of beagle dogs; evaluation studies have been conducted to compare the VUC recordings with an independent method for measuring blood flow and wall motion. Coupling of this catheter-tip device with the PUDVM and echo track provides chronic measurements of hemodynamic parameters in these deep vessels which were virtually impossible to obtain previously. This technique may prove useful in monitoring vessel pathology longitudinally as well as in basic experimental situations requiring flow and arterial wall mechanical properties. blood flow; pulsed
Doppler;
beagle dogs
HEMODYNAMIC MEASUREMENTS of deep-lying vessels such as the abdominal aorta and iliac arteries would be a valuable asset in determining severity of arterial disease, normal aging changes, and/or normal arterial behavior. Previous efforts have involved using transcutaneous ultrasonic crystals (7), intra-arterial catheters (1, 3, 5, 12), and surgically implanted flow cuffs. Each of these methods has inherent disadvantages which limits its usefulness. Transcutaneous methods are of limited value on large or obese animals as a result of the increased path length of the ultrasound beam which results in signal loss and uncertainties in the sound beam angle. The presence of intra-arterial catheters causes flow-stream disturbance and occasionally vasospasm. In addition to being invasive, the mere presence of implanted flow cuffs (11) hinders vessel wall dynamics. An advanced technique developed by Hodson and Duck (8) made ultrasonic measurement of arterial blood velocities of deep-lying vessels possible. By placing a continuous-wave Doppler ultrasound generating catheter-tip device in the vein adjacent to the artery and transmitting the ultrasound to the artery, recordings of blood velocities were made. The intravenous approach eliminates the possibilities of flow-stream disturbance and arterial vessel spasms due to the presence of the catheter. An improvement of this method has been developed by using a pulsed ultrasonic Doppler system’
l Pulsed
ultrasonic
Doppler
characteristics:
Emission
pulse
- cen-
for evaluation
JR. of Physiology
and Biophysics,
(10) in conjunction with the catheter-tip probe. This venous ultrasound catheter (VUC) permits the measurement of velocity distributions from which volume flow can be calculated. An additional feature of the probe is the incorporation of an echotrack scheme (9) which allows measurement of arterial wall motion and instantaneous vessel diameter. METHODS
In the initial evaluation of this method for measuring flow, a single-crystal catheter was used to determine the fi~asibility of the approach. Radiographs were taken to determine relationships between the catheter and the artery. Problems in determining accurate sound beam angles from the radiographs soon became apparent and led to the development of a modified catheter which provided more precise determination of the angle subtended by the flow stream and the ultrasound beam. Arteries of adult beagle dogs were injected with liquid plastic; and from these vascular models a probe was designed with two pulsed Doppler crystals (LTZ-2 lead titanate-zirconate, 7.5 MHz; Transducer Products, Torrington, Conn.) placed so their sound beams converged at 7.2 mm from the catheter, which was the average aorta center-line distance from the crystal face. By recording the center-line velocity with each of the two crystals, it is possible to calculate the sound beam angles for the two crystals. Figure 1 shows a schematic view of the current version of the VUC and Fig. 2 shows an actual photograph of the probe. Figure 3 depicts the relationship between the vessel flow stream and the transmitted ultrasound from the opposing crystals (1 and 2) housed within the venous ultrasound catheter. This scheme allows corrections to be made for the Doppler angles & and & if the original crystal angles are known within the VUC. The center crystal (LTZ-2 lead titanate zirconate, 5 MHz; Transducer Products) is perpendicular to the vessel wall and is used with the echo-track instrument for measuring wall displacement and instantaneous diameter change. The crystals are housed in a 3-mm-diameter brass tip which is connected to a 3.2-mm-OD flexible plastic tubing containing the lead wires. Markings were made on the tube so that insertion distances could be measured. The location of the venous ultrasound catheter probe tip was verified on unimplanted dogs by arteriography; while in the dogs with implanted flow cuffs, it is a simple matter to find the implanted cuff with the catheter-tip probe and use that as a reference point. ter frequency, 7.5 MHz; amplitude, 10 V p-p; duration, 4 cycles (0.53 ps); repetition frequencies, selectable 20, 40 kHz. Detectortype, range compensated, matched filter; input noise, 0.2 PV rms; bandwidth, 4 MHz; sample period, 0.25 ps. Frequency conversiondirectional zero crossing; linear range, -+ 1.6 m/s; frequency sponse, 20 Hz.
type, re-
Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (163.015.154.053) on October 10, 2018. Copyright © 1976 American Physiological Society. All rights reserved.
VENOUS
CATHETER
FOR
EVALUATION
OF
ARTERIAL
947
HEMODYNAMICS
FIG. 1. Schematic representation of venous ultrasound cathetertip probe (VUC). Top: probe as it would appear if viewed from above. Bottom: probe as seen from side. Two crystals for pulsed Doppler are oriented at an angle to flow stream while center echo-track crystal is at right angles to artery.
ments. Data were recorded on an analog magnetic tape recording system for subsequent computer processing and data reduction. Evaluation of the echo-track crystal for determining instantaneous diameter change and vessel wall motion was made by comparing these results with data from a 5-mm transducer which was placed directly on the artery during the terminal experimental session. Absolute measurements of the flowstream diameter with the echo track were not compared with the diameters calculated from velocity profile data using the pulsed Doppler because the echo-track instrument is designed for detecting motion rather than measuring absolute distances. Exact cuff placement and relative catheter-tip placements were determined on the implanted animals during the final recording session. The abdominal aorta and iliac arteries of two of the animals were injected with a plastic polymer (Batson’s no. 17 anatomical corrosion compound, Polysciences Inc., Warrington, Pa.), while arteriography was performed on the other three animals. RESULTS
FIG.
2. Actual photograph
of venous catheter-tip
probe.
The pulsed Doppler portions of both the single-crystal and the dual-crystal probes were evaluated for characteristics such as sound beam angles, crystal sensitivity, echo patterns, and accuracy of velocity, flow, and diameter measurements in a tube model with steady flow. Initial animal studies were performed on three normal adult beagles with methoxyflurane (Metofane) anesthesia. Arteriographs were used to calculate the crystal angle and determine probe placement. The results were used in determining requirements necessary for a dualcrystal catheter-tip probe and the areas of the abdominal aorta and iliac arteries accessible by this method. Evaluation of the dual-crystal pulsed Doppler catheter-tip probe was performed on five dogs, each implanted with two standard ultrasound flow cuffs (11). One cuff was located just distal to the left renal artery on the abdominal aorta and the other about 2 cm distal to the trifurcation on the left iliac artery. Comparisons of blood velocities, flow, and vessel diameters were made between the catheter-tip probe and the implanted flow cuffs. Recording sessions consisted of inserting the catheter-tip probe into the right femoral vein and positioning the catheter at varying locations along the inferior vena cava and iliac vein. Time-varying velocities were recorded with the pulsed Doppler in 0.5-ps (0.34-mm) increments across the artery on the slant diameter. Recordings were made with each of the two crystals housed within the implanted cuffs and from both the front and back crystal of the venous ultrasound catheter. The data from the two crystals within the cuff were averaged and compared with the averaged data from the front and back crystals of the VUC. All recordings for velocity and wall motion data were recorded sequentially rather than simultaneously due to interference between ultrasound instru-
Figure 4 shows a comparison between velocity profiles recorded on a model system with one crystal of the VUC and a theoretical profile constructed from the known tube diameter and flow. The measured profile compares favorably with the theoretical profile except near the walls of the tube where profile broadening occurs. The volume flow was calculated from the velocity distributions and the diameter of the flow stream. This diameter was calculated by extrapolating the tangent of the profile to zero velocity giving good agreement between the actual and measured parameters (4). From the profile in Fig. 4, an average flow of 12.0 cm% was computed compared with the actual flow of 13.2 cm3/s. Table 1 contains the vessel diameter, velocity, and flow parameters measured on each dog and averaged for the four recording sessions. This table gives a direct comparison of parameters measured with the implanted flow cuff on the abdominal aorta just distal to the left renal artery (AA@R) and the VUC about 2 cm distal to the implanted cuff. An example of the time-varying velocities, velocity profiles, and flow curves as displayed on microfilm generated from a CDC ULTRASOUND TRANSMISSION
BLOOD FLOW
c /\
VUC
Af, where then
where
PROBE
= Afp
V=AfC/2f,COSe Af,
/Af,
= COS of
V = Velocity Af = Frequency C = Speed e = Doppler
FIG. 3. Diagram of relationship sound beam and bloodstream.
0, /COS Blood
e2
Shaft
of Sound Angle
I” Blood
between VUC transmitted
ultra-
Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (163.015.154.053) on October 10, 2018. Copyright © 1976 American Physiological Society. All rights reserved.
948
NEALEIGH,
6400 computer system is shown in Figs. 5, 6, and 7. These figures are representative of the data and permit comparisons of the implanted cuff and the VUC. Recordings such as these are generally possible over the arterial segment length extending from the femoral artery to the abdominal aorta at a level between the two renal arteries. In addition to these computer-generated microfilm prints, numerical data such as flow-stream diameters, average blood flow, and other flowrelated data are provided on the computer output. Figures 8, 9, and 10 are graphs comparing the maximum forward velocity, vessel diameter, and the flow recorded from the implanted flow cuffs with the data from the VUC. AverVELOCITY (THEORETICAL
PROFILES vo MEASURED
1
60
MILLER,
AND
McLEOD
aged data from each recording session are depicted. Although there is not an excessive amount of error in the averaged data, each parameter as measured by the VUC is somewhat lower than for the implanted flow cuff. Figure 11 is a plot of average flow at the various locations along the artery as measured with the VUC and the cuff. The center crystal of the VUC was used to measure vessel wall motion. At termination prior to killing the animal, the wall motion was measured with the VUC and compared to the wall motion obtained using a patch transducer placed directly on the aorta. The results of these experiments are shown in Fig. 12. The wall motion was recorded on a strip-chart recorder which enabled wave form and amplitude to be compared directly. Figure 13 is an example of these data from one of the animals. Recordings on all subjects were similar to these and showed that the two devices produced almost identical wall motion wave forms and amplitudes. Problems might have TABLE 1. Averaged parameters of four recording sessions -
SO
40
Location
Implanted
30
Dog No.
Vessel Diam
Max Fwd Vel. cm/s
Max Rev Vel, cm/s
Peak Fwd Flow, ml/s
Peak Rev Flow, ml/s
Flow Rate, ml/s
cuff AA@R
1 2 3 4 5 Mean +SE
7.74 7.40 6.93 8.13 7.22 7.48 20.21
75.32 104.55 93.72 100.27 122.00 99.17 +7.58
-31.38 -32.75 -36.76 -39.77 -37.12 -35.56 + 1.53
19.26 25.75 17.59 27.35 28.94 23.78 ~2.26
-5.02 -4.48 -4.63 -7.77 -3.29 -5.04 20.74
4.76 5.53 3.67 5.32 8.13 5.48 20.74
1 2 3 4 5 Mean *SE
6.84 7.00 6.84 7.21 7.94 7.17 20.21
69.53 95.39 86.44 93.98 107.47 90.56 26.25
--21.16 - 19.86 -25.09 -23.37 - 15.69 -21.03 -+ 1.61
14.59 22.57 18.57 22.94 31.01 21.94 k2.73
-3.07 -3.48 -3.81 -4.26 +0.18 -2.89 +0.79
3.55 4.52 3.73 4.76 9.99 5.31 21.19
20
IO
vuc
- 2 cm distal cuff at renals
to
0
0
QI
a2
0.3
TU8E
Q4
DIAMETER
0.5
0.6
O.?
(cm)
FIG. 4. Comparison between theoretical velocity distribution velocity distribution as measured with VUC. These recordings made in a tube under steady flow.
100.
SET NO. ON 18698
90.
and were
2
Averaged hemodynamic the five dogs. Comparisons catheter can be made.
109.
r
A%
Measured
parameters obtained between the i mplanted
from four recording sessions flowcuffs and the intravenous
T O CENTER
N
SET
NO.
for each of ultrasonic
1
4-25-75 PDC IMPLANT AA-R ORG BLUE
80. 70. 60. : L u
50.
t
30.
;
20.
j w
10.
40.
0.
TIME
IN MILLISECONDS
FIG. 5. Comparisons of time-varying velocity wave forms from an implanted flow cuff (Left) and as recorded with VUC (right). Timevarying velocities were recorded in 05~s (0.34-mm) increments
TIME
across flow stream on angle plays depict numerous wave
IN MILLISECONDS
of transmitted forms.
sound
beam;
hence
dis-
Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (163.015.154.053) on October 10, 2018. Copyright © 1976 American Physiological Society. All rights reserved.
VENOUS
CATHETER VELOCITY I
180.1
FOR PROFILES I I
EVALUATION ACROSS 1
OF
VESSEL 1 1
ARTERIAL
I
O--PEAK I 1
180. -
150.
-
120.
-
llO.lOO.-
f
90.-
=
80.-
2
70.-
g
60.-
ACROSS I
VESSEL I I
O--PEAK I I
I
SET NO. ON 18698
1
160.150.
130.
PROFILES I 1
170.-
140.-
2
VELOCITY I
190.r
1
170.160.
;
949
HEMODYNAMICS
-
140.-
130.
-
s 120.v, 2 llO.”
lOO.-
z
90.-
E
80.-
g ci w
60.-
70.-
50. -
50.-
40.-
40.-
30.20.-
30.20.-
lO.-
10.-
O.-
O.-10
1
.
DISTANCE
z
0 0 c
iz
TIME
7. Instantaneous instantaneous velocity FIG.
E
0 z
c
blood profiles.
I
I
.
M
4
curves
I
.
u)
DISTANCE
0 iz
E
0 0 *
%
as recorded
I .
rT,
peak systole are displayed. Remainder during latter stages of cardiac cycle
1
.
P
0
I
.
I .
0
1
.
0
.
-
IN MM
of profiles which are recorded are suppressed for clarity.
0 iz
IN MILLISECONDS
flow
I
.
*
IN MM
6. Velocity profiles as recorded using implanted cuff (left) and VUC (right). Profiles are depicted in 20” intervals throughout cardiac cycle. In these representative figures, only profiles from 0 to FIG.
0
I
.
-
TIME
from
implanted
been encountered if the catheter moved significantly inside the vena cava. However, through visual observations and by tracing the motion of the VUC with an external transducer it was concluded that movement was negligible. DISCUSSION
A new venous ultrasonic catheter device (VUC) for obtaining hemodynamic data in relatively inaccessible arteries has been evaluated in a hydraulic model system and in beagle
flow
cuff
(left)
and
VUC
(right).
IN MILLISECONDS
Flow
curves
are calculated
from
dogs. These results have shown that arterial blood flow and arterial wall motion can be accurately measured with relative ease using the intravenous approach. Previous workers have recorded instantaneous arterial blood flow in the abdominal region with flow cuffs which encircle the artery or with an intra-arterial catheter. The intra-arterial catheter has the disadvantage that its presence interferes with normal hemodynamics. The lone report of an intravenous technique for measuring arterial blood flow (5) employed a continuous-wave Doppler which measures only blood velocity and does not
Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (163.015.154.053) on October 10, 2018. Copyright © 1976 American Physiological Society. All rights reserved.
950
NEALEIGH,
ILIAC
l
r=
.so
Y=
0.87X
AND
McLEOD
next time. Another problem is that any errors in calculating the sound beam angle will introduce considerable discrepancies into both velocity and dimension data. Failure to scan across the exact center line of the vessel will also result in underestimations of diameter and velocity. These problems can be reduced by using care in positioning the VUC. The lower blood flow measured with the venous ultrasound catheter compared to the cuff can be explained in part by the fact that all of the measurements with the probe were necessarily r 1.5 cm distal to the implanted cuff due to interference from the cuff and associated tissues. There are some small vessels which branch from the abdominal aorta immediately distal to the implanted cuff; thus lower blood flow and smaller vessel diameter would result. Similar values for arterial blood flows
AORTA
l
MILLER,
+
2.27
25
50
75 VELOCITY IMPLANTED
100 (cm/set) CUFFS
I25
150
FIG. 8. Comparison between maximum forward velocity as measured with VUC and implanted cuffs. Recordings were made in abdominal aorta at renal arteries (0) and in iliac artery (@. Regression line (); + 1 correlation line (- - -).
r=
.91
Y = 1.15 x
OO
I///.
,
4I
2
FLOW IMPLANTED a /
l
r= Y=
-0
2
4 DIAMETER
IMPLANTED
6
-
0.88
8
81 (cc/see
0.75
IO I
12 I
1 CUFFS
FIG. 10. Comparison between mean flow as measured with VUC and implanted cuffs. Recordings were made in abdominal aorta at renal arteries (0) and in iliac artery (w). Regression line (-); +1 correlation line (- - 4.
AORTA ILIAC .85 1.06X
61
+
IO
(mm) CUFFS
FIG. 9. Comparison between mean diameter as measured VUC and implanted cuffs. Recordings were made in abdominal at renal arteries (a) and in iliac artery ( n ) . Regression line (->; correlation line (- - 4.
with aorta + 1
provide information concerning blood flow or arterial wall motion. Hood flow. The accuracy of the technique for measuring blood flow in the abdominal aorta and iliac artery was comparable although greater errors in measurement of the vessel diameterin the smaller vessel could contribute to larger errors in flow than in the larger vessels. One of the problems encountered using this method was the difficulty in finding femoral venous branches large enough for probe insertion after the second and third recording sessions. Since the vein used in the previous recording session was simply tied off, it sclerosed and scar tissue covered the area causing bleeding and diffkulty in locating a suitable vein the
0' CUFF (AA-R)
FIG.
sured sured
vuc 2 cm DISTAL
vuc PROX.TO TRIF.
CUFF (LI)
vuc (RI)
11. Individual values for average blood flow (2 SE) as meaon all dogs and all recording sessions. Blood flows were meawith cuffs and VUC.
Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (163.015.154.053) on October 10, 2018. Copyright © 1976 American Physiological Society. All rights reserved.
VENOUS
CATHETER
FOR
EVALUATION
OF
ARTERIAL
951
HEMODYNAMICS
0.6
0.5 0 E 0.4 Y 0 :3 2" 3
0.3
$
0.2
IO z
vuc
I I
a 3
Stmm
0. I I
0.0
0.0
0.1
0.2
0.3 AMPLITUDE STANDARD
0.4
0.5
0.6
0.7
PATCH
(mm) CRYSTAL
FIG. 12. Comparison of abdominal aorta wall motion amplitude as measured with standard patch crystal directly on aorta and VUC from inside vena cava. Regression line (p); +1 correlation line (- - -).
have been reported by others (2) for conditions comparable to those reported here. W&Z motion. The excellent results obtained with the echotrack portion of the catheter are due in part to the fact that positioning is considerably easier than for the blood flow technique. When the sound beam is perpendicular to the vessel wall the greatest echo amplitude is received and the greatest motion of the vessel occurs. The values for wall motion recorded in the abdominal aorta agree favorably with those reported by Gow and Taylor (6). The pressure-strain elastic modulus E,, may be used for this comparison. Gow and Taylor reported mean values of 1.77 x lO”dyn/cm” compared with a mean value of 1.02 x lO”dyn/cm” reported here. The intravenous ultrasound catheter has been demonstrated to be a viable alternative for evaluating blood flow and wall motion dynamics in deep-lying vessels. Since the artery itself is not invaded, vasospasm does not occur, which would complicate the hemodynamic picture. Obviously flow disturbances likewise do not occur. The effect of venous pressure and flow changes on catheter movement are minimal and therefore minimal errors in arterial wall motion or blood velocity occur. In our laboratory we have been able to monitor aortic diameter and blood flow changes continuously over periods of up to several hours. The capability for continuously monitoring these events would be important in experiments designed to
.
.
TRANSDUCER
FIG. 13. Comparison between instantaneous diameter change as measured with intravenous ultrasound catheter and a transducer attached directly to aorta.
demonstrate the effect of drugs or other insults on the cardiovascular system. The catheter could also be adapted for evaluation of carotid artery dynamics via the superior vena caval approach. Such a capability opens up new avenues for the study of local flow properties associated with outflow tracts and with atherosclerotic lesions which may affect flow properties. Similarly, the mechanical properties of the arterial wall can be evaluated with this catheter during the course of aging or experimentally induced situations affecting the distensibility of the wall. For instance, such parameters as pulse wave velocity and instantaneous arterial diameter provide indices of arterial wall stiffness and can be provided with this technique. This method although currently used only in animal studies may be applicable to man where knowledge of vessel hemodynamics is desirable, e.g., in studies of aging, arterial stenosis, atherosclerotic plaque formation, and aneurysm development. This work was supported in part by Contract 223-76-6002 from the Bureau of Radiological Health, Food and Drug Administration, U. S. Department of Health, Education and Welfare. Address for reprint requests: R. C. Nealeigh, Collaborative Radiological Health Laboratory. Received
for publication
15 January
1976.
REFERENCES 1. ALFONSO,
S. A thermodilution flowmeter. J. Appl. Physiol. 21: 1883-1886, 1966. 2. ATTINGER, E. O., H. SUGAWARA, A. NAVARRO, A. RICCETTO, AND R. MARTIN. Pressure-flow relations in dog arteries. CircuZation Res. 19:230-246, 1966. 3. BENCHIMOL, A., H. F. STEGAL, P. R. MOROKI, J. L. GARTLAN, AND L. BRENER. Aortic flow velocity in man during cardiac arrhythmias measured with the Doppler catheter flowmeter system. Am. Heart J. 78: 649-659, 1969. 4. DAIGLE, R. E., C. W. MILLER, M. B. HISTAND, F. D. MCLEOD, AND D. E. HOKANSON. Nontraumatic aortic blood flow sensing by use
of an ultrasonic esophageal probe. J. AppZ. Physiol. 38: 11531160, 1975. 5. FIXLER, D. E., L. STAGE, A. M. RUDOLPH, G. D. BUCKBERG, AND J. P. ARCHIE. Evaluation of a Doppler catheter probe to measure cardiac output. J. Surg. Res. 15: 243-250, 1973. 6. Gow, B. S., AND M. G. TAYLOR. Measurement of viscoelastic properties of arteries in the living dog. CircuZation Res. 23: lll121, 1968. 7. HISTAND, M. B., C. W. MILLER, AND F. D. MCLEOD, JR. The transcutaneous measurement of blood velocity, profiles and flow. CardiovascuZar Res. 7: 703-712, 1973.
Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (163.015.154.053) on October 10, 2018. Copyright © 1976 American Physiological Society. All rights reserved.
952
NEALEIGH,
C. J., AND F. A. DUCK. Bloodflow monitoring in deeplyarteries. Invest. RadioZ. 8: 160-166, 1973. 9. HOKANSON, D. E., D. J. MOZERSKY, D. S. SUMMER, AND D. E. STRANDNESS. A phase-locked echo tracking system for recording arterial diameter changes in vivo. J. Appl. Physiol. 32: 728-733, 1972. 10. MCLEOD, F. D., JR. Comparison of pulse and continuous wave 8. HODSON,
situated
MILLER,
AND
McLEOD
Doppler flowmeters. Conf. Engr. Med. Biol., 23rd Washington, D. c., 1970. 11. MILLER, C. W., M. B. HISTAND, AND F. D. MCLEOD, JR. The chronic measurement of local flow properties in the abdominal aorta of dogs. Med. Res. Eng. 11: 17-24, 1972. 12. MILLS, C. J. A catheter-tip electromagnetic velocity probe. Phys. Med. BioZ. 11: 323, 1966.
Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (163.015.154.053) on October 10, 2018. Copyright © 1976 American Physiological Society. All rights reserved.
OF
Vol. 41, No.
A PPLIED
PHYSIOLOC
6, December
Y
1976.
Printed
in U.S.A.
Venous ultrasound catheter-tip of arterial hemodynamics
technique
R. C. NEALEIGH, C. W. MILLER, AND F. D. McLEOD, Collaborative Radiological Health Laboratory and Department Colorado State University, Fort Collins, Colorado 80523
NEALEIGH, R. C., C. W. MILLER, AND F. D. MCLEOD, kvenous ultrasound catheter-tip technique for evaluation of arterial hemodynamics. J. Appl. Physiol. 41(6): 946-952. 1976. -A variety of devices has been used for measuring flow properties of deep-lying arteries, but many have limitations. This paper describes a relatively nontraumatic intravenous approach which uses a catheter in connection with a pulsed ultrasonic Doppler velocity meter (PUDVM) and an ultrasound echo track. The venous ultrasound catheter (VUC) has permitted measurements of local instantaneous blood velocity, flow, and wall motion in the abdominal aorta and iliac arteries of beagle dogs; evaluation studies have been conducted to compare the VUC recordings with an independent method for measuring blood flow and wall motion. Coupling of this catheter-tip device with the PUDVM and echo track provides chronic measurements of hemodynamic parameters in these deep vessels which were virtually impossible to obtain previously. This technique may prove useful in monitoring vessel pathology longitudinally as well as in basic experimental situations requiring flow and arterial wall mechanical properties. blood flow; pulsed
Doppler;
beagle dogs
HEMODYNAMIC MEASUREMENTS of deep-lying vessels such as the abdominal aorta and iliac arteries would be a valuable asset in determining severity of arterial disease, normal aging changes, and/or normal arterial behavior. Previous efforts have involved using transcutaneous ultrasonic crystals (7), intra-arterial catheters (1, 3, 5, 12), and surgically implanted flow cuffs. Each of these methods has inherent disadvantages which limits its usefulness. Transcutaneous methods are of limited value on large or obese animals as a result of the increased path length of the ultrasound beam which results in signal loss and uncertainties in the sound beam angle. The presence of intra-arterial catheters causes flow-stream disturbance and occasionally vasospasm. In addition to being invasive, the mere presence of implanted flow cuffs (11) hinders vessel wall dynamics. An advanced technique developed by Hodson and Duck (8) made ultrasonic measurement of arterial blood velocities of deep-lying vessels possible. By placing a continuous-wave Doppler ultrasound generating catheter-tip device in the vein adjacent to the artery and transmitting the ultrasound to the artery, recordings of blood velocities were made. The intravenous approach eliminates the possibilities of flow-stream disturbance and arterial vessel spasms due to the presence of the catheter. An improvement of this method has been developed by using a pulsed ultrasonic Doppler system’
l Pulsed
ultrasonic
Doppler
characteristics:
Emission
pulse
- cen-
for evaluation
JR. of Physiology
and Biophysics,
(10) in conjunction with the catheter-tip probe. This venous ultrasound catheter (VUC) permits the measurement of velocity distributions from which volume flow can be calculated. An additional feature of the probe is the incorporation of an echotrack scheme (9) which allows measurement of arterial wall motion and instantaneous vessel diameter. METHODS
In the initial evaluation of this method for measuring flow, a single-crystal catheter was used to determine the fi~asibility of the approach. Radiographs were taken to determine relationships between the catheter and the artery. Problems in determining accurate sound beam angles from the radiographs soon became apparent and led to the development of a modified catheter which provided more precise determination of the angle subtended by the flow stream and the ultrasound beam. Arteries of adult beagle dogs were injected with liquid plastic; and from these vascular models a probe was designed with two pulsed Doppler crystals (LTZ-2 lead titanate-zirconate, 7.5 MHz; Transducer Products, Torrington, Conn.) placed so their sound beams converged at 7.2 mm from the catheter, which was the average aorta center-line distance from the crystal face. By recording the center-line velocity with each of the two crystals, it is possible to calculate the sound beam angles for the two crystals. Figure 1 shows a schematic view of the current version of the VUC and Fig. 2 shows an actual photograph of the probe. Figure 3 depicts the relationship between the vessel flow stream and the transmitted ultrasound from the opposing crystals (1 and 2) housed within the venous ultrasound catheter. This scheme allows corrections to be made for the Doppler angles & and & if the original crystal angles are known within the VUC. The center crystal (LTZ-2 lead titanate zirconate, 5 MHz; Transducer Products) is perpendicular to the vessel wall and is used with the echo-track instrument for measuring wall displacement and instantaneous diameter change. The crystals are housed in a 3-mm-diameter brass tip which is connected to a 3.2-mm-OD flexible plastic tubing containing the lead wires. Markings were made on the tube so that insertion distances could be measured. The location of the venous ultrasound catheter probe tip was verified on unimplanted dogs by arteriography; while in the dogs with implanted flow cuffs, it is a simple matter to find the implanted cuff with the catheter-tip probe and use that as a reference point. ter frequency, 7.5 MHz; amplitude, 10 V p-p; duration, 4 cycles (0.53 ps); repetition frequencies, selectable 20, 40 kHz. Detectortype, range compensated, matched filter; input noise, 0.2 PV rms; bandwidth, 4 MHz; sample period, 0.25 ps. Frequency conversiondirectional zero crossing; linear range, -+ 1.6 m/s; frequency sponse, 20 Hz.
type, re-
Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (163.015.154.053) on October 10, 2018. Copyright © 1976 American Physiological Society. All rights reserved.
VENOUS
CATHETER
FOR
EVALUATION
OF
ARTERIAL
947
HEMODYNAMICS
FIG. 1. Schematic representation of venous ultrasound cathetertip probe (VUC). Top: probe as it would appear if viewed from above. Bottom: probe as seen from side. Two crystals for pulsed Doppler are oriented at an angle to flow stream while center echo-track crystal is at right angles to artery.
ments. Data were recorded on an analog magnetic tape recording system for subsequent computer processing and data reduction. Evaluation of the echo-track crystal for determining instantaneous diameter change and vessel wall motion was made by comparing these results with data from a 5-mm transducer which was placed directly on the artery during the terminal experimental session. Absolute measurements of the flowstream diameter with the echo track were not compared with the diameters calculated from velocity profile data using the pulsed Doppler because the echo-track instrument is designed for detecting motion rather than measuring absolute distances. Exact cuff placement and relative catheter-tip placements were determined on the implanted animals during the final recording session. The abdominal aorta and iliac arteries of two of the animals were injected with a plastic polymer (Batson’s no. 17 anatomical corrosion compound, Polysciences Inc., Warrington, Pa.), while arteriography was performed on the other three animals. RESULTS
FIG.
2. Actual photograph
of venous catheter-tip
probe.
The pulsed Doppler portions of both the single-crystal and the dual-crystal probes were evaluated for characteristics such as sound beam angles, crystal sensitivity, echo patterns, and accuracy of velocity, flow, and diameter measurements in a tube model with steady flow. Initial animal studies were performed on three normal adult beagles with methoxyflurane (Metofane) anesthesia. Arteriographs were used to calculate the crystal angle and determine probe placement. The results were used in determining requirements necessary for a dualcrystal catheter-tip probe and the areas of the abdominal aorta and iliac arteries accessible by this method. Evaluation of the dual-crystal pulsed Doppler catheter-tip probe was performed on five dogs, each implanted with two standard ultrasound flow cuffs (11). One cuff was located just distal to the left renal artery on the abdominal aorta and the other about 2 cm distal to the trifurcation on the left iliac artery. Comparisons of blood velocities, flow, and vessel diameters were made between the catheter-tip probe and the implanted flow cuffs. Recording sessions consisted of inserting the catheter-tip probe into the right femoral vein and positioning the catheter at varying locations along the inferior vena cava and iliac vein. Time-varying velocities were recorded with the pulsed Doppler in 0.5-ps (0.34-mm) increments across the artery on the slant diameter. Recordings were made with each of the two crystals housed within the implanted cuffs and from both the front and back crystal of the venous ultrasound catheter. The data from the two crystals within the cuff were averaged and compared with the averaged data from the front and back crystals of the VUC. All recordings for velocity and wall motion data were recorded sequentially rather than simultaneously due to interference between ultrasound instru-
Figure 4 shows a comparison between velocity profiles recorded on a model system with one crystal of the VUC and a theoretical profile constructed from the known tube diameter and flow. The measured profile compares favorably with the theoretical profile except near the walls of the tube where profile broadening occurs. The volume flow was calculated from the velocity distributions and the diameter of the flow stream. This diameter was calculated by extrapolating the tangent of the profile to zero velocity giving good agreement between the actual and measured parameters (4). From the profile in Fig. 4, an average flow of 12.0 cm% was computed compared with the actual flow of 13.2 cm3/s. Table 1 contains the vessel diameter, velocity, and flow parameters measured on each dog and averaged for the four recording sessions. This table gives a direct comparison of parameters measured with the implanted flow cuff on the abdominal aorta just distal to the left renal artery (AA@R) and the VUC about 2 cm distal to the implanted cuff. An example of the time-varying velocities, velocity profiles, and flow curves as displayed on microfilm generated from a CDC ULTRASOUND TRANSMISSION
BLOOD FLOW
c /\
VUC
Af, where then
where
PROBE
= Afp
V=AfC/2f,COSe Af,
/Af,
= COS of
V = Velocity Af = Frequency C = Speed e = Doppler
FIG. 3. Diagram of relationship sound beam and bloodstream.
0, /COS Blood
e2
Shaft
of Sound Angle
I” Blood
between VUC transmitted
ultra-
Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (163.015.154.053) on October 10, 2018. Copyright © 1976 American Physiological Society. All rights reserved.
948
NEALEIGH,
6400 computer system is shown in Figs. 5, 6, and 7. These figures are representative of the data and permit comparisons of the implanted cuff and the VUC. Recordings such as these are generally possible over the arterial segment length extending from the femoral artery to the abdominal aorta at a level between the two renal arteries. In addition to these computer-generated microfilm prints, numerical data such as flow-stream diameters, average blood flow, and other flowrelated data are provided on the computer output. Figures 8, 9, and 10 are graphs comparing the maximum forward velocity, vessel diameter, and the flow recorded from the implanted flow cuffs with the data from the VUC. AverVELOCITY (THEORETICAL
PROFILES vo MEASURED
1
60
MILLER,
AND
McLEOD
aged data from each recording session are depicted. Although there is not an excessive amount of error in the averaged data, each parameter as measured by the VUC is somewhat lower than for the implanted flow cuff. Figure 11 is a plot of average flow at the various locations along the artery as measured with the VUC and the cuff. The center crystal of the VUC was used to measure vessel wall motion. At termination prior to killing the animal, the wall motion was measured with the VUC and compared to the wall motion obtained using a patch transducer placed directly on the aorta. The results of these experiments are shown in Fig. 12. The wall motion was recorded on a strip-chart recorder which enabled wave form and amplitude to be compared directly. Figure 13 is an example of these data from one of the animals. Recordings on all subjects were similar to these and showed that the two devices produced almost identical wall motion wave forms and amplitudes. Problems might have TABLE 1. Averaged parameters of four recording sessions -
SO
40
Location
Implanted
30
Dog No.
Vessel Diam
Max Fwd Vel. cm/s
Max Rev Vel, cm/s
Peak Fwd Flow, ml/s
Peak Rev Flow, ml/s
Flow Rate, ml/s
cuff AA@R
1 2 3 4 5 Mean +SE
7.74 7.40 6.93 8.13 7.22 7.48 20.21
75.32 104.55 93.72 100.27 122.00 99.17 +7.58
-31.38 -32.75 -36.76 -39.77 -37.12 -35.56 + 1.53
19.26 25.75 17.59 27.35 28.94 23.78 ~2.26
-5.02 -4.48 -4.63 -7.77 -3.29 -5.04 20.74
4.76 5.53 3.67 5.32 8.13 5.48 20.74
1 2 3 4 5 Mean *SE
6.84 7.00 6.84 7.21 7.94 7.17 20.21
69.53 95.39 86.44 93.98 107.47 90.56 26.25
--21.16 - 19.86 -25.09 -23.37 - 15.69 -21.03 -+ 1.61
14.59 22.57 18.57 22.94 31.01 21.94 k2.73
-3.07 -3.48 -3.81 -4.26 +0.18 -2.89 +0.79
3.55 4.52 3.73 4.76 9.99 5.31 21.19
20
IO
vuc
- 2 cm distal cuff at renals
to
0
0
QI
a2
0.3
TU8E
Q4
DIAMETER
0.5
0.6
O.?
(cm)
FIG. 4. Comparison between theoretical velocity distribution velocity distribution as measured with VUC. These recordings made in a tube under steady flow.
100.
SET NO. ON 18698
90.
and were
2
Averaged hemodynamic the five dogs. Comparisons catheter can be made.
109.
r
A%
Measured
parameters obtained between the i mplanted
from four recording sessions flowcuffs and the intravenous
T O CENTER
N
SET
NO.
for each of ultrasonic
1
4-25-75 PDC IMPLANT AA-R ORG BLUE
80. 70. 60. : L u
50.
t
30.
;
20.
j w
10.
40.
0.
TIME
IN MILLISECONDS
FIG. 5. Comparisons of time-varying velocity wave forms from an implanted flow cuff (Left) and as recorded with VUC (right). Timevarying velocities were recorded in 05~s (0.34-mm) increments
TIME
across flow stream on angle plays depict numerous wave
IN MILLISECONDS
of transmitted forms.
sound
beam;
hence
dis-
Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (163.015.154.053) on October 10, 2018. Copyright © 1976 American Physiological Society. All rights reserved.
VENOUS
CATHETER VELOCITY I
180.1
FOR PROFILES I I
EVALUATION ACROSS 1
OF
VESSEL 1 1
ARTERIAL
I
O--PEAK I 1
180. -
150.
-
120.
-
llO.lOO.-
f
90.-
=
80.-
2
70.-
g
60.-
ACROSS I
VESSEL I I
O--PEAK I I
I
SET NO. ON 18698
1
160.150.
130.
PROFILES I 1
170.-
140.-
2
VELOCITY I
190.r
1
170.160.
;
949
HEMODYNAMICS
-
140.-
130.
-
s 120.v, 2 llO.”
lOO.-
z
90.-
E
80.-
g ci w
60.-
70.-
50. -
50.-
40.-
40.-
30.20.-
30.20.-
lO.-
10.-
O.-
O.-10
1
.
DISTANCE
z
0 0 c
iz
TIME
7. Instantaneous instantaneous velocity FIG.
E
0 z
c
blood profiles.
I
I
.
M
4
curves
I
.
u)
DISTANCE
0 iz
E
0 0 *
%
as recorded
I .
rT,
peak systole are displayed. Remainder during latter stages of cardiac cycle
1
.
P
0
I
.
I .
0
1
.
0
.
-
IN MM
of profiles which are recorded are suppressed for clarity.
0 iz
IN MILLISECONDS
flow
I
.
*
IN MM
6. Velocity profiles as recorded using implanted cuff (left) and VUC (right). Profiles are depicted in 20” intervals throughout cardiac cycle. In these representative figures, only profiles from 0 to FIG.
0
I
.
-
TIME
from
implanted
been encountered if the catheter moved significantly inside the vena cava. However, through visual observations and by tracing the motion of the VUC with an external transducer it was concluded that movement was negligible. DISCUSSION
A new venous ultrasonic catheter device (VUC) for obtaining hemodynamic data in relatively inaccessible arteries has been evaluated in a hydraulic model system and in beagle
flow
cuff
(left)
and
VUC
(right).
IN MILLISECONDS
Flow
curves
are calculated
from
dogs. These results have shown that arterial blood flow and arterial wall motion can be accurately measured with relative ease using the intravenous approach. Previous workers have recorded instantaneous arterial blood flow in the abdominal region with flow cuffs which encircle the artery or with an intra-arterial catheter. The intra-arterial catheter has the disadvantage that its presence interferes with normal hemodynamics. The lone report of an intravenous technique for measuring arterial blood flow (5) employed a continuous-wave Doppler which measures only blood velocity and does not
Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (163.015.154.053) on October 10, 2018. Copyright © 1976 American Physiological Society. All rights reserved.
950
NEALEIGH,
ILIAC
l
r=
.so
Y=
0.87X
AND
McLEOD
next time. Another problem is that any errors in calculating the sound beam angle will introduce considerable discrepancies into both velocity and dimension data. Failure to scan across the exact center line of the vessel will also result in underestimations of diameter and velocity. These problems can be reduced by using care in positioning the VUC. The lower blood flow measured with the venous ultrasound catheter compared to the cuff can be explained in part by the fact that all of the measurements with the probe were necessarily r 1.5 cm distal to the implanted cuff due to interference from the cuff and associated tissues. There are some small vessels which branch from the abdominal aorta immediately distal to the implanted cuff; thus lower blood flow and smaller vessel diameter would result. Similar values for arterial blood flows
AORTA
l
MILLER,
+
2.27
25
50
75 VELOCITY IMPLANTED
100 (cm/set) CUFFS
I25
150
FIG. 8. Comparison between maximum forward velocity as measured with VUC and implanted cuffs. Recordings were made in abdominal aorta at renal arteries (0) and in iliac artery (@. Regression line (); + 1 correlation line (- - -).
r=
.91
Y = 1.15 x
OO
I///.
,
4I
2
FLOW IMPLANTED a /
l
r= Y=
-0
2
4 DIAMETER
IMPLANTED
6
-
0.88
8
81 (cc/see
0.75
IO I
12 I
1 CUFFS
FIG. 10. Comparison between mean flow as measured with VUC and implanted cuffs. Recordings were made in abdominal aorta at renal arteries (0) and in iliac artery (w). Regression line (-); +1 correlation line (- - 4.
AORTA ILIAC .85 1.06X
61
+
IO
(mm) CUFFS
FIG. 9. Comparison between mean diameter as measured VUC and implanted cuffs. Recordings were made in abdominal at renal arteries (a) and in iliac artery ( n ) . Regression line (->; correlation line (- - 4.
with aorta + 1
provide information concerning blood flow or arterial wall motion. Hood flow. The accuracy of the technique for measuring blood flow in the abdominal aorta and iliac artery was comparable although greater errors in measurement of the vessel diameterin the smaller vessel could contribute to larger errors in flow than in the larger vessels. One of the problems encountered using this method was the difficulty in finding femoral venous branches large enough for probe insertion after the second and third recording sessions. Since the vein used in the previous recording session was simply tied off, it sclerosed and scar tissue covered the area causing bleeding and diffkulty in locating a suitable vein the
0' CUFF (AA-R)
FIG.
sured sured
vuc 2 cm DISTAL
vuc PROX.TO TRIF.
CUFF (LI)
vuc (RI)
11. Individual values for average blood flow (2 SE) as meaon all dogs and all recording sessions. Blood flows were meawith cuffs and VUC.
Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (163.015.154.053) on October 10, 2018. Copyright © 1976 American Physiological Society. All rights reserved.
VENOUS
CATHETER
FOR
EVALUATION
OF
ARTERIAL
951
HEMODYNAMICS
0.6
0.5 0 E 0.4 Y 0 :3 2" 3
0.3
$
0.2
IO z
vuc
I I
a 3
Stmm
0. I I
0.0
0.0
0.1
0.2
0.3 AMPLITUDE STANDARD
0.4
0.5
0.6
0.7
PATCH
(mm) CRYSTAL
FIG. 12. Comparison of abdominal aorta wall motion amplitude as measured with standard patch crystal directly on aorta and VUC from inside vena cava. Regression line (p); +1 correlation line (- - -).
have been reported by others (2) for conditions comparable to those reported here. W&Z motion. The excellent results obtained with the echotrack portion of the catheter are due in part to the fact that positioning is considerably easier than for the blood flow technique. When the sound beam is perpendicular to the vessel wall the greatest echo amplitude is received and the greatest motion of the vessel occurs. The values for wall motion recorded in the abdominal aorta agree favorably with those reported by Gow and Taylor (6). The pressure-strain elastic modulus E,, may be used for this comparison. Gow and Taylor reported mean values of 1.77 x lO”dyn/cm” compared with a mean value of 1.02 x lO”dyn/cm” reported here. The intravenous ultrasound catheter has been demonstrated to be a viable alternative for evaluating blood flow and wall motion dynamics in deep-lying vessels. Since the artery itself is not invaded, vasospasm does not occur, which would complicate the hemodynamic picture. Obviously flow disturbances likewise do not occur. The effect of venous pressure and flow changes on catheter movement are minimal and therefore minimal errors in arterial wall motion or blood velocity occur. In our laboratory we have been able to monitor aortic diameter and blood flow changes continuously over periods of up to several hours. The capability for continuously monitoring these events would be important in experiments designed to
.
.
TRANSDUCER
FIG. 13. Comparison between instantaneous diameter change as measured with intravenous ultrasound catheter and a transducer attached directly to aorta.
demonstrate the effect of drugs or other insults on the cardiovascular system. The catheter could also be adapted for evaluation of carotid artery dynamics via the superior vena caval approach. Such a capability opens up new avenues for the study of local flow properties associated with outflow tracts and with atherosclerotic lesions which may affect flow properties. Similarly, the mechanical properties of the arterial wall can be evaluated with this catheter during the course of aging or experimentally induced situations affecting the distensibility of the wall. For instance, such parameters as pulse wave velocity and instantaneous arterial diameter provide indices of arterial wall stiffness and can be provided with this technique. This method although currently used only in animal studies may be applicable to man where knowledge of vessel hemodynamics is desirable, e.g., in studies of aging, arterial stenosis, atherosclerotic plaque formation, and aneurysm development. This work was supported in part by Contract 223-76-6002 from the Bureau of Radiological Health, Food and Drug Administration, U. S. Department of Health, Education and Welfare. Address for reprint requests: R. C. Nealeigh, Collaborative Radiological Health Laboratory. Received
for publication
15 January
1976.
REFERENCES 1. ALFONSO,
S. A thermodilution flowmeter. J. Appl. Physiol. 21: 1883-1886, 1966. 2. ATTINGER, E. O., H. SUGAWARA, A. NAVARRO, A. RICCETTO, AND R. MARTIN. Pressure-flow relations in dog arteries. CircuZation Res. 19:230-246, 1966. 3. BENCHIMOL, A., H. F. STEGAL, P. R. MOROKI, J. L. GARTLAN, AND L. BRENER. Aortic flow velocity in man during cardiac arrhythmias measured with the Doppler catheter flowmeter system. Am. Heart J. 78: 649-659, 1969. 4. DAIGLE, R. E., C. W. MILLER, M. B. HISTAND, F. D. MCLEOD, AND D. E. HOKANSON. Nontraumatic aortic blood flow sensing by use
of an ultrasonic esophageal probe. J. AppZ. Physiol. 38: 11531160, 1975. 5. FIXLER, D. E., L. STAGE, A. M. RUDOLPH, G. D. BUCKBERG, AND J. P. ARCHIE. Evaluation of a Doppler catheter probe to measure cardiac output. J. Surg. Res. 15: 243-250, 1973. 6. Gow, B. S., AND M. G. TAYLOR. Measurement of viscoelastic properties of arteries in the living dog. CircuZation Res. 23: lll121, 1968. 7. HISTAND, M. B., C. W. MILLER, AND F. D. MCLEOD, JR. The transcutaneous measurement of blood velocity, profiles and flow. CardiovascuZar Res. 7: 703-712, 1973.
Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (163.015.154.053) on October 10, 2018. Copyright © 1976 American Physiological Society. All rights reserved.
952
NEALEIGH,
C. J., AND F. A. DUCK. Bloodflow monitoring in deeplyarteries. Invest. RadioZ. 8: 160-166, 1973. 9. HOKANSON, D. E., D. J. MOZERSKY, D. S. SUMMER, AND D. E. STRANDNESS. A phase-locked echo tracking system for recording arterial diameter changes in vivo. J. Appl. Physiol. 32: 728-733, 1972. 10. MCLEOD, F. D., JR. Comparison of pulse and continuous wave 8. HODSON,
situated
MILLER,
AND
McLEOD
Doppler flowmeters. Conf. Engr. Med. Biol., 23rd Washington, D. c., 1970. 11. MILLER, C. W., M. B. HISTAND, AND F. D. MCLEOD, JR. The chronic measurement of local flow properties in the abdominal aorta of dogs. Med. Res. Eng. 11: 17-24, 1972. 12. MILLS, C. J. A catheter-tip electromagnetic velocity probe. Phys. Med. BioZ. 11: 323, 1966.
Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (163.015.154.053) on October 10, 2018. Copyright © 1976 American Physiological Society. All rights reserved.
