gina pectoris, and 10 patients treated postopera- tively after ..... Since the geometry of the mitral annulus ... Goldberg SJ, Dickinson DF, Wilson N: Evaluation.
」Clin Ultrasound 20:263-270,May 1992 CCC 0091-2751/92/04263‐ 08$04.00 ◎ 1992 by Johnヽ Viley&Sons,Inc
Doppler Echocardiographic Measurement of Cardiac Outputin Man Using Mitral
Annulus Method
Hiroyuki Shimamoto,MD,*Yoriko Shimamoto,PID,十 11iroyuki Kito,MD,‡ Kohei Kawazoe,MD,I TsuyOShi Fttita,MDl,and DIlitsunori Okamoto,MD§
Abstract: The rnitral inflow lnethod of measuring cardiac output with pulsed Dop‐ pler two‐ dilnensional echocardioraphy Was developed and validated against the ther‐ modilution technique in 42 patients.A Initral inflow method combined the velocity of left ventricular inflow at the lnitral annulus(apical 10ng― axis view)with the cross‐ sectional area of the annulus calculated froln its diameter(parasterna1 long― axis
view).A good correlation was observed between thermodilution and Doppler ineasure‐ ments of cardiac output(′ =0.93)The rati0 0f Doppler measurements to thermodilu‐ tion lneasurements will be between O.76 and l 15 for about 959ら
ofcases There was a
good correlation between percentage change in thermal cardiac output and those in Doppler cardiac output(r=0.95);the lilnits of agreement were-11.5%to 10 5%,sug― gesting that this lnethod can be lnost useful for assessing relative changes in cardiac
output. Indexing Words: Cardiac output・ Doppler echocardiography・ tral flow・ Mitral in■ ow method
Translni―
The lrnost accurate lnethods for lFneaSurement of
which tends to limit the applicability of these
cardiac output are considered to be the Fick
techniques in clinical practice.
method,the dye― dilution method and the ther¨ modilution lnethod.In recent years,quantitative
uation of transIInitral flow velocities with trnini‐
Doppler echocardiography has been used in
mal angulation between the ultrasound beam
many centers.l The principle of this lnethod is to lrneasure the diallneter of the ascending aorta or
and flow,and lnay be ideal for llneasurements of
The apical window provides access to the eval―
cardiac output.Fisher et al.2 applied a correction
pulllnonary artery by two¨ dilrnensional echocardi―
factor derived from the mitral valve M¨ mode to
ography and to measure blood flow velocity by the Doppler method. Images adequate for mea‐
the maximal mitral valve orince area deter¨
surement of a cross― sectional area of the ascend―
accurately trneasured llnitral inflow volume from
ing aorta fronl the suprasternal notch or of the
the cardiac apex in experimental animals.How‐ ever, the method described by Fisher et al. for measuring lnitral annular area is complex and
pull■ onary
artery fronl the parasternal window
cannot always be obtained in adult patients,
*Department of Cardiovascular Surgery,National Cardiovas― cular Center,Suita 565,5-7-1,FttishirOdai,」 apan Present:
Division of Physiology and Pharmac01ogy, Departinent of BiOmedical Sciences,Faculty of Health Sciences,McMaster University,Hamilton,Ontario,Canada lDepartment of ln― ternal Medicine, Nakamura Hospital, HirOshilna, Japan lDepartment of CardiOvascular Surgery,National Cardiovas― cular Center Suita」 apan §Department oflnternal Medicine, HirOshilna Prefectural Hospital, Hiroshima,」 apan For re―
prints cOntact Hiroyuki Shimamoto,MD, 1-5-17,Kusunoki, Nishi― ku,IIiroshilna 733,」
apan.
mined frorn cross― sectional echocardiograms and
technically applicable to a small subgroup of pa‐ tients studied.Lewis et al.3 deVeloped a lnodinca‐ tiOn ofFisher's lnethod for FneaSuring translrnitral flow at the level ofthe lnitral annulus and one an‐
nular diameter from the apical four‐ chamber view.Although their llnethod sillnplines previous methods of measuring cardiac output,llnitral an‐ nulus obtained froln the cardiac apex is located far fronl the transducer,which lnay lnake it dif‐ ■cult to measure the llnitral annular diameter.In contrast,the parasterna1 long‐ axis view has been
useful in the assessment of the llnitral valve 263
SHIMAMOTO ET AL.
264
。 ‘・ ，
こ
蝋 鬱
魃 1嗽
FIGURE l Measurements of parameters from mlra1 low pattern(A)Apica eft vent百
cularlong axis view The sampling vo ume{SV)for Dop―
pleris indicated by a white arrow SV is setin the center ofthe mitral ring(LV=ieft ventricle′ LA=left atrium′ Ao=ascending aorta′
RV=right
ventricle)(3)Dopplerflow signal recorded at SV shown in{A)The dots of Doppler flow signal shovv the instantaneous mean mitralflow veloc―
ly MIra1 low velocly― ■me integral is measured(ECG=eiectrocardiogram′ 刊ow components away from the transducer)
TOWARD=刊 ow
components toward the transducer′
」OURNAL
AWAY=
OF CLINICAL ULTRASOUND
CARDIAC OUTPUT BY MITRAL ANNULUS METHOD
FIGURE l{C}Measurement of anterior― LA=left at● um′
265
posterior diameter(D)of the rnitral va ve annulus in the parasternallong‐ axis view (LV=left ventricle′
Ao=ascending aorta)
compared with the apical two― chamber or four‐
involved in the Doppler deterllnination of cardiac
chamber vlew.
output were blinded to the result of the ther_ modilution deterlnination.Images were obtained
In this investigation、 ve have developed and validated the approach to llneasuring translnitral
using a directional pulsed DOppler■ owmeter in‐
flow with single annular diameter obtained frOm
vide angle phased array echo‐ corporated in a 、
the parasternal long― axis view.
graph (Yokogawa 1/1edical Systems U‐ sonic hIIodel RT5000,Tokyo,」 apan,or Toshiba SonO‐ layergraph 65A and Toshiba SDS21A,Tokyo,」 a‐
METHODS SutteCtS
pan).These systems use a fast Fourier transform spectral analyzer system,which displays the dif― ferent red blood cell velocities within the artery or
The clinical population consisted of 42 patients
cardiac chambers.The ultrasound frequency of
vomen, 58.5 ± 12.3 years of age) (24 men, 18 、
the Doppler■ oM/11neter was 2.5 ⅣIHz.The cut¨ off
who had undergone deterlnination of cardiac out¨ put by thermodilution either while in an inten‐
frequency ofthe high‐ pass■ lter was setto 400Hz.
sive care unit or as part of diagnOstic cardiac catheterization.In none of the patients、 vas clin¨
ical, echocardiographic, or Doppler evidence of llnitral stenOsis found.No patients had rnore than l+llnitral regurgitatiOn on left ventriculography (Sellers criteria)5。 r On pulsed Doppler echocar‐
The pulse repetition rate ranged fron■ lto 20 kHz for Yokogawa Ⅳledical Systems and 4 or 6 kHz for
Toshiba systellns.The depth ofsample volume was about 3 11nln.The smallest feasible angle was
maintained during all Doppler measurements.
diography6.All suttects ShOWed normal sinus rhythm.Twenty one patients were studied while
Assunling that the presumed direction oftranslni― tral■ ow is a linejoiningthe apex ofthe left ventri‐ cle to the center Ofthe lnitral annulus in the api‐ cal long― axis view, correction for angle of inci‐
in the intensive care unit being treated for a
dence was llnade in all measurements.The Dop―
myocardial infarction,1l patients treated for an― gina pectoris,and 10 patients treated postopera― tively after coronary artery graft surgery.After the nature and the purpose ofthe study were ex‐ plained,informed consent、 vas obtained.
pler signals were displayed silnultaneously with the electrocardiograIIl and M― mOde echocardio‐
Doppler Echocardiography All of the echocardiographic studies were per―
gram using a strip chart recorder(Yokogawa blledical Systems U― sonic Line Scan Recorder SR-02,or Toshiba LSR20A)at a paper speed of 50 mln/sec. Flo、 v velocity colnponents toward and a、vay froln the transducer were displayed above and below the basehne,respectively.
The transducer was placed at the apical iln,
formed within trninutes before the deterlnination
pulse and the apical long‐ axis view was imaged.
of cardiac output by thermodilution. Operators
The Doppler sample volume was placed in the
VOL 20.N0 4′ MAY 1992
SHllⅦ AMOT0
266
ET AL.
center of the lnitral ring.Lllitral in■ ow velocity
values were discarded and the 3 remaining val―
was recorded over a nulnber of cardiac cycles
ues were averaged, as 10ng as the variability of
(Figure lA).The mitral flow velocity― time inte―
these deterlninations did not exceed 15%. If
gral was measured by planilrnetry(Figure lB). Five beats 、 vere averaged for each determina―
there was excessive variability, 5 new deterllni―
tion.
peated.
The parasternal long― axis view of the left ven― tricle was obtained,which provided good visual¨ ization of the left ventricular and atrial cavities.
The lrnitral annulus diameter was taken as an
nations were made and the procedure was re―
DeterIIlination of Percent Changes in Cardiac Output
oblique line joining the twO sides of the annulus frorn the parasterna1 long― axis vie、 v with maxi‐
Changes in cardiac output induced by■
mal excursion of the lnitral valve leanets. The anterior and posterior landmarks of the lnitral annulus were recognized as the insertion of the
pressure or intra¨ aortic balloon pumping were
uid infu¨
sions or associated with positive end― expiratory measured in each of 42 sutteCtS using thermodi¨ lution and Doppler technique.
anterior and posterior leaflets,respectively,from the real― tilne,two‐ dilnensional
image.An ECG―
gated stop frame of the trnitral annulus at the tilne of peak rapid■ 1ling■ oⅥ″velocity on a t、 vo‐
dilnensional image was recorded,and the annu‐
lar diameter was lrneasured (Figure lC).The
Statistics
Data are expressed as mean ± standard devia‐ tion. Data were analyzed by linear regression
velocity when
analysis.Furthermore,the extent of the agree‐ ment between the lneasurements using the ther‐
measurements ofthe annular diameter were per‐ formed isjust after the time when maximal ini‐
modilution and Doppler rnethods、 vas assessed by the rnethod of Bland and Altlnan.9
tillne of peak rapid
■1ling ■ow
tial excursion of the initral valve occurs.7,8 NIlea―
surements frolln a lninilnuln of■ ve cardiac cycles
RESULTS
were averaged and the cross― seCtiOnal area of the
annulus was derived as πr2,where r represents half of the annular diameter.This method as‐ sumes a circular shape for the lnitral annulus and a constant cross― sectional area throughout diastole. Cardiac output was calculated as the product of the lnitral diastolic flow velocity¨ tillne integral and the cross¨ sectional area of the lni‐
tral annulus to yield stroke volume which was multiplied by the heart rate.
A second independent observer made cardiac output deterlninations in 15 patients to test for interobserver differences.Interobserver variabil―
ity was expressed as a percent error for each
No patients were excluded frolln the study due to poor quality echo or Doppler data.
Evaluation of Thermodilution and Doppler PIethods by Linear RegressiOn Analysis A high correlation was observed between ther_ modilution‐ derived cardiac output and Doppler― determined cardiac output over a wide range of cardiac outputs,with an r value of O.93(Thermal
CO[L/1nin]=1.11・ Doppler CO[L/1nin]-0.12,Fig‐ ure 2). There was an excellent correlation be― tween percent change in thermal cardiac output
measurement and was deterlnined as the differ‐ ence between the t、 vo observers divided by the
and that in Doppler cardiac output(%Thermal
mean value ofthe t、 vo observations.
3).The percent differences were 4.8± tween the 2 independent observers.
CO=0.97・ %Doppler CO+1.5,r=0.95,Figure 3.79ぢ be‐
Therlmodilution… Derived Cardiac Output
A Swan‐ Ganz thermodilution catheter(Ameri‐ can Edwards Laboratory)was inserted into the pull■ onary artery. Cardiac output was deter‐ Inined by thermOdilution using 5% glucose in water at O° C as the indicator.Computations 、 vere performed using a bedside thermodilution
cardiac output computer (American Edwards Laboratory,COC¨ 9520¨ A).Five measurellnents of cardiac output were obtained.The high and low
Evaluation of Thermodilution and Doppler Methods by the Method of Bland Altlnan Table l lists the results of this statistical analy‐ SlS.
The scatter of the differences between Doppler
and thermal cardiac outputs tended to increase as the cardiac output increased Therefore,the analysis of Bland and Altman was applied to the data after logarithtrnic transformation.Figure 4 」O∪ RNAL
OF CLINICAL ULTRASOUND
CARDIAC OUTPUT BY MITRAL ANNULUS METHOD (′
267
/min)
“Ｆ こ ①エト 一
ＰＣＯ ｏＬ① 住
葛 Ｅ ＬΦ 〓ト
０“ 一 ＯＬｏ〇
Ｃ一 ①”Ｃ”〓Ｏ
Ｐ⊃Ｑや⊃０
ｏヽ“
００
0
20
2468(2/min) Doppler CO
40
60
80 %
Percent Change ln Doppler Cardiac Output
FIGURE 2 Relationship betvveen Doppler cardiac output(Doppler CO)and thermodilution cardiac output(thermal CO)〔 γ=11l χ ― 012;′ =093(ρ
