Measurement of Flow in Single Blood Vessels IncludingCardiac Output by Local Thermodilution

By A. FRONEK, M.D., AND V. GANZ, M.D.

On the basis of local thermodilution a method for measuring blood flow in a single vesselwas elaborated. The measurement was accomplished by a special catheter with an in-jection orifice and a thermistor. The accuracy of the method was studied both in modeland in vivo experiments on dogs (in jugular vein, in carotid and pulmonary artery) andgood correlation was found with direct methods.

BLOOD flow in single vessels can be meas-ured by direct methods such as rota-

meter, drop-flow meter, Rein-thermostromuhr,Venturi principle, electromagnetic flow meterand bubble-floAV meter.1 Such methods areusually characterized by great accuracy, buttheir applicability is considerably limited inman by the fact that the appropriate vesselmust be made accessible. Blood flow in someregions was determined by methods applyingStewart-Hamilton's principle.2'3 On this back-ground we attempted to elaborate a methodfor measuring blood flow in any single vessel.

MethodsPrinciple

Measurement in a single vessel is made possibleby mixing an indicator with blood at the site ofinjection, and then detecting the resulting changein the immediate neighborhood of the site of mix-ing. The injecting orifice and detector were lo-cated on a specially prepared catheter. For tech-nical reasons a thermal indicator was chosen,introduced into dilution technics by Fegler.4 Athermistor was used for detecting temperaturechanges.

Method of InjectionA condition for correct measurement is homoge-

neous mixing of the indicator with blood through-out the cross-section of the vessel. This can bedone by transforming laminar into turbulent flow,if the Reynolds number of the injected indicatoris increased above a critical value. In model ex-

From the Institute for Cardiovascular Research,Prague-Krc, Czechoslovakia.

Presented at the 2nd and 3rd Meeting of theCzechoslovak Physiological Society (1958-HradecKralove), (1959-Brno) and published as abstracts inCsl. fysiologie 7: 455, 1958 and 8: 189, 1959.

Received for publication August 11, 1959.

Circulation Research, Volume VIII, January 19(0 175

periments homogeneity of mixing was judged ac-cording to whether the same dilution curves wereobtained for a constant position of the injectingorifice, but with the thermistor placed at differentpoints of the cross-section of the tube. Threemethods of injection were used to achieve reliablemixing. It was found that on injection from thetip of the catheter, correct mixing could not beachieved if the tip pointed against the wall of thetube, even when the Reynolds number of the indi-cator was 5000. On injecting from 4 orifices,placed radially in the wall of the catheter, themixture was also inhomogeneous if the catheter waslocated close to the wall of the tube. When meas-uring in vivo, the possibility that injection mightoccur against the wall of a vessel must be takeninto consideration. We therefore used a catheterwith a curved tip, and with an injection orificelocated in the concavity (fig. 1). TVith this ar-l'angement, the. indicator jet is oriented towardsthe longitudinal axis of the. vessel, against the di-rection of the current, at an angle of 30 .to 45°.In order to satisfy this condition, the oblique ori-fice in the wall of the catheter must have a certainminimum length (1.5 to 2 mm.).

The Effect of Injection on FlowDue to the small distance of the thermistor from

the injection orifice, part of the dilution curve isregistered during the injection of the indicator. Inorder to calculate the flow correctly, it was neces-sary to know how the injection of the indicator in-fluences flow below the site of injection, i.e., inthe region of the thermistor. This question wasstudied both in vivo in arteries and veins, and inmodel experiments. The changes in flow duringinjection were studied with a rotameter or drop-flow meter connected with the appropriate vessel,the point of injection being upstream from theflowmeter. In this way we found that during in-jection the flow in arteries either decreases or doesnot change. The decrease in the' flow increaseswith the linear velocity of flow and the kineticenergy of the indicator, These relations were-veri-

by guest on May 29, 2018

http://circres.ahajournals.org/D

ownloaded from

176 FRONEK, GANZ

OWECTOW OF flUJOO FWT THERfiSt Table 1

BFigure 1

A-downstream catheter. B-upstream catheter.

fied in model experiments. If the kinetic energyof the injected liquid indicator was smaller than13,000 Gm./cm.2/sec.~2, there was no change inflow in the region of the thermistor, even for alinear velocity of 356 cm./sec, which in vivo doesnot enter into consideration. For correct calcu-lation, therefore, an indicator kinetic energygreater than 13,000 Gni./cm.2/sec.~2 must not beused.

In veins, however, flow in the region of thethermistor increases during injection. An increasein flow during injection in the model experimentswas achieved by narrowing the tube above thesite of injection. The increase in flow in veinsis probably caused by injection against the "pe-ripheral resistance," similarly as in the model ar-rangement.

IndicatorAs an indicator 5 per cent glucose or physiolog-

ical saline at a temperature of 18 to 22 C wasinjected. The rate of injection was chosen so thatReynolds number was at least 3,000, while in thearteries we had to keep the condition that thekinetic energy did not exceed 13,000 Gm./cm.2/sec."2. The corresponding data of indicator in-jection are given in table 1.

The volume of the injectate may be chosen ac-cording to the expected blood flow. For measuringblood flow in the range of tens of ml. we used upto 1 ml., in the range of hundreds up to 3 ml. andfor flows in the range of liters up to 5 ml. of in-jectate. The dead space of the catheter must besubtracted from the injected volume. Since partof the catheter is external to the body and thecontained liquid cools off, the catheter must befilled immediately before injection with uncooledblood drawn in with a second syringe. Whenpassing through the catheter the indicator isslightly heated. The heating depends on the lengthof the catheter, the length of the part introducedinto the vessel, the amount of indicator and itsrate of flow through the catheter, and the differ-ence in temperature between the blood and the

Diameter of injectionorifice in mm.

Injection ratein ml./sec.

0.80.70.60.50.40.3

6.5—4.05.0—3.03.5—2.52.5—1.51.5—1.01.0—0.6

indicator. In extreme cases the heating may be asmuch as 1.5 to 2.0 C, which for a temperaturedifference between blood and indicator of 15 to20 C would mean an error of 10 to 13 per cent.The degree of heating of the indicator, however,can easily be determined. After completingmeasurements, that part of the catheter which wasin the vessel is immersed in water at the tempera-ture of the blood during measurement. Afterdrawing in, the same amount of indicator at thesame temperature as during measurement is in-jected into the catheter. The liquid leaving thecatheter is caught in a small bowl and its tempera-ture measured.

Example of calculation of indicator temperatureafter passage through catheter (tj) :

Injected, 7 ml. of indicator at temperature 2.1C. Dead space of catheter 2 ml. Blood tempera-ture 37 C. Temperature of liquid caught in bowl26 C.

t : = 2 1 . 6 C .

Detector and Its LocationTo detect the temperature changes we used a

thermistor, with a temperature coefficient of 3 to3.5 per cent. The thermistor is carefully insu-lated with a layer of PVC or silicone. The insu-lation of the thermistor is tested in the followingmanner: One terminal of the thermistor is con-nected through a microammeter to a source of d-cvoltage (of the order of 5 V), while the secondpole of the source is immersed in physiologicalsaline. If a deflection appears on immersing thethermistor into this solution, the insulation mustbe regarded as faulty; it can easily be repairedwith a thin film of PVC dissolved in cyclohexanon.Quick drying is achieved by heating to about 70 Cfor a few minutes. The insulating layer must bethin so that the time constant of the thermistoris as short as possible. The time constant doesnot influence the accuracy of measurement (calcu-lation is carried out on the basis of the area de-scribed), but with a short time constant the curvesare more ideal, and the demands made on the

Circulation Research, Volume VIII, January 1960

by guest on May 29, 2018

http://circres.ahajournals.org/D

ownloaded from

LOCAL THERMODILUTION METHOD 177

sensitivity of the recording equipment smaller. Thetime constant of thermistors used in our experi-ments was 0.2 to 0.5". A small working current(of the order of 1 niA) must be used, since athigher currents the thermistor is overheated. Inthis state the resistance of the thermistor dependsnot only on the temperature of the medium, butalso on the velocity of the flowing blood. The re-sistance of thu thermistor was measured with abridge circuit and changes were recorded. Thenon-linearity of the calibrating curve of the ther-mistor was compensated by a. logarithmic compen-sation. In order that the flow could also be meas-ured in shorter vessels, the thermistor was locatedwithin 5 to 10 mm. of the injection orifice. Toprevent close contact between the thermistor andthe wall of the vessel, which might influence theaccuracy of measurements, the thermistor is alsomounted in the concavity of the catheter. If thecatheter is introduced in the direction of the stream,the thermistor must be located distally to the in-jection orifice (downstream catheter—fig. 1 A) ;when introducing the catheter against the directionof the blood stream in a vessel the flow in whichis to be measured, the thermistor must be locatedproximally to the injection orifice (upstreamcatheter—fig. I B ) . As stated above, the termi-nal part of the catheter should be bent. This canbe done by applying a layer of PVC to the con-cavity of the bend.

Derivation of Formulae for Calculating FlowArterial

If the indicator is injected into the arteryagainst the stream with a kinetic energy of 13,000Grm./cm.2/sec"

2 or less, flow below the site ofinjection does not change. This means that in theregion of mixing the indicator replaces an exactlycorresponding amount of blood.

During homogeneous mixing, blood of initialtemperature tb is cooled, and the indicator of tem-perature t, is warmed to an average resulting tem-perature t. The blood taking part in the mixingloses the same number of calories as are gained bythp indicator. Therefore

( j - • T - m) • (tb - t ) Sb • Cb = m( t - t , ) Si • C,

(Caloric loss of blood) (Caloric gain of indicator)

F = flow in ml./l min., T = time in seconds be-tween beginning and end of dilution curve, Sb, S,= specific heat of blood and indicator respectively(cal/g), Cb, Ci = specific gravity of blood andindicator respectively (g/cm.3), m = amount ofindicator in ml.

t - ti can be substituted by : tb - ti - (tb - t)

tb - t can be substituted by : — • f

Circulation Research, Volume VIII, January 1960

F mt/min r ml/min

torw

Figure 2During indicator injection (T sec.) in veins, fl

F

( —r?' T ml.) is increased by a portion of the indi-

cator (x ml.).

where A = area below dilution curve in mm.2, r =speed of registration paper in mm./sec, f = tem-perature change in C corresponding to 1 mm. ofcurve amplitude.

On substitution, the final formula for calculat-ing the flow in arteries is

m • 60 • r ( t b - t i ) Si ' CiF =

VenousA • f • Sb • Cb

During injection of indicator (T sec.) againstthe blood stream in veins the flow at the time of

injection — T, below the injection site is in-

creased by a portion of the indicator (x ml.) (fig.2). The time of injection (T) agrees with thetime between the beginning of the dilution curve(C) and the beginning of the return of the curveto the initial level (D). After injection has ended,the flow is again at its original level. Accordingto this, the area under the dilution curve can bedivided into 2 parts: At and A2. Area Al cor-responds to the increased flow. Area A2 corre-sponds to the unchanged flow, i.e., part of theblood is replaced by the rest of the indicator(m—x) ml. According to the area A2, flow canbe expressed by the formula derived for arteries.

• 60 » r « ( t b - t i ) » Si « C

A a * I * S b * Cb

On the basis of the caloric interchange betweenblood and indicator, it is also possible to writefor the time TF — —— • x • (U - t) Sb • Cb = x (t - t,) • Si • Ci

t - ti can be substituted by : U - ti - (tb • t)

tb - t can be substituted by : —•T • r

After substitutionF • A, • f • Sb

F _ ( I.

60= x ( t b - t , - S, • C,

II.

by guest on May 29, 2018

http://circres.ahajournals.org/D

ownloaded from

178 FRONEK, GANZ

Figure 3Calculation of flow in an artery:

f =}•''' — 0.15 C/mm.; r — 10 mm./sec; t{

22 C ; in = 5 ml.; hematocrit : 40%

F _ m ' 60 ' r (th—h) ' S, ' C-, _A • / • Sb • Cb

_ 5 • 60 • 10 (37—22) • 0 935 • 1.0181000 ' 0.15 • O.S7 ' 1.058

= 320 ml./min.

tooo

3000

/OOO

1000

0

. !

1 /

/

1000

/ . 0 5• 0V

ROTAMETSR AC/MM

2000 3000

MM WATERMM BEEF BLOODMM WiTERKM mTER

WOO

f • Si, • G, •

Figure 4Comparison of LTD and rotaineter values inmodel experiments.

After solving equations I and II, the final formulafor calculating flow in veins is:•̂ _ _ m • 60 • i • Si • Ci

/ A2 A , - i ' i \\ t i , - ti (tb - t,) • T • r - A, • f/

The validity of the equations for blood flow in ar-teries and veins was confirmed on a correspondingarrangement in model experiments.

Procedure During MeasurementThe measuring catheter (upstream or down-

stream) is introduced into the vessel, the flow ofwhich is to be measured. The initial temperatureof the blood (th) is found by reading the resist-ance of the thermistor on the bridge. This valueis simultaneously recorded, and eventual changesin the initial temperature of the blood up to themoment of injecting the indicator can then be read

Table 2Specific Heat and Gravity of Blood as Function ofHematocrit. An Example of the Calculation ofFlow is given in Fig. 3

Hematocrit%

Specific heat of blood Spec, gravity of bloodcal./Gm. Cm./cm.

30354045505560

0.890.880.870.865O.Sfi0.850.84

1.0401.0531.0581.0611.0641.0681.073

from the record. The catheter is filled with un-cooled blood by drawing in with a syringe and theindicator is injected with a second syringe. Thedilution curve is recorded and the resistance ofthermistor is calibrated by adding an external re-sistance. After measurement is completed the de-gree of heating of the indicator during passagethrough the catheter is found.

Constants Needed for Calculation5 per cent glucose: spec, heat 0.965 cal./Gm.,

spec, gravity 1.018 Gm./cm.3. The specific heatand specific gravity of the blood depend on thehematocrit. The values of the specific heat of theblood, calculated according to Mendlowitz's formu-la,0 are given for hematocrit values of 30 to 60per cent in table 2. The same table gives the val-ues for the specific gravity of the blood, calculatedon the assumption that in human blood0 and dog'sblood7 the specific gravity of the plasma is 1.026and the specific gravity of the erythrocytes 1.10S.

ResultsModel Experiments

The model measurements were carried outto verify the accuracy of the method. Wateror citrated beef blood at a temperature of37 to 40 C flowed from a reservoir under ahydrostatic gradient in glass and rubbertubes. The magnitude of the flow was regu-lated by a screw clamp and read off on arotameter, calibrated with a reproducibilityof ±1 per cent. Measurements were carriedout in tubes of different diameter 5, 10, 15.6.and 24.4 mm., placed during measurement inthe reservoir of warm liquid. As indicatorwater was injected; whenever measurementswere made with beef blood, 5 per cent glucoseor physiological saline solution was used.

Figure 4 gives a comparison of the values

Circulation Research, Volume VIII, January I960

by guest on May 29, 2018

http://circres.ahajournals.org/D

ownloaded from

LOCAL TIIERMODILUTION METHOD 179

Figure 5Comparison of LTD and rotameter values in thecarotid artery.

obtained by the local therinodilution (LTD)method and those read off on the rotameter.A statistical analysis of 111 comparativemeasurements did not show a systematic devi-ation of the LTD values from the actualvalues of the flow either as a whole (A percent = 0.189 is not significant, t< l ) , or as afunction of the diameter of the tube in whichmeasurement was carried out. The standarddeviation of both methods was ±3.14 per cent.

Measurement of Blood Flow in DogsMeasurement in Carotid Artery

The experiments were carried out on dogs.Both the carotid artery and the left jugularvein were dissected to a length of 5 to 6 cm.under chloralose anesthesia. A Y-shaped can-nula was introduced into the right carotidartery. An upstream catheter was introducedinto the artery through one arm of the can-nula and the other arm was connected to therotameter. The blood passing through therotameter returned to the left carotid arteryor to the jugular vein (in order to achievehigher flows). Coagulation of blood was pre-vented with heparin. After each measurementcalibration of the rotameter value was carriedout at least twice with a reproducibility ofabout ±5 per cent.

CONTROL ME7HOOC

5& 165 ! ^5 lad ao

Figure 6Comparison of LTD and, control method valuesin the jugular vein.

Altogether 70 comparative measurements ina flow range of 21.5 to 538 ml./min. werecarried out on 5 dogs. Figure 5 gives a com-parison of the values obtained by the LTDmethod and those read from the rotameterat the moment of injecting the indicator.According to statistical analysis the 2 meth-ods did not differ systematically (A p e r cent= +0.071 per cent, not significant, t< l ) . Thestandard deviation of both methods was ±4.61per cent with a maximum deviation of +10.95per cent and —11.1 per cent. :

Measurement in Jugular Vein

Measurements were carried out on 7 anes-thetized dogs. A Y-shaped cannula was intro-duced into the jugular vein, prepared to alength of 5 to 6 cm. An upstream catheter wasintroduced into the A'ein through one ai'in ofthe cannula and through the other bloodflowed out during measurement into gradu-ated cylinder for 30 sec.

Altogether 108 comparative measurementswere carried out for a range of flows from9.6 to 204 ml./min. (fig. 6). The values cal-culated by the LTD method did not deviatesystematically from those found rotametri-cally (A per cent = +0.096 per cent, notsignificant, t< l ) . The standard deviation of

Research, Volume VIII, January I960

by guest on May 29, 2018

http://circres.ahajournals.org/D

ownloaded from

180 FBONEK, GANZ

c o for u^m

Figure 7Comparison of Fick and LTD values.

both methods was ±5.23 per cent, the maxi-mum deviations being +11.85 per cent and—12.2 per cent.Measurement in Pulmonary Artery

On 7 dogs under chloralose narcosis 86comparative measurements were carried out.Fick method: Pure oxygen was led through atracheal cannula and consumption was deter-mined by a Koth-Benedict spirometer. A de-flection of the spirometer pen by 1 mm.represented a consumption of 20.4 ml. of oxy-gen. To test for leaks, at the end of each ex-periment the apparatus was run for 10 min.with the dead dog in the circut. Heparin (10mg./Kg.) was used as anticoagulant. Mixedvenous blood samples were drawn with acatheter from the root of the pulmonary ar-tery for 1 min. and arterial blood from thefemoral artery. Oxygen saturation was deter-mined on a hemoreflectometer according toBrinkman and Zijlstra. The hemoglobin wasdetermined photoelectrically for each sampleof arterial blood.

LTD method: A downstream catheter wasintroduced from the jugular vein under X-ray and blood pressure control into the rootof the pulmonary artery. The following pro-cedure proved suitable for preventing theshifting of the catheter into a branch of thepulmonary artery as a result of ventricularcontraction: on introducing the catheter the

Figure 8Deviation of LTD values from the average ofLTD and Fick value in %.

position at which a change from ventricularto arterial blood pressure occurs is deter-mined. The catheter is then pushed furtherinto the branch and again drawn out so thatthe tip of the catheter is about 1 cm. distallyfrom the position ascertained. With this pro-cedure the loop of the catheter in the rightventricle is much smaller, and the probabilityof a shift of the catheter into the branch isminimized. However, occasional control ofthe position of the catheter is recommended.As indicator, 3 ml. of 5 per cent glucose atroom temperature was injected.

Blood could not be drawn simultaneouslywith the recording of the dilution curves asthe indicator was injected through the samecatheter into the pulmonary artery. Thespirometer being constantly attached to thecircuit, we measured cardiac output by bothmethods at 15 min. intervals. Cardiac out-put was decreased by repeated removal ofblood in amounts of 30 to 300 ml.

Calculation of the flow in the pulmonaryartery from the thermodilution curves ob-tained was carried out according to the for-mula valid for arteries.

Since the flow in the pulmonary artery ex-hibits rapid periodic fluctuations, we regu-larly made 3 to 4 LTD measurements a min-ute or 5 to 8 measurements in 2 min. Theaverage of the values obtained in this waywas compared with the value calculated ac-cording to the Fick method, which representsthe average flow over a few minutes.

The fluctuation of the different LTD valuesabout the average was ±6.7 per cent. Figure 7gives a comparison of the values of cardiacoutput determined by both methods.* Stand-

*The table representing these data can be sent onrequest to the authors.

Circulation Research, Volume VIII, January I960

by guest on May 29, 2018

http://circres.ahajournals.org/D

ownloaded from

LOCAL THERMODILUTION METHOD 183

10 sec

ADRENALINE0,1 mg

Figure 9Changes of arterial blood 'pressure and coronary sinus flow, recorded by LTD method,after 0.1 mg. adrenaline.

ard deviation of the 2 methods was ±9.30per cent with greatest deviation +23.0 percent and —22.4 per cent. There was no sys-tematic difference between the two methods(A per cent = +0.49, insignificant, t< l ) .Figure 8 represents the percentage deviationof the thermodilution values from the aver-age value obtained from the two methods.

DiscussionThe principle of local dilution and local

detection, which forms the basis of this meth-od for measuring blood flow in a single vessel,required the solution of 2 main problems: 1)mixing at the site of injection, and 2) influ-ence of indicator-injection on flow.Mixing

When measuring cardiac output by theclassical indicator-dilution method, mixing isnot a particular problem. The indicator isthoroughly mixed with the blood in the car-diac chambers and in the pulmonary bed.Despite some theoretical objections,8-9 themixing is obviously sufficient since the cardiacoutput can be measured with satisfactory ac-curacy. When measuring blood flow in asingle vessel, however, homogeneous mixingmust take place immediately at the site ofinjection. A condition of this is turbulentflow. However, calculations10' - and experi-ments10' u have shown that there is probablyturbulent flow only in some sections of thevascular bed, e.g., in the root of the aorta.

Circulation Research, Volume VIII, January I960

The linear velocity of the blood stream invessels therefore had to be greatly increasedin order to exceed the critical Reynolds num-ber at which laminar flow is transformed intoturbulent flow. With injection of the indica-tor into the blood stream, however, turbu-lence can also be obtained by increasing theReynolds number of the indicator above thecritical limit (about 1000). Under the presentcondition, the Reynolds number of the indi-cator was 3,000 to 5,000; moreover, injectingthe indicator against the blood stream is afurther factor facilitating mixing. For re-liable mixing throughout the cross-section,however, particularly in wider vessels, suit-able spatial orientation of the indicator jetalso had to be ensured, i.e., the direction inrelation to the axis of the vessel. This Avasdone by bending the terminal part of thecatheter, as described above.

The Effect of Injection of Indicator on Blood FlowBecause of the proximity of the thermistor

and injection orifice, part of the dilutioncurve is recorded while the indicator is stillbeing injected. It was necessa^ to investigatehow, and under what conditions, blood flowin arteries and veins changes as a result ofinjecting the indicator. The data enabled usto derive relations for calculating flow bothin arteries and in veins. As a result of thesea maximum permissible kinetic energy of in-jection into arteries was determined. Atten-

by guest on May 29, 2018

http://circres.ahajournals.org/D

ownloaded from

182 PRONBK, GANZ

tion must also be paid to the changes in bloodflow, arising when using modifications of theclassical dilution principle with constant in-jection of the indicator, where injection anddetection occur simultaneously.

This method enables flow to be measuredin any part of the vascular bed into which acatheter can be introduced under more phy-siological conditions than has been possibleup to now. When measuring in small vessels,of course, a catheter of an appropriate sizemust be used so that its presence should notaft'ect the flow in the vessel to any great ex-tent. Measurement of blood flow can be re-peated several times a minute, and one canthus study rapid changes in flow (fig. 9).Evaluation of the dilution curves is facili-tated by the fact that no ascertainable recircu-lation of the indicator occurs even when in-jecting into the pulmonary artery. This factcan be explained, apart from the considerabledilution of the indicator, by its escape fromthe vascular bed and by heat conduction fromthe vessels and their surroundings on passageof the indicator into further sections of thevascular bed.

SummaryA method was elaborated for measuring

blood flow in single blood vessels on the basisof local thermodilution. Measurements werecarried out with a specially prepared catheterwith an injection orifice and a thermistor fordetecting temperature changes.

The method was elaborated in model ex-periments (s.d. ± 3.14 per cent) and verifiedby comparative measurements in the carotidartery (s.d. ± 4.61 per cent), jugular vein(s.d. ± 5.23 per cent) and pulmonary artery(s.d. ± 9.30 per cent). The method permitsflow to be measured in any vessel into whicha catheter can be introduced, and enablesrapid changes in flow to be studied.

Summario in InterlinguaEsseva elaborate un methodo pro le mesuration del

fluxo de sanguine in vasos individual super le basedo tliermodilution local. Le mesurationes esseva ef-fectuate per medio de un specialmente preparatecatheter con un orificio de injection e un thermistor

pro le detection de alterationes de temperatura. Lemethodo esseva elaborate in experimeiitos con niodellos(deviation standard ±3,14 pro cento) e verificate inmesurationes comparative in canes in le arteria caro-tidic (deviation standard ±4,61 pro cento), le venajugular (deviation standard ±5,23 pro cento), e loarteria pulmonar (deviation standard ±9,30 procento). Le methodo permitte le mesuration del fluxoin omne vaso in que un catheter pote esscr intro-ducite. Illo rende possibile le studio dc alterationesrapide del fluxo.

References1. GREEN, H. D.: Methods in Medical Research I,

Chicago, The Year Book Publishers, 1948.2. ANDRES, R., ZIERLER, K. L., ANDERSON, H. M.,

STAINSBY, W. N., CARER, 0., GHRAYYIB, A. S.,AND LILIENTHAL, J. L., JR . : Measurement ofblood flow and volume in the forearm of man;with notes on the thcorj' of indicator-dilutionand on production of turbulence, hemolysis,and vasodilation by intra-vascular injection.J. Clin. Invest. 33: 482, 1954.

3. FEGLER, G., AND HILL, K. J.: Measurement of

blood flow and heat production in the splanch-nic region of the anesthetized sheep. Quart. J.Physiol. 43: 189, 1958.

4. —: Measurement of cardiac output in anesthe-tized animals by a tliermo-dilution method.Quart. J. Exp. Physiol. 39: 153, 1954.

5. FELIX, W., AND GROLL, H.: Die Mcssung des

Blutstromes mit Thermistoren. Ztschft. Biol.105: 208, 1953.

6. MENDLOWITZ, M.: The specific heat of humanblood. Science 107: 97, 1948.

7. DALY, I. DE BURGH, EGGLETON, P., ELSDEN, S.R., AND HEBB, C. O.: A biochemical study ofisolated perfused lungs with special referenceto the effect of phosgene. Quart. J. Physiol.33: 215, 1946.

8. BUCHER, K., AND EMMENEGGER, H.: Besoiuler-

heiten der Lungcnclurchblutiing. Arch. f.Kreislaufforschg. 18: 94, 1952.

9. LOCHNER, W.: Untersuchungon iiber die unvoll-stiindige Misehung des Blutes im rcclit.cnHerzon mit Hilfe von Temperaturinessungcn.Pfliig. Arch. 256: 296, 1953.

10. HESS, W. R.: iiber die periphcre Begulierung derBlutzirkulation. Pfliig. Arch. 168: 439, 1917.

11. RALSTON, J. H., TAYLOR, A. N., AND ELLIOTT,

H. W.: Further studies on streamline bloodflow in the arteries of the cat. Am. J. Physiol.150: 52, 1947.

12. PETERSON, L. H., HELRICH, M., GREENE, L., TAY-LOR, C, AND CHOQUETTE, G.: Measurement ofleft ventricular output. J. Appl. Physiol. 7:258, 1954.

Circulation Research, Volume VIII, January I960

by guest on May 29, 2018

http://circres.ahajournals.org/D

ownloaded from

A. FRONE and V. GANZThermodilution

Measurement of Flow in Single Blood Vessels Including Cardiac Output by Local

Print ISSN: 0009-7330. Online ISSN: 1524-4571 Copyright © 1960 American Heart Association, Inc. All rights reserved.is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231Circulation Research

doi: 10.1161/01.RES.8.1.1751960;8:175-182Circ Res. 

http://circres.ahajournals.org/content/8/1/175World Wide Web at:

The online version of this article, along with updated information and services, is located on the

  http://circres.ahajournals.org//subscriptions/

is online at: Circulation Research Information about subscribing to Subscriptions: 

http://www.lww.com/reprints Information about reprints can be found online at: Reprints:

  document. Permissions and Rights Question and Answer about this process is available in the

located, click Request Permissions in the middle column of the Web page under Services. Further informationEditorial Office. Once the online version of the published article for which permission is being requested is

can be obtained via RightsLink, a service of the Copyright Clearance Center, not theCirculation Research Requests for permissions to reproduce figures, tables, or portions of articles originally published inPermissions:

by guest on May 29, 2018

http://circres.ahajournals.org/D

ownloaded from