Pulmonary arterial compliance at rest and exercise in normal humans

DAVID M. SLIFE, RICKY D. LATHAM, PIETER SIPKEMA, AND NICO WESTERHOF Cardiology Service and Department of Clinical Investigation, Brooke Army Medical Center, Fort Sam Houston, Texas 78234; and Laboratory for Physiology, Free University, 1081 BT Amsterdam, The Netherlands

M., RICKY D. LATHAM, PIETER SIPKEMA, AND Pulmonary arterial compliance at rest and exercise in normaL humans. Am. J. Physiol. 258 (Heart Circ. Physiol. 27): H.1823-H1828, 1990.-We evaluated the feasibility of determining pulmonary arterial compliance (C,) by a parameter estimation procedure based on the three-element windkessel model. Eight normal patients studied with multisensor micromanometry technology had simultaneous rest and exercise pulmonary artery pressures (PAP) and flows recorded. These were submitted to the model and independent methods to determine C,, pulmonary characteristic impedance (Z,), and pulmonary vascular resistance (PVR). Significant changes in heart rate, PAP, and stroke volume (P < 0.05) occurred with exercise. In comparing rest and exercise 2, and PVR values determined by the model and independent methods, and in comparing each method for these values, there was no significant difference. Model-derived and independently derived estimates of C, were significantly different at rest (P < 0.04) and exercise (P < 0.001). There was no significant difference between rest and exercise values of C, by either method. The model estimates of PVR at rest (64 t 11 dyne so cms5) and exercise (41 t 7 dyn . so cm-) (P = 0.06) and the model 2, value at rest (22 k 3 dyne so cmW5) were appropriate. The model C, values at rest (0.22 t 0.05 ml.mmHg-lo kg-) correlated with previously reported normalized values in other species. This study reports the successful use of a parameter estimation procedure based on the three-element windkessel model to describe pulmonary artery compliance in normal humans.
SLIFE, DAVID NICO WESTERHOF.

with a computer-processing algorithm to obtain the three parameters of vascular function on a beat-to-beat basis and during transients from measured pressure and flow data (20). To date, the application has generally been limited to the systemic arterial tree in the animal studies (7, 20) with one report in humans (6). Detailed studies on the pulmonary input impedance spectrum in humans have been limited to the resting state (12) or in vitro electrical analog methods (15). A reliable method to determine pulmonary arterial compliance in humans has not been suggested. An accurate measurement of the arterial compliance may provide useful information in the evaluation of pulmonary hypertension or in determining medical therapy for pulmonary hypertension. Furthermore, the contribution of these pulmonary arterial parameters, pulmonary arterial compliance (C,), characteristic impedance (2,)) vascular resistance (PVR) at rest and during submaximal supine exercise, has not been investigated and might provide insight into the basic pulmonary physiology. The present study in normal humans was designed to evaluate the utility of a parameter estimation procedure based on the three-element windkessel model applied to the pulmonary arterial circulation to describe the C,, Z,, and PVR at rest and during supine submaximal exercise.
METHODS

pulmonary model

characteristic

impedance;

modeling;

windkessel

DESCRIPTIONS of the structure and function of the circulatory system have been attempted since the early part of the 18th century beginning with the efforts of Reverend Stephen Hales (5a). He emphasized the functional property of elasticity of the vascular tree and believed that the arterial system behaved as a large, distensible reservoir, later referred to as a windkessel in the German literature. Since that time, numerous mathematical models have been contrived to describe and evaluate the properties of pulsatile flow and pressure within the arterial system (2, 19, 20, 22, 24). One such mathematical approach has been to apply an electrical analogue of a three-element windkessel model to give estimates of characteristic impedance, vascular resistance, and compliance of the systemic arterial circulation (22). Recently, this model has been integrated DETAILED

Eight patients were studied during elective routine cardiac catheterization performed for various clinical indications, the most common of which was a chest pain syndrome. Medicines were discontinued before study. There was no evidence of organic heart disease by hemodynamics at rest or during supine submaximal exercise or as judged by left ventricular cineangiography or coronary arteriography. Therefore, all patients were considered physiologically normal and without demonstrable cardiovascular pathology. All patients were studied in a fasting basal state and were unsedated or very lightly sedated with diazepam (10 mg orally -1 h before the procedure). All patients were studied with informed consent. This protocol was reviewed and approved by our medical center Institutional Review Board. Right and left heart cardiac catheterizations were performed using the Sones technique from a right brachial approach. A balloon-tipped, flow-directed, thermodilution catheter was advanced through the right heart
H1823

H1824

PULMONARY

ARTERIAL

COMPLIANCE

IN

HUMANS

PULMONARY EMF VELOCDTY

ARTERY PROBE ARTERY

PULMONARY SENSOR RIGHT SENSOR R1GHT SENSOR ATRIAL

VENTRICULAR

FIG. 1. Typical eter catheter.

placement

of right

heart

multisensor

micromanom-

chambers in routine fashion, and the tip was positioned in the distal pulmonary artery. A second right-heart catheter, which contained three solid-state pressure sensors and an electromagnetic flow velocity probe, was used to record high-fidelity right heart hemodynamics. The three pressure sensors were mounted laterally with the distal sensor, and the electromagnetic flow velocity probe was mounted at the same site (model no. SSD-192, Millar Instruments, Houston, TX). The distance between the tip of the catheter and the distal pressure sensor with the electromagnetic flow probe was 9 cm. The distance between the distal and middle sensors was 5 cm; between middle and proximal sensors, there was a 12-cm distance. The catheter was manipulated under fluoroscopic guidance so that the distal sensor and electromagnetic flow velocity probe were positioned in the main pulmonary artery just above the pulmonary valve, the middle sensor in the right ventricular outflow tract, and the proximal sensor in the right atrium (Fig. 1). Details of the technical characteristics of these sensors, including frequency response, drift characteristics, and calibration techniques, have been previously described (8, 13). During the rest and exercise studies, oxygen consumption was determined by collecting a 5-min Douglas bag and measuring oxygen content using mass spectrometry. Simultaneously with the gas collection, arteriovenous oxygen content difference was derived from hemoglobin oxygen saturation determinations from the brachial and pulmonary artery. Right heart oximetry was used to document absence of a left-to-right shunt. Cardiac output was t/hen calculated using the direct Fick method.

Cardiac outputs were also measured using the thermal dilution technique averaging three or more output determinations. The bolus was 10 ml of dextrose water chilled to 6C. The cardiac outputs by the two methods agreed with 600 ml/min difference, and the cardiac outputs quoted are those from the Fick method. When proper positioning of the catheters was achieved and a steady state confirmed by stable heart rate and pulmonary artery oxygen saturation determinations, simultaneous pulmonary artery flow and pressure, right ventricular, and right atria1 pressures were recorded. After the hemodynamics were recorded at rest, patients had their legs elevated and feet positioned on a bicycle ergometer (Quinton Instruments, Seattle, WA). The average work load given for submaximal supine exercise was 2,000 ft lb* s-l ranging between 710 and 3,250 ft. lb s-l depending on the judgment of the attending physician. Subjects were exercised a full 5 min after a steady state was verified by a stable heart and pulmonary artery oxygen saturation determination. Hemodynamics were recorded in similar fashion to the rest study. Finally, ventriculography and coronary cineangiography were performed after rest, and exercise hemodynamics were completed for verification of normality. Data analysis. Data were originally recorded and saved on l-in. 14-channel FM analog tape (Honeywell model 101). The cases were identified and a representative steady-state portion of the rest and exercise data were digitized at a 500-Hz sample rate using a MASSCOMP 5500 mainframe computer (MASSCOMP, Westford, MA). Using a generic signal processing program (DaDisp version l.O5B, DSP, Cambridge, MA), four representative beats at the end-expiratory phase of simultaneous pulmonary artery pressure and flow at rest and during exercise were selected. The mean pulmonary capillary wedge pressure was subtracted from the individual beats for both rest and exercise states. The pulmonary arterial flow velocity was calibrated to the Fick-derived stroke volume. Four representative beats were then averaged. The averaged simultaneous pressure and flow waveforms were used in a custom-programmed parameter identification algorithm based on the three-element windkessel model that has been previously described (7). The curvefitting algorithms adjust the three parameters through several iterations until the best fit (minimal of the chisquare) of the calculated flow to the measured flow has been achieved (see Fig. 2). Independent calculations of
l

500

MW8.C

ZC

500

YLhoc

PAF

FIG. 2. Illustration of 3-element windkessel model (middle) and example of fit to pulmonary artery flow calculated by model at rest and exercise. Mid&e: C, compliance; R, resistance; Z,, characteristic impedance. Left and right: C, calculated; M, measured; PAF, pulmonary artery flow; PAP, pulmonary artery pressure.

PULMONARY

ARTERIAL

COMPLIANCE

IN

HUMANS

HE325

1. Summary of demographic and hemodynamic data at rest and during supine submaximal exercise
TABLE
Patient

RESULTS

Age
Sex wt Ft-lb/s HR R E co R E SV R E PAS R E PAd R E PAm R E LVes R E LVed R E Aos R E Aod R E Aom R E

34 M 82 3,200 67 108 6.5 18.8

34 F 85 710 59 82 4 6.9

47 M 79 1,640 78 108 8.6 15.2 110 141 19 26 7 11 13 20 110 136 6 15 110 136 74 85 88 95

38 M 83 2,170 66 114 7.2 15.5 109 136 19 26 8 11 13 19 116 143 12 12 116 142 72 88 95 127

53 F 95 1,200 72 90 7.2 10.5

67 F 75 1,200 90 122 8.1 10.3

46 M 99 2,250 65 91 4.2 9.7 65 106 25 34 12 15 16 24 119 160 7 15 119 160 74 94 96 122

29 M 107 3,250 58 140 8.5 19.4 146 139 22 31 9 9 18 12 118 143 13 16 118 143 75 98 96 118

97
174 18 30 8 12 12 24 112 135 10 16 112 135 75 82 88 105

69
84 20 31 7 11 11 18 122 145 10 15 122 145 78 85 101
111

99
117 22 35 8 12 12 23 148 174 6 14 148 174 76 95 103 125

90
84 16 22 7 13 11 13 100 112 9 12 100 112 70 78 83 89

the three parameters were also performed to allow comparison of values determined by traditional methodology with those estimated by the computer model. Calculation of characteristic impedance was performed by applying a Fourier analysis to averaged pulmonary artery pressure and flow waveforms. Amplitude of moduli of pressure and flow harmonics were divided. The impedance modulus between 3rd and 12th harmonic were then averaged to give characteristic impedance. Pulmonary vascular resistance was considered the 0th harmonic of the impedance modulus. The arterial compliance was estimated by dividing the right ventricular stroke volume by the pulmonary arterial pulse pressure (systolic minus diastolic pressure). For comparison with other species, compliance estimates were normalized by dividing by the body weight in kilograms. The paired t test was used to test for a significant difference between calculated and modeled values of the three parameters and for the differences between rest and exercise values of those parameters. Values of P < 0.05 were considered significant. All values are presented as the mean t SE.

The subject population consisted of five men and three women aged 40 t 3 yr. All subjects had normal hemodynamic parameters (Table 1) at rest and with supine submaximal exercise. The subjects had normal left ventriculography and coronary angiography. The average heart rate increased by 37% (P < 0.001) with exercise. The mean cardiac output at rest was 6.5 l/min, which increased 50% (P c 0.005) with exercise with stroke volume increasing 20% (P < 0.03). The mean pulmonary artery pressure increased 29% (P < 0.02) during exercise. Pulmonary root flow signals are broad at the base with slower ascending and descending segments. The diastolic pulmonary artery pressure decays at rest were slow, indicating longer time constants. Individual values of the three parameters derived from the windkessel model and the calculated method during rest and exercise are shown in Table 2. The averaged parameters by the modeled and calculated methods at rest and with supine submaximal exercise are shown in Table 3. The relative error for each parameter estimation was <IO%, and the regression coefficient >0.90. The differences in characteristic impedance and vascular resistance at rest or during exercise using the estimations returned by the model and the independently calculated values were not statistically significant (see Fig. 3). There were also no significant differences between rest and exercise characteristic impedance determinations by either method. Pulmonary vascular resistance, however, decreased by an average of 50% from rest to exercise that did not quite meet significance (P = 0.06). In contrast, there was a significant difference between arterial compliance values determined by the model and those calculated (see Fig. 4). The modeled pulmonary arterial compliance at rest gave a mean value of 19.9 -+ 5.4 ml/mmHg, which was 2.5 times greater (P < 0.05) than the arterial compliance (8.0 t 0.87 ml/ mmHg) determined by independent estimation. During exercise, this difference increased to 3.4 times (P c 0.02) with the modeled arterial compliance giving a mean value of 27.5 t 3.63 ml/mmHg and the independently calculated arterial compliance estimating a mean of 8.12 t 0.66 ml/mmHg. However, the difference from rest to exercise by either method was not statistically significant. The modeled arterial compliance did reveal a trend (P = 0.06) to increase from rest conditions by -30% during exercise. When both modeled and independently derived arterial compliances were normalized for the subjects weight in kilograms (Table 4), there was still a significant difference between modeled and calculated methods at rest (P < 0.02) and with exercise (P c 0.003). However, there were no significant differences between the rest and exercise values for each respective method.
DISCUSSION

The present study evaluated the pulmonary arterial compliance by a parameter estimation procedure based on the three-element windkessel model, which additionally determined the characteristic impedance and the vascular resistance, at rest and during supine submaxi-

PULMONARY

ARTERIAL

COMPLIANCE LEGEND

IN

HUMANS

lz!izzl PVR,MODEL
PVRwCALC Zc-MODEL
Zc-CALC FIG. 3. Comparison of characteristic impedance and pulmonary vascular resistance determined by model and calculated methods at rest and during exercise (mean & SE). There was no significant difference between modeled and calculated values or between rest and exercise values (talc, calculated; PVR, pulmonary vascular resistance; Z,, characteristic impedance).

REST

EXERCISE

CONDITION

OF

STUDY

2. Modeled and calculated parameters at rest and with supine submaximal exercise
TABLE zz
Patient m 15 24 24 17 38 18 22 17 R c 15 23 19 26 28 12 25 26 m 23 18 12 20 16 7 17 10 E c 43 25 16 25 19 17 29 9 -m 19 9 9 20 34 13 5 50 R c 10 6 6 8 7 10 5 12

C
~ m 34 33 32 42 13 9 14 45 E c 10 9 9 10 5 9 6 7 R _______ m 50 75 75 86 25 60 119 25

PVR E
C

~ m 21 41 32 16 75 35 40 45

c 39 40 38 30 86 45 50 35

58 63 86 78 64 54 148 36

Z,, characteristic impedance (dyn . s. cma5); C, compliance (ml. mmHg- . kg-l); PVR, pulmonary vascular resistance (dyn +s. cmW5); R, rest; E, exercise; c, calculated; m, modeled.

ma1exercise in normal humans. We found model-derived estimations of characteristic impedance and vascular resistance were similar to independent calculations of these variables at rest or during exercise. Furthermore, the variables of these three parameters were statistically unchanged with supine submaximal exercise despite a significant change in the pulmonary artery pressure, stroke volume, and heart rate. The values at rest for Z,, both model derived and independently calculated, are similar to reported values at rest in humans ranging from 20 to 23 dyn. s.cmB5 (9,
TABLE

10, 15). Model-derived and calculated PVR at rest and with exercise correlated with previously reported values (5). Previous mathematical models have been used to predict the pressure and flow of the pulmonary circulation (3, 14, 16, 23). Piene (15) used a four-element model and validated this approach successfully in cats. Unlike parameters of resistance, however, there is no easily validated method for the determination of absolute pulmonary arterial compliance in vivo for humans. Limited studies on pulmonary arterial compliance have been performed in animals (12, 17). Milnor et al. (11) evaluated pulmonary compliance at rest in patients with valvular heart disease and pulmonary hypertension. The total pulmonary compliance was estimated by dividing the total pulmonary blood volume (determined by a dye dilution technique) by the mean intravascular pressure calculated as the mean pulmonary artery pressure plus left atria1 pressures divided by two. The averaged total pulmonary compliance (arterial plus venous compliance) normalized to body weight was found to be 0.408 ml. mmHg- kg-l in patients with normal PVR. Shoukas (17) found in dogs the mean total pulmonary vascular compliance that included arterial and venous contributions approximating 0.30 ml mmHg- .kg-, similar to the results of Milnor in humans. Shoukas (17) further demonstrated that the portion of total pulmonary vascular compliance attributed to the pulmonary arterial tree was 0.18 ml. mmHg-l . kg-, which closely approximates 0.22 ml. mmHg- kg- found in the present study. Utilizing a normalization of compliance values to body
l l l

3. Averaged modeled and calculated parameters at rest and with supine submaximal exercise
PVR R E 16t1.89 23t3.64 R 64t11.18 73211.92 E 38t6.34 45t6.18 R 20t5.4 8-c-0.87* C E 28t4.89 8t0.67-f

Modeled Calculated

22t2.6 2222.0

Z,, characteristic impedance (dyn . so cmm5); PVR, pulmonary vascular resistance kg-l); R, rest; E, exercise. * Calculated compliance at rest was significantly different with exercise was significantly different (P < 0.003) from modeled compliance.

(dyne s m cmM5); C, pulmonary arterial compliance (ml. mmHg- (P = 0.04) from modeled compliance. t Calculated compliance

PULMONARY

ARTERIAL

COMPLIANCE LEGEND

IN

HUMANS

HI827

Eza COMPLIANCFMODEL
COMPLIANCLCALC

FIG. 4. Comparison of pulmonary arterial compliance determined by model and calculated methods at rest and with exercise (mean t SE). When comparing modeled and calculated arterial compliance, there was a significant difference at rest and with exercise (*P < 0.05). There was, however, no significant difference between rest and exercise by either method (talc, calculated).

EXERCISE

OF

STUDY

TABLE
Patient

4. Normalized compliances
Rest m
C

Exercise m 0.42 0.39 0.38 0.5 0.14

C

1
2 3 4 5 6 7 8

0.23 0.11 0.11 0.24 0.36 0.17 0.05 0.46 0.22t0.05

0.12 0.07 0.07 0.01 0.07 0.13 0.05 0.11 0.08t0.01*

0.12
0.14 0.42 0.30t0.05
l

0.12 0.11 0.11 0.12 0.05 0.12 0.06 0.06 0.09+0.01j-

Mean

Normalized compliances (ml. mmHg- kg-) during exercise by modeled and calculated methods. c, calculated; m, modeled; * Mean calculated compliance was significantly different (P < 0.02) from mean modeled compliance. t Mean calculated compliance was significantly different (P < 0.003) from mean modeled compliance with exercise.

weight or surface area allows for the comparison of results between species. When the modeled and calculated C, were normalized to body weight and compared with the pulmonary arterial compliances obtained by Shoukas (17), then the modeled C, appeared a more reasonable estimate of the arterial compliance than that approximated by the independently derived method. Because the independently derived method provides only a crude estimate of compliance that dynamically changes throughout the cardiac cycle, a discrepancy between methods was expected. The trend seen in the pulmonary artery compliance may have been influenced by an increase in the cross-sectional area of the distal pulmonary arteriolar bed. Thus the trend toward an increase in arterial compliance may have been determined more by a peripheral phenomenon rather than one involving the main pulmonary artery and major branches. Our findings and hypothesis are further supported by Yu et al. (25), who noted significant increases in pulmonary blood flow in humans during submaximal supine

exercise with minimal changes in pressure. In addition, it has been shown that the pulmonary capillary blood volume increases in exercise secondary to opening of parallel channels (4, 21). These changes would cause an increase in the compliance peripherally. The resultant increase in the arteriolar cross-sectional area probably also explains the trend (P = 0.06) toward a decrease in the PVR. Theoretically, this trend may occur because there is redistribution of the vascular volume when supine causing an increase in the arteriolar cross-sectional area. Thus the change is not as great as expected with upright exercise (5). There are several limitations to the present study. Patients in this study received an elective cardiac catheterization because of atypical chest pain syndromes. Therefore, these patients may not necessarily represent normal patients hemodynamics at rest or with exercise. Also, because the parameter estimation procedure is based on the three-element windkessel model, it does not analyze the reflected waves that may represent a major portion of the pressure and flow waveform of the pulmonary system, and one might anticipate erroneous estimations of the parameters. However, given the excellent correlation of all three parameters with previously reported values at rest, and given the high correlation coefficients and low relative errors by the model, the model does describe the circulatory dynamics appropriately. It is possible, however, that the pulmonary circulation may be better defined by a distributed circulatory model (I) that attempts to compensate for the reflections in the analysis. Another limitation is that no previous study has analyzed these parameters during exercise; therefore, the explanation of the physiological mechanism during exercise is hypothetical. This study is one of the most extensive evaluations to date of pulmonary vascular dynamics at rest and with exercise in a patient population without overt organic heart disease. This study also demonstrated the success-

Hl828

PULMONARY

ARTERIAL

COMPLIANCE

IN

HUMANS

ful use of a parameter estimation procedure based on the three-element windkessel model to model pulmonary circulatory dynamics in humans. This approach, although invasive, is possibly the simplest and most accurate method currently available for the evaluation of the arterial compliance of the pulmonary circulatory system in vivo. In humans, this model is capable of describing the pulmonary vasculature system more completely than previously and may prove to be beneficial in the evaluation of pulmonary hypertension (reactive vs. fixed hypertension) as well as perhaps defining a better therapeutic approach to patients with pulmonary hypertension. Studies are ongoing to test the applicability of a distributed model to the pulmonary vascular tree and to evaluate subjects with pulmonary hypertension.
We are grateful for the graphics work by Roberto Rios of the Department of Clinical Investigation, and we appreciate the editorial assistance of Margaret Latham in preparation of the manuscript. This work was supported in part by a grant from North American Treaty Organization Scientific Affairs Division RG 86/006, N. Westerhof and R. D. Latham. The views expressed herein are the private views of the authors and are not to be construed as official or representing those of the Department of the Army or the Department of Defense. Address for reprint requests: D. M. Slife, Cardiology Service, Beach Pavilion, Brooke Army Medical Center, Fort Sam Houston, TX 782346200. Received 9 June 1989; accepted in final form 16 January 1990.

8. 9.

10.

11.

12.

13.

14.

15.

16.

17.

19.

REFERENCES 1. BURRATTINI, R., AND S. DICARLO. Effective length of the arterial circulation determined in the dog by aid of a model of the systemic input impedance. IEEE Trans. Biomed. Eng. 35: 53-61,1988. 2. COX, R. H. Comparison of linearized wave propagation models for arterial blood flow analysis. J. Biomech. 2: 251-265, 1969. 3. ENGLEBERG, J., AND A. DUBOIS. Mechanics of pulmonary circulation in isolated rabbit lung. Am. J. Physiol. 196: 401-414, 1967. 4. GLAZIER, J. B., J. M. B. HUGHES, J. E. MALONEY, AND J. B. WEST. Measurements of capillary dimensions and blood volume in rapidly frozen lungs. J. Appl. Physiol. 26: 65-76, 1969. 5. GURTNER, H. P., P. WALSER, AND B. FASSLER. Normal values for pulmonary hemodynamics at rest and during exercise in man. Prog. Respir. Res. 9: 295-315, 1975. 5a. HALES, S. (Editor). StatisticaL Essays Containing Haemostatics. London: Wilson and Nichol, 1969. 6. LATHAM, R. D., B. J. RUBAL, AND R. S. SCHWARTZ. Gravitational influence of systemic arterial system in normal man. Circ. Res. 54: 666-673,1984. 7. LATHAM, R. D., B. J. RUBAL, N. WESTERHOF, P. SIPKEMA, AND 20.

21.

22.

23.

24.

25.

R. A. WALSH. Nonhuman primate model for regional wave travel and reflections along aortas. Am. J. PhysioZ. 253 (Heart. Circ. Physiol. 22): H299-H306, 1987. MILLAR, H. D., AND L. E. BAKER. A stable ultraminiature cathetertip pressure transducer. Med. Biol. Eng. Comp. 11: 86-89, 1973. MILLS, C. J., I. T. GABE, J. H. GAULT, D. T. MASON, J. Ross, JR., E. BRAUNWALD, AND J. P. SHILLINGFORD. Pressure-flow relationships and vascular impedance in man. Circ. Res. 4: 405-417, 1970. MILNOR, W. R., R. C. CONTI, K. B. LEWIS, AND M. F. OROURKE. Pulmonary arterial wave velocity and impedance in man. Circ. Res. 25: 637-649,1969. MILNOR, R. W., A. D. JOSE, AND C. J. MCGRAFF. Pulmonary vascular volume, resistance and compliance in man. Circulation 22: 130-137,1969. MURGO, J. P., AND N. WESTERHOF. Input impedance of the pulmonary arterial system in normal man. Circ. Res. 54: 666-673, 1984. NICHOLS, W. W., AND W. E. WALKER. Experience with the Millar PC-350 catheter-tip pressure transducer. Biomed. Eng. 9: 58-60, 1974. PATEL, D. J., J. C. GREENFIELD, JR., AND D. L. FRY. Hydraulic input impedance to aorta and pulmonary artery in dogs. J. Appl. Physiol. 18: 134-140, 1964. PIENE, H. Some physical properties of the pulmonary arterial bed deduced from pulsatile arterial flow and pressure. Acta PhysioZ. Stand. 98: 295-306, 1976. POLLACK, G. H., R. V. REDDY, AND A. WOODERGRAAF. Input impedance wave travel and reflections in the human pulmonary arterial tree: studies using an electrical analog. Trans. Biomed. Eng. Biomed. Eng. 15: 151-164, 1968. SHOUKAS, A. A. Pressure-flow and pressure-volume relations in the entire pulmonary vascular bed of the dog. Determined by twopart analysis. Circ. Res. 37: 809-818, 1975. TAYLOR, M. G. An approach to the analysis of the arterial pulse wave. I. Oscillations in an attenuating line. Phys. Med. Biol. 1: 258-269,1957. TOROOP, G. P., N. WESTERHOF, AND G. ELZINGA. Beat-to-beat estimation of peripheral resistance and anterior compliance during pressure transients. Am. J. Physiol. 252 (Heart Circ. Physiol. 21): H1275-H1283,1987. WEST, J. B., C. T. DOLLARY, AND A. NAIMARK. Distribution of blood flow in isolated lung: relation to vascular and alveolar pressures. J. Appl. Physiol. 19: 713-724, 1964. WESTERHOF, P., AND A. NOORDERGRAAF. Arterial viscoelasticity: a generalized model. Effect in input impedance and wave travel in the systemic tree. J. Biomech. 3: 357-379, 1970. WIENER, G., E. MORKIN, R. SKALAK, AND A. P. FISHERMAN. Wave propagation in the pulmonary circulation. Circ. Res. 19: 834-850, 1966. WOMERSLEY, S. R. Oscillatory flow in arteries: the constrained elastic tube as a model of arterial flow and pulse transmission. Phys. Med. Biol. 2: 178-187, 1957. Yu, P. N., G. W. MURPHY, B. F. SCHREINER, AND D. H. JAMES. Distensibility characteristics of the human pulmonary vascular bed. Circulation 35: 710-723, 1967.