evaluation of lv diastolic function from color m-mode ... · analysis of ﬁlling velocities...

J A C C : C A R D I O V A S C U L A R I M A G I N G V O L . 4 , N O . 1 , 2 0 1 1

© 2 0 1 1 B Y T H E A M E R I C A N C O L L E G E O F C A R D I O L O G Y F O U N D A T I O N I S S N 1 9 3 6 - 8 7 8 X / $ 3 6 . 0 0

P U B L I S H E D B Y E L S E V I E R I N C . D O I : 1 0 . 1 0 1 6 / j . j c m g . 2 0 1 0 . 0 9 . 0 2 0

Evaluation of LV Diastolic Function FromColor M-Mode Echocardiography

Kelley C. Stewart, MS,* Rahul Kumar, MD,† John J. Charonko, PHD,*Takahiro Ohara, MD, PHD,† Pavlos P. Vlachos, PHD,* William C. Little, MD†

Blacksburg, Virginia; and Winston-Salem, North Carolina

O B J E C T I V E S This study evaluated early diastolic filling dynamics using a semiautomated objective

analysis of filling velocities obtained from color M-mode echocardiography.

B A C K G R O U N D Diastolic function can be evaluated from color M-mode echocardiography by

measuring the early diastolic flow propagation velocity (Vp) from the slope of a single linear

approximation of an isovelocity contour. However, this method has limitations and may not accurately

represent diastolic filling.

M E T H O D S We used a semiautomated objective analysis of color M-mode echocardiograms from a

development cohort of 125 patients with varying diastolic function to quantify left ventricular filling

velocities. Early diastolic filling was not accurately described with a single propagation velocity; instead,

the rapid initial filling velocity abruptly decelerated to a slower terminal velocity. Then, we evaluated a

new measure of diastolic function in a separate group of 160 patients.

R E S U L T S Compared with normal filling, diastolic dysfunction with restricted filling had a lower

initial velocity (53 � 21 cm/s vs. 87 � 29 cm/s, p � 0.001), and the deceleration point occurred closer

to the mitral annulus (2.4 � 0.6 cm vs. 3.1 � 0.7 cm, p � 0.05). The product of the initial velocity and

the distance to the deceleration point from the mitral annulus, indicating the strength of the early filling

(Vs), was progressively reduced with diastolic dysfunction. In a separate validation cohort of 160 patients,

Vs better recognized diastolic dysfunction (classified by reduced diastolic intraventricular pressure

gradient, elevated pulmonary capillary wedge pressure, or elevated B-type natriuretic peptide) than

Vp did.

C O N C L U S I O N S Early diastolic flow propagation occurs with an initial rapid velocity that abruptly

decelerates to a terminal velocity. With diastolic dysfunction, the initial velocity is slower and the

deceleration point occurs closer to the mitral annulus than with normal filling. A new parameter that

combines these 2 effects (Vs) provides a more accurate assessment of diastolic function than the

conventional propagation velocity. (J Am Coll Cardiol Img 2011;4:37–46) © 2011 by the American

College of Cardiology Foundation

From the *Department of Mechanical Engineering, School of Biomedical Engineering and Sciences, Virginia Tech,Blacksburg, Virginia; and the †Cardiology Section, Wake Forest University School of Medicine, Winston-Salem, NorthCarolina. This work is partially supported by a National Science Foundation Graduate Research Fellowship Grant (0547434).Any opinions, findings, conclusions or recommendations expressed in this publication are those of the authors and do notnecessarily reflect the views of the National Science Foundation. Financial support has also been received from Wake ForestTranslational Science Institute. An application for a provisional patent covering the analysis program has been filed by VirginiaTech and Wake Forest Universities, which employ the authors. The authors have reported that they have no relationships todisclose.

Manuscript received July 23, 2010; revised manuscript received September 17, 2010, accepted September 21, 2010.

LctfTc

idaia(

pedditodavo

uwoflsn

at1

M

EctMdpailamtraTlaBteOLurirvga

taprstc1fEwriiCo

A

A

B

p

C

E

H

c

I

d

L

P

w

V

Vs � early filling strength

J A C C : C A R D I O V A S C U L A R I M A G I N G , V O L . 4 , N O . 1 , 2 0 1 1

J A N U A R Y 2 0 1 1 : 3 7 – 4 6

Stewart et al.

Stewart-Color M-Mode Diastolic Function

38

eft ventricular (LV) diastolic function can benoninvasively evaluated from LV filling dy-namics determined by Doppler echocardiogra-

phy (1–3). Color M-mode (CMM) echo-ardiography provides a spatiotemporal map ofhe velocities of the blood flow along the scan linerom the mitral annulus to the LV apex (4,5).he current method of analyzing this data is to

alculate the propagation velocity (Vp) of the

See page 47

nflow jet traveling toward the apex during earlyiastole (1–3,5–11). Vp is measured as the slope oflinear approximation of an isovelocity contour. Vp

s reduced in patients with diastolic dysfunction,nd the ratio of peak transmitral E-wave velocityE) to Vp is elevated when left atrial pressure is

increased (3,12–16).There are several potential limitations

to the use of Vp as a measure of LVdiastolic function. First, in many situa-tions, the isovelocity contour may not beaccurately described by a straight line (17–19). Thus, assignment of a single slope(i.e., Vp) may not be accurate. Further-more, Vp is subject to variation based onhow the isovelocity contour is determined(1,17,20). Finally, Vp has been found tobe normal in patients with hypertrophiccardiomyopathy (HCM) who have dia-stolic dysfunction apparent by other meth-ods (21).

Under normal circumstances, early di-astolic filling results from a progressive

ressure gradient from the left atrium (3,22) thatxtends toward the LV apex. With diastolicysfunction, the magnitude of the pressure gra-ient is reduced, and it does not extend as deeplynto the left ventricle (9). Thus, we hypothesizedhat: 1) with diastolic dysfunction, the initial Vpf the filling wave is reduced; and 2) it wouldecelerate to a lower velocity prior to reaching thepex. In contrast, in normal subjects, the initialelocity would be higher and deceleration wouldccur closer to the LV apex.Accordingly, we evaluated early diastolic filling

sing CMM echocardiograms from 125 patientsith a range of diastolic function. Consistent withur hypothesis, we found that the early diastolicow velocities were not accurately described by aingle slope. We used this information to develop a

sure

ew metric of early diastolic filling that more c

ccurately recognized diastolic dysfunction. Wehen tested this new measure in a second cohort of60 patients.

E T H O D S

cho Doppler. Echo Doppler examinations wereompleted using an iE33 ultrasound imaging sys-em with a multiple frequency transducer (Philips

edical Systems, Andover, Massachusetts). Stan-ard 2-dimensional images were obtained in thearasternal long and short axes and in the apical 4-nd 2-chamber views. Pulsed-wave Doppler trac-ngs of mitral valve inflow were recorded at theeaflet tips. A CMM ultrasound was obtained in thepical long-axis view with a sweep speed of 100m/s with a scale that optimized visualization of

he isovelocity color contour as judged by theecording sonographer. Recordings of the septalnd lateral mitral annular velocities were averaged.he LV volumes and Doppler tracings were ana-

yzed using a digital echocardiography workstations previously described (23,24).rief description of the automated algorithm. An au-omated data analysis algorithm was developed toxamine the CMM echocardiograph images.riginal CMM images were analyzed in MAT-AB (The Mathworks, Natick, Massachusetts)sing in-house developed image processing algo-ithms. The algorithm is used to crop a region ofnterest (Fig. 1B) and the velocity color scaleegion. With this information, a point-by-pointelocity reconstruction is completed on the re-ion of interest (see the Online Appendix fordditional information).

The image was reconstructed using a dealiasingechnique similar to that used by Thomas et al. (25)nd Rovner et al. (26) (Fig. 1D). Using image-rocessing tools, the E-wave velocity field waseduced to a series of 27 isovelocity contours evenlypaced between 45% and 55% of the peak E-waveransmitral velocity shown in Figure 1E. The re-onstructed velocity contours are shown in FigureF with the 45% to 55% isovelocity contours shownrom light to dark.nsemble contour methodology. A smoothing splineas fit to the series of isovelocity contours and is

eferred to as the ensemble contour in the remain-ng analysis (see the Online Appendix for additionalnformation).hange-point methodology. Previous observationsf a change in slope or curvilinear isovelocity

B B R E V I A T I O N S

N D A C R O N YM S

NP � B-type natriuretic

eptide

MM � color M-mode

� E-wave inflow velocity

CM � hypertrophic

ardiomyopathy

VPD � intraventricular pres

ifference

V � left ventricular

CW � pulmonary capillary

edge

p � propagation velocity

ontour (17–19) are consistent with an abrupt

dstapttl

cmttchsePpwdtf

srpl

ittRbmoanhEfialp


J A N U A R Y 2 0 1 1 : 3 7 – 4 6

Stewart et al.


39

eceleration of the LV filling wave. We used atatistical change-point analysis method (27,28) onhe derivative of the ensemble contour (Figs. 2Dnd 2I) to objectively determine the decelerationoint. The method is based on a cumulative sum ofhe difference between the value of interest (xi) andhe mean value (x�). Equation 1 displays the cumu-ative sum equation:

Cumulative Sumi � Cumulative Sumi�1 � (xi � x�)

The waveform produced by the output of theumulative sum equation was plotted to deter-ine the significance of the change throughout

he signal (Figs. 2E and 2J). The peaks withinhis cumulative sum waveform were sorted ac-ording to their magnitude. The peak with theighest magnitude signifies the most statisticallyignificant change and was labeled as the decel-ration point.ressure calculations. The 1-dimensional incom-ressible Euler equation, shown in Equation 2,here p is the pressure, � is constant bloodensity, and � is velocity, was used to calculatehe relative pressures within the region of interest

Figure 1. Color M-Mode Echocardiography Analysis Method Ove

Color M-mode echocardiogram analysis of a restrictive filling patienafter a point-by-point velocity reconstruction. (D) Antialiased ROI. (Epeak transmitral velocity. (F) Isovelocity contours shown from light(G) Initial (pink) and terminal (green) propagation velocities.

rom the reconstructed velocity field. The pres- t

ure at each point along a scan line was calculatedelative to the position of the mitral annulus justrior to mitral valve opening by calculating the

ine integral between them (9,29,30).

�p

�s� ��v

�t� �

�v

�s �From the relative pressures, the peak diastolic

ntraventricular pressure difference (IVPD) fromhe left atrium to the LV apex was calculated similaro the calculations of Greenberg et al. (30) andovner et al. (26). This method has been validatedy comparison to direct measurements with micro-anometers (9,30). The IVPD provides a measure

f the strength of LV diastolic suction (9). We usedn IVPD of 2.2 mm Hg as the lower limit oformal based on the observations of the IVPD inealthy volunteers by Yotti et al. (9).arly filling velocity strength. We observed that earlylling is characterized by a point at which the Vpbruptly decelerates, indicating that a single straightine is not an accurate approximation for the Vparameter. The ensemble contour was divided at

w

) Original image. (B) Extracted region of interest (ROI). (C) ROII displaying isovelocity contours from 45% to 55% of the E-waveark. Arrows in E and F indicate the deceleration point location.

rvie

t. (A) ROto d

he deceleration point into 2 segments: the initial

prevaaj

tmps

PtccfptCpFt

oaar

ocfig�cdp(dawov

fiid

n p


J A N U A R Y 2 0 1 1 : 3 7 – 4 6

Stewart et al.


40

ropagation region and the terminal propagationegion. A line was fit to the ensemble contour forach filling region to calculate the propagationelocities before (initial Vp) and after the deceler-tion point (terminal Vp), as shown in Figure 2Cnd 2H as the 2 intersecting lines where theunction is the deceleration point.

We calculated the distance from the position ofhe mitral annulus just prior to the opening of theitral valve to the deceleration point, Li. The

roduct of this distance and the initial Vp repre-ents the strength of early filling (Vs).

Vs � InitialVp · Li

atient population. Two independent groups of pa-ients were used in this study: a development cohortonsisting of 125 patients and a validation cohortonsisting of 160 patients. These were selectedrom patients undergoing clinically indicated com-rehensive echocardiography and Doppler evalua-ion at the Wake Forest University Baptist Medicalenter. The study was conducted according torotocols approved by the Virginia Tech and Wakeorest University Baptist Medical Center institu-

Figure 2. Propagation Velocities and Deceleration Point Locatio

(A, F) Color M-mode echocardiograms of a normal (top) and restricplaying conventional propagation velocity. (C, H) Dealiased regioneration point located at their intersection. (D, I) Ensemble contourensemble contour slope. Vertical dotted line represents deceleratio

ional review boards. t

Diastolic dysfunction stages were assigned basedn the mitral valve inflow, tissue Doppler mitralnnular velocities, and conventional Vp measuredccording to the American Society of Echocardiog-aphy guidelines (3).

DEVELOPMENT COHORT. The patients in the devel-pment cohort were selected to create 5 equally sizedategories. Due to the wide range of ages, the healthylling category was divided into 2 categories: the firstroup was �30 years of age, and the second group was30 years of age. The remaining 3 categories were

omposed of 25 patients with each of the 3 stages ofiastolic dysfunction: delayed relaxation filling (stage 1);seudonormal filling (stage 2); and restrictive fillingstage 3) (Table 1). Seventeen patients from theevelopment cohort were randomly selected for thessessment of parameter reproducibility. The patientsere analyzed 3 independent times by 3 differentbservers to assess the interobserver and intraobserverariability.

VALIDATION COHORT. The patient populationor the validation cohort consisted of 160 patients,ncluding 10 patients with HCM (Table 2). It ismportant to note that development cohort CMMata included only good image quality scans and

filling (bottom) patient. (B, G) Dealiased region of interest dis-terest displaying initial and terminal propagation velocities, decel-ensemble contour slope. (E, J) Cumulative sum calculated fromoint time.

n

tiveof inand

hat CMM images with merged E and A waves

wsfwope

ehm2hmt

edP�mSSgT

tie

d


J A N U A R Y 2 0 1 1 : 3 7 – 4 6

Stewart et al.


41

ere not included. The validation cohort’s CMMelection criteria were less strict, and patients withair to poor image quality and merged E and Aaves were included in the analysis. Other analysisf observations in some of the validation patientopulation has been previously published by Bruckst al. (31).

Each of these patients had a clinically indicatedcho-Doppler examination. Fifty-six of these patientsad undergone cardiac catheterization with measure-ent of the pulmonary capillary wedge (PCW) withindays of the echocardiography; 117 of the patients

ad a serum B-type natriuretic peptide (BNP) deter-ination on the same day as the echo. In the valida-

ion cohort, we assessed the ability of CMM param-

Table 1. Clinical Characteristics of the Testing Cohort

Diastolic Dysfunction Stage Number of Patients Ag

0—Healthy, age �30 yrs 25 24.3

0—Healthy, age �30 yrs 25 47.3

1—Delayed relaxation 25 68.3

2—Pseudonormal 25 66.2

3—Restrictive filling 25 59.4

Patients are classified based on clinically diagnosed diastolic dysfunction stage.transmitral to E-wave mitral annulus velocity ratio.

Table 2. Clinical Characteristics of the Validation Cohort

Subset With IV

n Age, yrs Sex, n (%)

160 57 � 13 Male 92 (58)

Female 68 (42)

Subset With B

n Age, yrs Sex, n (%)

117 58 � 14 Male 61 (52)

Female 56 (48)

Subset With PCW P

n Age, yrs Sex, n (%)Period Betweeand Catheteriz

56 56 � 19 Male 38 (68) 20 � 15

Female 18 (32)

BNP � B-type natriuretic peptide; EF � ejection fraction; IVPD � intraventricular pr

ters to recognize diastolic dysfunction defined in 3ifferent ways. These included: 1) elevated meanCW pressure �18 mm Hg; 2) elevated serum BNP100 pg/ml (31–33); and 3) reduced IVPD �2.2m Hg, indicating reduced LV suction (9).tatistical analysis. Data are expressed as mean �D. We analyzed statistical significance amongroups using a 1-way analysis of variance and theukey-Kramer honest significant difference test.The squared correlation coefficient, R2, was used

o compare a single linear fit and the combinednitial and terminal fit of the leading edge of thearly filling wave (Table 3).

The ability to distinguish patients with diastolicysfunction from normal filling was analyzed by

rs) E/A* E/E=† Ejection Fraction

4.25 1.98 � 0.69 6.53 � 1.47 0.57 � 0.05

14.33 1.54 � 0.31 7.88 � 2.66 0.62 � 0.05

9.62 0.79 � 0.11 13.00 � 4.22 0.57 � 0.13

12.94 1.58 � 0.31 16.40 � 5.27 0.40 � 0.15

18.50 2.91 � 1.02 18.77 � 7.38 0.28 � 0.08

es represent mean � SD. *E-wave to A-wave transmitral velocity ratio. †E-wave

Measurement

EF <0.4, n (%) Diastolic function, n (%)

29 (18) Normal 52 (33)

Impaired relaxation 51 (32)

Pseudonormal filling 26 (16)

Restricted filling 24 (15)

Not determined 7 (4)

Measurement

EF <0.4, n (%) Diastolic function, n (%)

25 (21) Normal 40 (34)




Not determined 1 (0.9)

ure Measurement

chon, h EF <0.4, n (%) Diastolic Function, n (%)

14 (25) Normal 20 (36)




Not determined 6 (11)

e (y

6 �

6 �

2 �

0 �

4 �

Valu

PD

NP

ress

n Eatio

essure difference; PCW � pulmonary capillary wedge.

rnccaCSCs

R

Actsetpaeharcr

tpdtsiBdspi(Va

cdc

9Ao


J A N U A R Y 2 0 1 1 : 3 7 – 4 6

Stewart et al.


42

eceiver-operator characteristic curves. The diag-ostic ability of the parameters was assessed byomparing the area under the receiver-operatorharacteristic curves with the method of DeLong etl. (34) using MedCalc Statistical Software (Med-alc Software bvba, Mariakerke, Belgium). JMPtatistical Discovery Software (SAS Institute Inc.,ary, North Carolina), was used for the other

tatistical analyses.

E S U L T S

nalysis of the development cohort. The isovelocityontours were not linear as the 2-stage initial anderminal fit provided considerably better fits than aingle linear approximation (Table 3). This nonlin-arity was also present by visual inspection. Becausehe slope abruptly decreased at the discontinuityoint, we termed this the deceleration point. Usingchange-point analysis, the location of the decel-

ration point occurred furthermost into the LV forealthy patients and patients with delayed relax-tion filling (3.1 � 0.7 cm and 3.3 � 0.8 cm,espectively) and progressively decreased with in-reased diastolic dysfunction (2.4 � 0.6 cm forestrictive filling), shown in Figure 3A.

The inflow wave was split into an initial anderminal section before and after the decelerationoint. The initial Vp progressively decreased withiastolic dysfunction shown in Figure 3B. Theerminal Vp was lower than the initial Vp and wasimilar in normal subjects and patients with increas-ng diastolic dysfunction shown in Figure 3C.ecause both the initial Vp and the distance to theeceleration point decreased with increasing dia-tolic dysfunction, we calculated their product as aotential new diastolic parameter. This productndicates the strength of the initial flow propagationVs) and represents the magnitude of LV suction.s was progressively decreased with increasing di-

Table 3. Coefficient of Determination of Linear and 2-StageInitial and Terminal Contour Fits

DiastolicDysfunction Stage

Linear FitR2 Value

Initial and TerminalFit R2 Value

pValue

Healthy, age �30 yrs 0.77 0.84 0.0098

Healthy, age �30 yrs 0.77 0.87 0.0328

Delayed relaxation 0.76 0.84 0.0461

Pseudonormal 0.82 0.89 0.0716

Restrictive filling 0.78 0.84 0.1065

stolic dysfunction (Fig. 4B). In the development

ohort, Vs was better at distinguishing patients withiastolic dysfunction from normal subjects thanonventional Vp (Fig. 5).

The variability of repeated measures of Vs was.9% for intraobserver and 13.8% for interobserver.dditional analysis of the intraobserver and inter-bserver variability values of Vs parameter, deceler-

Figure 3. DP Location and Initial and TerminalPropagation Velocities

The distance into the left ventricle where the velocity decelera-tion point (DP) occurs is progressively decreased with moresevere diastolic dysfunction, as is the initial propagation velocity.In contrast, the terminal propagation velocity is similar in all
groups.

avofvAcddI(o5I

w5W��w�

D

CttsUCappssdttavsf

ifwVafrldfitsl


J A N U A R Y 2 0 1 1 : 3 7 – 4 6

Stewart et al.


43

tion point location, and initial Vp and terminal Vpalues are in the Online Appendix. The variabilityf all parameters was calculated from a single beator each patient and does not include beat-to-beatariability.nalysis of the validation cohort. In the validationohort, Vs was consistently better at recognizingiastolic dysfunction than conventional Vp whetheriastolic dysfunction was defined as a reducedVPD, elevated PCW pressure, or elevated BNPFig. 6). The pairwise agreement of the 3 measuresf diastolic dysfunction was: IVPD versus BNP �7%; PCW versus BNP � 55%; PCW versusVPD � 65%.

We separately analyzed 10 patients with HCMho had abnormal diastolic filling patterns and0 normal subjects with normal filling pattern.

e found that Vs (area under the curve [AUC]0.76) was superior to conventional Vp (AUC0.58, p � 0.022) in detecting the 7 patients

ith HCM who had diastolic dysfunction (IVPD

Figure 4. Conventional Vp and Vs

The conventional flow propagation velocity (Vp) is reduced fromnormal �30 years in patients with restricted filling patterns. Theearly filling strength (Vs) progressively declines with increasingdiastolic dysfunction.

2.2 mm Hg).

I S C U S S I O N

olor M-mode echocardiography provides a spa-iotemporal map of early diastolic filling. Conven-ionally, this has been evaluated by calculating thelope of an isovelocity contour representing the Vp.sing an objective quantitative analysis of theMM data, we found that a single slope does not

ccurately represent the Vp for normal subjects oratients with diastolic dysfunction. Instead, the flowropagation is characterized by a more rapid initiallope indicating rapid flow propagation that abruptlylows after a deceleration point. In the presence ofiastolic dysfunction, the initial slope is reduced andhe deceleration point moves progressively closer tohe mitral annulus. The product of the initial slopend the distance to the deceleration point (Vs) pro-ides a measure of the strength of early diastolic LVuction and may provide a better measure of diastolicunction than the conventional Vp.

Under normal circumstances, early diastolic fill-ng results from a progressive pressure gradientrom the left atrium (4,22) that extends most of theay to the LV apex. This results in a rapid initialp that extends 3.1 � 0.7 cm from the mitral

nnulus toward the LV apex. With diastolic dys-unction, the magnitude of the pressure gradient iseduced, and it does not extend as deeply into theeft ventricle (9). Thus, we observed with diastolicysfunction that the initial Vp is reduced and thatlling wave decelerates to a lower velocity closer tohe mitral annulus. After termination of the pres-ure gradient, the terminal Vp is reduced to similarevels in all subjects regardless of diastolic function.

Figure 5. Development Cohort ROC Curves

Receiver-operator characteristic (ROC) curves displaying theability of Vp and Vs to discriminate normal subjects fromsubjects with abnormal diastolic filling patterns. Vs performs
better than Vp.

TiVf

tmbPpAnfims

bdaiapr

vdaHcf


J A N U A R Y 2 0 1 1 : 3 7 – 4 6

Stewart et al.


44

his terminal Vp may represent slower inertial flown the absence of a pressure gradient. The terminalp may also be reduced due to deviation of flow

rom the M-mode scan line.Propagation velocity is conventionally calculated as

he linear slope of an isovelocity contour from theitral annulus to 4 cm into the LV as recommended

y the American Society of Echocardiography (3).atients with normal filling display a decelerationoint that occurs 3.1 � 0.7 cm from the annulus.ccordingly, the initial Vp and conventional Vp forormal filling are similar. In contrast, in restrictivelling, the deceleration point occurs closer to theitral annulus (2.4 � 0.6 cm). Thus, in this circum-

1-Specificity

Reduced Intraventricular Pressure Difference <2.2 mmHg

Vs vs. E/E’Vs vs. Vp

P=0.021P=0.008

Vs (AUC=0.76)

E/E’ (AUC=0.58)

Vp (AUC=0.64)

0.0

0.2

0.4

0.6

0.8

1.0

0.0 0.2 0.4 0.6 0.8 1.0

Vs=199 cm2/sS

ensi

tivi

ty

Elevated BNP >100 pg/ml

0.0

0.2

0.4

0.6

0.8

1.0

0.0 0.2 0.

Sen

siti

vity

A

C

Figure 6. Validation Cohort ROC Curves

ROC curves displaying the ability of color M-mode parameters to detesure gradient �2.2 mm Hg, (B) an elevated pulmonary capillary wedge�100 pg/ml. Regardless of how diastolic dysfunction is defined, Vs wavelocity; other abbreviations as in Figures 4 and 5.

tance, the conventional Vp will be determined by w

oth the initial Vp and the terminal Vp, furtherecreasing the conventional Vp below the initial Vplone. Clinically, Vp may be measured using only thenitial linear portion. In patients with impaired fillingnd a short distance to the deceleration point, this willroduce a higher value for Vp and potentially fail toecognize diastolic dysfunction.

The conventional Vp is frequently normal in se-erely hypertrophied ventricles that have clear evi-ence of diastolic dysfunction (3). Accordingly, wessessed the new Vs parameter in 10 patients withCM and found that its performance was superior to

onventional Vp in accurately detecting diastolic dys-unction in HCM. These results should be interpreted

1-Specificity

Elevated Pulmonary Capillary Wedge Pressure >18 mmHg


P=0.719P=0.004

Vs (AUC=0.80)

E/E’ (AUC=0.77)

Vp (AUC=0.62)

0.0

0.2

0.4

0.6

0.8

1.0

0.0 0.2 0.4 0.6 0.8 1.0

Vs=155 cm2/s

Sen

siti

vity

pecificity


P<0.001P<0.001

Vs (AUC=0.70)

E/E’ (AUC=0.75)

Vp (AUC=0.44)

0.6 0.8 1.0

astolic dysfunction recognized as: (A) a reduced intraventricular pres-ssure �18 mm Hg, and (C) an elevated B-type natriuretic peptideperior to Vp. BNP � B-type natriuretic peptide; E � E-wave inflow

1-S

4

B

ct dipre

s su

ith caution because of the small sample size.

Sdmcmmnaae(ombmirtngmtmtpDitsnc

ptdbradi

pitdr

C

Tdvsitefisf

RtM

R


J A N U A R Y 2 0 1 1 : 3 7 – 4 6

Stewart et al.


45

tudy limitations. We analyzed the degree of diastolicysfunction in the development cohort based on theitral inflow and tissue Doppler. In the validation

ohort, we used BNP, IVPD, and invasively deter-ined PCW pressure to provide 3 independenteans of recognizing diastolic dysfunction. However,

one of these standards is a perfect method of evalu-ting LV diastolic dysfunction. Increased BNP valuesre correlated with diastolic dysfunction (32); how-ver, BNP values can be influenced by other factors35). The IVPD provides another objective measuref the degree of diastolic dysfunction. We used aethod to calculate the IVPD that has been validated

y comparison to micromanometer pressure measure-ents (9,30). However, the values are not completely

ndependent of the echocardiography analysis becauseelative pressures were calculated from CMM usinghe Euler relationship. The PCW measurements wereot performed simultaneously with the echocardio-rams, thus diminishing the accuracy of the measure-ent in assessing diastolic dysfunction at the time of

he echo-Doppler examination. The limited agree-ent of the 3 methods of evaluating diastolic dysfunc-

ion may indicate that diastolic function is a complexrocess that cannot be evaluated by a single parameter.espite the potential limitations of the 3 methods of

ndependently defining diastolic function, we foundhat Vs consistently performed better than Vp. Thisuggests that Vs provides a superior method of recog-izing diastolic dysfunction from CMM echo-

relaxation: a color M-mode Dopplercolor M-mode Dopraphy in first myo

We used an algorithm to analyze a single beat ofreviously acquired clinical studies. It is possiblehat the algorithm could be implemented online asata are being acquired and applied to multipleeats. This has the potential to improve its accu-acy. Although we used a quantitative analysislgorithm, the initial Vp and the distance to theeceleration point can be recognized by visualnspection (Fig. 2).

In the development cohort, we included onlyatients with high-quality CMM images. However,n the evaluation cohort, we did not exclude pa-ients based on the quality of the images. Thisemonstrates the robustness of the analysis algo-ithm and the potential utility of Vs.

O N C L U S I O N S

he propagation of flow into the LV in early diastoleoes not have a single velocity. The initial rapid flowelocity suddenly slows at a deceleration point. Dia-tolic dysfunction is characterized by a reduction of thenitial Vp and the deceleration point occurring closero the mitral annulus. The product of these 2 param-ters, Vs, which reflects the strength of early diastoliclling, provides a more accurate assessment of dia-tolic function than conventional measurement of Vprom CMM echocardiography.

eprint requests and correspondence: Dr. William C. Lit-le, Cardiology Section, Wake Forest University School of

edicine, Medical Center Boulevard, Winston-Salem,
ardiography. North Carolina 27157-1045. E-mail: [email protected].
1

1

1

E F E R E N C E S

1. Thomas JD, Popovic ZB. Assessmentof left ventricular function by cardiacultrasound. J Am Coll Cardiol 2006;48:2012–25.

2. Oh JK, Hatle L, Tajik AJ, Little WC.Diastolic heart failure can be diagnosedby comprehensive two-dimensional andDoppler echocardiography. J Am CollCardiol 2006;47:500–6.

3. Nagueh SF, Appleton CP, GillebertTC, et al. Recommendations for theevaluation of left ventricular dia-stolic function by echocardiography.Eur J Echocardiogr 2009;10:165–93.

4. Little WC. Diastolic dysfunction be-yond distensibility: adverse effects ofventricular dilatation. Circulation2005;112:2888–90.

5. Brun P, Tribouilloy C, Duval AM,et al. Left ventricular flow propagationduring early filling is related to wall

analysis. J Am Coll Cardiol 1992;20:420–32.

6. Stugaard M, Greenberg, NL, ZhouJH, Thomas JD. Automated eigen-vector analysis for quantification ofcolor M-mode Doppler filling pat-terns of the left ventricle in an ische-mic canine model. Paper presented at:24th Annual Computers in Cardiol-ogy Conference; September 7–10,1997; Lund, Sweden.

7. Takatsuji H, Mikami T, Urasawa K, etal. A new approach for evaluation of leftventricular diastolic function: spatial andtemporal analysis of left ventricular fill-ing flow propagation by color M-modeDoppler echocardiography. J Am CollCardiol 1996;27:365–71.

8. Moller JE, Sondergaard E, Seward JB,Appleton CP, Egstrup K. Ratio of leftventricular peak E-wave velocity toflow propagation velocity assessed by

pler echocardiog-cardial infarction:

prognostic and clinical implications.J Am Coll Cardiol 2000;35:363–70.

9. Yotti R, Bermejo J, Antoranz JC, etal. A noninvasive method for assessingimpaired diastolic suction in patientswith dilated cardiomyopathy. Circula-tion 2005;112:2921–9.

0. De Mey S, De Sutter J, VandervoortP, De Buyzere M, Verdonck P. As-sessment of LV diastolic filling usingcolor M-mode Doppler echocardi-ography: validation in a new hydrau-lic model. Biomech Model Mech-anobiol 2004;127–38.

1. Oh JK. Echocardiography in heartfailure: beyond diagnosis. Echocardi-ography 2007;8:4–14.

2. Chapman JN, Mayet J, Foale RA,Thom SA. Intraventricular dispersionof E wave velocity: an alternative mea-sure of left ventricular diastolic func-tion in hypertensive patients. J Hum
Hypertens 1999;13:867–9.
mailto:[email protected]

1

1

1

1

1

1

1

2

3

3

3

3

3

3

Key

‹

Fm


J A N U A R Y 2 0 1 1 : 3 7 – 4 6

Stewart et al.


46

3. Bella JN, Palmieri V, Roman MJ,et al. Mitral ratio of peak early tolate diastolic filling velocity as apredictor of mortality in middle-aged and elderly adults: the StrongHeart Study. Circulation 2002;105:1928 –33.

4. Galderisi M. Diastolic dysfunctionand diastolic heart failure: diagnostic,prognostic and therapeutic aspects.Cardiovasc Ultrasound 2005;3:9.

5. Garcia MJ, Ares MA, Asher C, Ro-driquez L, Vandervoort P, ThomasJD. An index of early left ventricularfilling that combined with pulsedDoppler peak E velocity may estimatecapillary wedge pressure. J Am CollCardiol 1997;29:448–54.

6. Claessens TE, De Sutter J, VanherckeD, Segers P, Verdonck PR. Newechocardiographic applications for as-sessing global left ventricular diastolicfunction. Ultrasound Med Biol 2007;33:823–41.

7. Quinones MA. Assessment of dia-stolic function. Prog Cardiovasc Dis2005;47:340–55.

8. Sessoms MW, Lisauskas J, Kovacs SJ.The left ventricular color M-modeDoppler flow propagation velocityV(p): in vivo comparison of alternativemethods including physiologic impli-cations. J Am Soc Echocardiogr 2002;15:339–48.

9. Asada-Kamiguchi J, Jones M, Green-berg NL, et al. Intraventricular pres-sure gradients in left ventricular aneu-rysms determined by color M-modeDoppler method: an animal study.J Am Soc Echocardiogr 2006;19:1112–8.

0. Seo Y, Ishimitsu T, Ishizu T, et al.Assessment of propagation velocity bycontrast echocardiography for stan-dardization of color Doppler propaga-tion velocity measurements. J Am Soc
Echocardiogr 2004;17:1266–74.
21. Barbier P, Grimaldi A, Alimento M,Berna G, Guazzi MD. Echocardio-graphic determinants of mitral earlyflow propagation velocity. Am J Car-diol 2002;90:613–9.

22. Little WC, Oh JK. Echocardio-graphic evaluation of diastolic func-tion can be used to guide clinical care.Circulation 2009;120:802–9.

23. Gandhi SK, Powers JC, Nomeir AM,et al. The pathogenesis of acute pul-monary edema associated with hyper-tension. N Engl J Med 2001;344:17–22.

24. Warner JG Jr., Metzger DC, KitzmanDW, Wesley DJ, Little WC. Losar-tan improves exercise tolerance in pa-tients with diastolic dysfunction and ahypertensive response to exercise.J Am Coll Cardiol 1999;33:1567–72.

25. Thomas JD, Greenberg NL, Vander-voort PM, Aghassi DS, Hunt BF.Digital analysis of transmitral colorDoppler M-mode data: a potentialnew approach to the noninvasive as-sessment of diastolic function. Con-ference on Computers in CardiologyProceedings. Durham, NC: October11–14, 1992:631–4.

26. Rovner A, Smith R, Greenberg NL,et al. Improvement in diastolic intra-ventricular pressure gradients in pa-tients with HOCM after ethanol sep-tal reduction. Am J Physiol Heart CircPhysiol 2003;285:H2492–9.

27. Hinkley DV. Inference about thechange-point from cumulative sumtests. Biometrika 1971;58:509–23.

28. Taylor WA. Change-Point Analysis: APowerful New Tool for DetectingChanges. 2000. Available at: http://w w w . v a r i a t i o n . c o m / c p a / t e c h /changepoint.html. Accessed July 16,2008.

29. Thomas JD, Popovic ZB. Intraven-tricular pressure differences: a newwindow into cardiac function. Circu-
lation 2005;112:1684–6. a
0. Greenberg NL, Vandervoort PM,Firstenberg MS, Garcia MJ, ThomasJD. Estimation of diastolic intraven-tricular pressure gradients by DopplerM-mode echocardiography. Am JPhysiol Heart Circ Physiol 2001;280:H2507–15.

1. Brucks S, Little WC, Chao T, et al.Contribution of left ventricular dia-stolic dysfunction to heart failure re-gardless of ejection fraction. Am JCardiol 2005;95:603–6.

2. Dokainish H, Zoghbi WA, LakkisNM, et al. Optimal noninvasive as-sessment of left ventricular fillingpressures: a comparison of tissueDoppler echocardiography and B-typenatriuretic peptide in patients withpulmonary artery catheters. Circula-tion 2004;109:2432–9.

3. Maeda K, Tsutamoto T, Wada A,Hisanaga T, Kinoshita M. Plasmabrain natriuretic peptide as a bio-chemical marker of high left ventric-ular end-diastolic pressure in patientswith symptomatic left ventricular dys-function. Am Heart J 1998;135:825–32.

4. DeLong ER, DeLong DM, Clarke-Pearson DL. Comparing the areasunder two or more correlated receiveroperating characteristic curves: a non-parametric approach. Biometrics1988;44:837–45.

5. Cowie MR, Mendez GF. BNP andcongestive heart failure. Curr ProbCardiol 2003;28:264–311.

ey Words: diastole ychocardiography y heart failure

imaging.

A P P E N D I X

or detailed methodology and analysis infor-ation, please see the online version of this
rticle.

evaluation of lv diastolic function from color m-mode ... · analysis of ﬁlling velocities...

Documents