Donate Help Contact The AHA Sign In Home
American Heart Association
Hypertension
Search: search_blue_button Advanced Search
« Previous Article | Table of Contents | Next Article »
Hypertension. 1997;30:788-795

This Article
Right arrow Abstract Freely available
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow Request Permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Kovács, S. J.
Right arrow Articles by Hall, A. F.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Kovács, S. J.
Right arrow Articles by Hall, A. F.

(Hypertension. 1997;30:788-795.)
© 1997 American Heart Association, Inc.

Articles

Can Transmitral Doppler E-Waves Differentiate Hypertensive Hearts From Normal?

Sándor J. Kovács; Jose Rosado; Abigail L. Manson McGuire; ; Andrew F. Hall

From the Cardiovascular Biophysics Laboratory, Cardiovascular Division, Barnes-Jewish Hospital at Washington University Medical Center, St Louis, Mo.

Correspondence to Dr Sándor J. Kovács, Cardiovascular Biophysics Laboratory, Cardiovascular Division, Barnes-Jewish Hospital at Washington University Medical Center, 216 South Kingshighway Blvd, St Louis, MO 63110. E-mail sjk{at}howdy.wustl.edu


*    Abstract
up arrowTop
 *Abstract
down arrowIntroduction
 down arrowMethods
 down arrowResults
 down arrowDiscussion
 down arrowReferences
 
Abstract Physiological models of transmitral flow predict E-wave contour alteration in response to variation of model parameters (stiffness, relaxation, mass) reflecting the physiology of hypertension. Accordingly, analysis of only the E-wave (rather than the E-to-A ratio) should be able to differentiate between hypertensive subjects and control subjects. Conventional versus model-based image processing methods have never been compared in their ability to differentiate E-waves of hypertensive subjects with respect to age-matched control subjects. Digitally acquired transmitral Doppler flow images were analyzed by an automated model-based image processing method. Model-derived indexes were compared with conventional E-wave indexes in 22 subjects: 11 with hypertension and echocardiographically verified ventricular hypertrophy and 11 age-matched nonhypertensive control subjects. Conventional E-wave indexes included peak E, {int}E, and acceleration and deceleration times. Model-based image processing–derived indexes included acceleration and deceleration times, potential energy index, and damping and kinematic constants. Intergroup comparison yielded lower probability values for model-based compared with conventional indexes. In the subjects studied, Doppler E-wave images analyzed by this automated method (which eliminates the need for hand-digitizing contours or the manual placement of cursors) demonstrate diastolic function alteration secondary to hypertension made discernible by model-based indexes. The method uses the entire E-wave contour, quantitatively differentiates between hypertensive subjects and control subjects, and has potential for automated noninvasive diastolic function evaluation in large patient populations, such as hypertension and other transmitral flow velocity–altering pathophysiological states.


Key Words: diastole • models, physiological • ultrasonics • hypertension, model-based image processing


*    Introduction
up arrowTop
 up arrowAbstract
 *Introduction
down arrowMethods
 down arrowResults
 down arrowDiscussion
 down arrowReferences
 
Abnormalities of left ventricular diastolic function have been described in patients with hypertension and LV hypertrophy.1 2 3 4 5 6 7 8 9 10 One manifestation of diastolic dysfunction in these patients is alteration of the Doppler-derived early peak transmitral flow to late transmitral flow velocity ratio (E/A). LV diastolic filling is affected by numerous diastolic function parameters such as ventricular mass, compliance, elastance, and relaxation, as well as variables such as volume status, age, heart rate, and systolic function.11 12 13 In addition, anatomic structures such as the pericardium, lung, and thorax are also known to affect diastolic filling and function.14 One limitation in basing the diagnosis of diastolic dysfunction on Doppler indexes such as the E/A ratio or deceleration time is the fact that these indexes are determined by the value of the flow velocity at only one or two single points rather than the entire E- and/or A-wave contours. Furthermore, E/A may become pseudonormalized11 12 over time in association with advancing dysfunction, which limits its reliability for distinguishing the presence of diastolic dysfunction. Under these conditions the pulmonary venous flow contour may provide additional information, but it may be impossible or often difficult and time-consuming to record.15 16 17 18 Moreover, essentially all Doppler indexes are computed manually (ie, the contour is determined `by eye' and is hand-traced using a trackball) or by manual placement of electronic cursors on Doppler images. The complex and incompletely understood dependence of the contour, and Doppler indexes computed from its shape, on parameters, variables, anatomic structures, and even settings of the echocardiography machine itself, further complicates matters.19

The PDF formalism is a lumped parameter mathematical model for the kinematics of filling which reproduces the E- and A-wave contours in their entirety, and has been shown to accurately reproduce in vivo transmitral flow velocities.20 The method has been developed in such a way that input data are in the form of digitally acquired audio or video21 Doppler flow images rather than hand-traced contours. The model-based image processing strategy solves the "inverse-problem" for diastole22 and determines the PDF parameters in an automated fashion for both the early rapid filling (E-wave)23 and the late atrial filling (A-wave) contours.24 The method of E-wave onset determination using an automated method has been previously described,25 and progress toward complete automation, including the effects of noise, and E- and A-wave parsing has been reported.26 The method's practical use rests in its ability to characterize the entire E- and A-wave contour in mathematical terms. Any pathophysiological state that may or may not alter the E-to-A ratio (determined by two points of the entire contour), but alters the shape of the Doppler contour relative to control, can be quantitatively differentiated using our method. Therefore, this method is not susceptible to the "E/A ratio problem" due to pseudonormalization. It can quantify contour differences relative to normal even in pseudonormalized states (where E/A was <1 and has become >1) because the model reproduces the entire E- and A-wave contour.

Consideration of the physiological determinants of the Doppler E-wave in concert with theoretical modeling of the effect that variation of these determinants has in altering the rapid filling process, and hence the entire E-wave contour,27 28 29 indicates that differences in the E-wave contours of hypertensive subjects versus control subjects should be discernible. In other words, if the peak E-wave amplitude in a subject with hypertension and one without hypertension were the same, we would still expect, based on physiological principles, that the contours would be different (ie, different shapes, curvatures, etc). This is because in modeling hypertensive physiology, values different from normal are required for model parameters (relaxation, compliance, resistance to inertia ratio, etc). Therefore, consideration of the shape of the entire E-wave should suffice to differentiate hypertensive from normal contours. In this study we sought to establish our methodology by investigating the ability of E-wave–derived model-based indexes versus that of conventional Doppler indexes, to differentiate (in terms of the probability value from an unpaired Student's t test) transmitral Doppler tracings of hypertensive subjects from those of age-matched control subjects.


*    Methods
up arrowTop
 up arrowAbstract
 up arrowIntroduction
 *Methods
down arrowResults
 down arrowDiscussion
 down arrowReferences
 
Patient Groups
Consent for participation in this study was obtained according to Barnes-Jewish Hospital/Washington University Medical Center Human Studies Committee criteria. Twenty-two subjects were evaluated: 11 hypertensive subjects and 11 age-matched normal control subjects. Control subjects had no clinical evidence of cardiac disease. Subjects, referred to the Barnes-Jewish Hospital echocardiography laboratory for evaluation, were selected from review of recently archived clinical echocardiography laboratory data. Clinical questionnaires detailed pertinent clinical history, medications, and epidemiological attributes. Subjects with technically suboptimal studies were not considered.

Echocardiographic Examination
The echocardiographic examination was performed with the patient supine. The subject's age, height, weight, and systolic and diastolic blood pressures were recorded. Doppler, two-dimensional, and M-mode echocardiograms were obtained in all study patients using a commercially available Hewlett-Packard Sonos 1500 or Acuson 128XP echocardiography machine. The two-dimensional directed M-mode echocardiograms of the left ventricle were obtained below the level of the midpapillary muscle. The LV mass index, average LV wall thickness, and indexes of systolic function were calculated according to American Society of Echocardiography standards.30

The left lateral position was used to obtain optimal Doppler image quality. The LV inflow tract was interrogated from the apical four-chamber view with the sample volume at the tips of the mitral leaflets. Doppler examinations were performed with a 2.5-MHz transducer with the baseline filter set at the lowest level, recorded on videotape in super VHS format. End-expiratory images, most representative of the transmitral Doppler signals, were used for analysis.

Doppler Image Processing
Analysis of frame-grabbed Doppler images was performed off-line in the Cardiovascular Biophysics Laboratory according to the schema in Fig 1Down. Representative transmitral flow profile images from each subject were used. Examples of transmitral Doppler images in hypertensive subjects and normal subjects, and the fit to the data using the model-based image processing method, are provided in Figs 2Down and 3Down. The MBIP method determined the PDF parameters that characterize the kinematics of the filling process.20 Three PDF parameters are required to characterize the E-wave in its entirety: xo, initial spring displacement (cm), equivalent to the velocity time integral of the E-wave; c, the damping constant (g/s); and k, the spring constant (g/s2). From these, c, the stored energy index 1/2 kxo2 (ergs), and the parameter ß=c2–4mk were chosen for comparison with conventional indexes. Peak E-wave flow velocities were determined using a previously described automated method.23 24 Acceleration and deceleration times were computed using two different methods: the MBIP method and the conventional method. The MBIP method determined the best fit to the E-wave contour and used the time interval from onset of the best fit to its peak as the acceleration time. The deceleration time using the MBIP method was determined similarly, ie, the time from peak to 10% of baseline. The conventional method of acceleration and deceleration time determination consisted of placing cursors, "by eye," at the onset, peak, and termination of the E-wave. These measurements were performed by three experienced echocardiographers, the measurements were averaged before comparison between hypertensive subjects and control subjects.



View larger version (30K):
[in this window]
[in a new window]
  		Figure 1. Flowchart of the MBIP method for transmitral Doppler flow-velocity analysis.



View larger version (83K):
[in this window]
[in a new window]
  		Figure 2. Pulsed Doppler recordings of normal (a) and hypertensive (b) subjects with MBIP contours superimposed. Note that E-wave amplitudes are comparable, but E-to-A ratios differ. The PDF parameters: for panel a are xo=11.34 cm, k=262.54 g/s2, c=19.45 g/s, Fo=4066 dyne, {omega}=17.48 rad/s; for panel b xo=14.61 cm, k=376.6 g/s2, c=52.77 gm/s, Fo=4424 dyne, {omega}=18.08 rad/s. See text for details.



View larger version (42K):
[in this window]
[in a new window]
  		Figure 3. Pulsed Doppler recordings of normal (a) and hypertensive (b) subjects with MBIP contours superimposed. Note similarity of E-wave amplitudes and overall features with E-to-A ratio <1. The PDF parameters were for panel a xo=9.96 cm, k=139.82 g/s2, c=17.98 g/s, Fo=3381 dyne, {omega}=16.92 rad/s; for panel b xo=11.02 cm, k=248.9 g/s2, c=34.30 g/s, Fo=4251 dyne, {omega}=17.45 rad/s. See text for details.

In an effort to determine the effect of beat averaging, we selected 10 consecutive beats from a subject's data set. The maximum velocity envelope (see Fig 1Up) was determined, and PDF parameters were computed as previously described.23 24 25 26 Due to respiratory variation, both the amplitude and the duration of the E-wave, and primarily the amplitude of the A-wave, were noted to vary. To illustrate the magnitude of the variation, we superimposed the images, using the peak of the E-wave as the fiducial point (see Fig 4aDown). This demonstrates the physiological variability of the overall features of E-waves in this subject. For completeness, we superimposed the same 10 images using the peak of the A-wave as the fiducial point (see Fig 4bDown), demonstrating the beat-to-beat variability associated with A-waves. It shows that A-wave variability is most significant in the alteration of amplitude, and not duration, which is to be expected in light of the stable nature of the electrocardiographic PR interval.



View larger version (32K):
[in this window]
[in a new window]
  		Figure 4. Ensemble of 10 consecutive beats from one subject are superimposed to illustrate the magnitude of beat-to-beat variability. A, Use of the E-wave peak as a fiducial marker illustrates the physiological (respiratory) variability among beats. Note slight variation in amplitude and more marked variation in duration. B, Use of the A-wave peak as the fiducial marker illustrates beat-to-beat variability. Note minimal variation in duration and slight variation in amplitude. See text for details.

These data illustrate the magnitude and nature of the variability of E- and A-wave contours. Furthermore, for this ensemble of beats, the standard deviation of each of the model parameters (as a percentage of the mean) was smaller than the corresponding standard deviation between the patient groups for those parameters by a factor of between 3 and 8. Therefore, using single-beat rather than beat-averaged data has no discernible effect on the comparison statistics between the groups studied.

We have also performed a flow-phantom study to assure that our results are not echo machine dependent. That is, we have shown that when a flow-phantom–generated velocity profile is imaged by our Hewlett-Packard or Acuson machine, including maximal variation of machine-dependent controls such as baseline filter, gain, and reject controls, the slight PDF parameter machine-dependence is less than the difference between clinical subsets.19 The assurance of machine independence in analysis of transmitral flow data is unique to our method, and has never been considered in any prior study of Doppler-based indexes.

Statistical Analysis
Statistical comparisons were performed using an unpaired Student's t test in the Statview program on a Macintosh Quadra 950 computer. A value of P<.05 was considered significant. All values are reported as mean±SD.


*    Results
up arrowTop
 up arrowAbstract
 up arrowIntroduction
 up arrowMethods
 *Results
down arrowDiscussion
 down arrowReferences
 
Patient Characteristics
All subjects were in normal sinus rhythm. Systolic and diastolic blood pressures were both slightly higher in the hypertensive group, as was the mean age and heart rate, but only the diastolic blood pressure appeared significant, P=.0361 (see Table 1Down). All study patients in the hypertensive group had a documented history of systolic and diastolic hypertension treated medically, whereas none of the control subjects had been treated with antihypertensive drugs. None of the subjects in this study had a prior clinical history of angina, myocardial infarction, or valvulopathy.


View this table:
[in this window]
[in a new window]
  		Table 1. Clinical Attributes

Echocardiographic Measurements
The average posterior LV wall thickness and the LV mass index were significantly greater in the hypertensive group than in the normotensive group (see Table 2Down).


View this table:
[in this window]
[in a new window]
  		Table 2. Doppler Echocardiographic Indexes

There was no significant difference between groups in peak velocity of early ventricular filling (E-wave magnitude), in E-wave integral, in E-wave acceleration or deceleration times determined by the conventional method, or in E-wave deceleration times determined by the MBIP method. E-wave acceleration time, determined using the MBIP method, was longer in normal subjects than in hypertensive subjects, P=.0032 (see Table 2Up).

Model-Based Image Processing
The choice of MBIP parameters used for analysis was based on their relationship to conventional indexes (acceleration and deceleration times) or the parameters' physical meaning and ability to characterize the kinematics of the system20 (see Table 3Down). The E-wave parameters selected for comparison are acceleration and deceleration times; 1/2 kxo2 (ergs), the potential energy stored in the spring before release; c (g/s), the damping constant; and the index ß=c2–4mk (g/s)2. This last index is a measure of system kinematics in the sense of underdamped (ß <0) or overdamped (ß >0) motion of the oscillator. Underdamped kinematics correspond to spring (ie, suction/restoring) forces dominant over damping (ie, delayed relaxation), whereas overdamped kinematics indicates a dominance of damping relative to restoring forces.


View this table:
[in this window]
[in a new window]
  		Table 3. MBIP Indexes


*    Discussion
up arrowTop
 up arrowAbstract
 up arrowIntroduction
 up arrowMethods
 up arrowResults
 *Discussion
down arrowReferences
 
The clinical diagnosis of isolated LV diastolic dysfunction is based on the presence of elevated LV end-diastolic pressure, resulting in pulmonary congestion in the presence of preserved LV systolic function. Clinically, the most striking examples involve patients with acute pulmonary edema who have normal LV ejection fraction. Historically, cardiac catheterization has been the conventional method for assessing diastolic LV chamber properties.31 32 The development of Doppler echocardiography has permitted noninvasive assessment of diastolic transmitral flow patterns and development of strategies for the analysis of diastolic LV function.33 The PDF formalism was developed for the evaluation of diastolic function using Doppler echocardiography.20 The kinematic lumped parameter approach was selected because it lumps all anatomic/physiological components into inertia, restoring forces, and damping terms, and therefore need not consider anatomic,14 27 28 29 34 35 physiological,36 37 38 and epidemiological39 40 details. The formalism and the MBIP method can not only provide kinematic indexes but can also be used to automatically, rather than manually, compute the traditional Doppler measures (E/A, {int}E, {int}A, etc). In this study we compared MBIP-derived parameters with conventional Doppler indexes in their ability to differentiate transmitral flow contours in hypertensive subjects versus age-matched control subjects.

Mathematical Models of Transmitral Flow
The PDF formalism quantifies lumped aspects of cardiovascular motion (kinematics) during the filling process and characterizes transmitral flow velocities in terms of the dynamics of a damped harmonic oscillator.20 According to the model, three parameters (xo, c, and k) suffice to characterize E-wave contours in their entirety. Therefore, E-wave contours can be uniquely characterized in mathematical terms using three easily understood attributes of harmonic oscillator motion: amplitude, frequency (ie, width), and rate of decay. The formalism predicts the clinical transmitral flow contour accurately, and its parameter determination method has been automated.21 22 23 24 25 As an added assurance, we have performed a flow-phantom study to characterize the effect of echo-machine setting and machine type on the PDF parameters and have found machine-dependence to be insignificant. Hence, indexes constructed from the PDF parameters allow use of various manufacturers' machines and facilitate the computation of physiologically interesting variables within the machine-dependent limits observed.19

In contrast to the 36 or more conventionally derived Doppler indexes of diastolic function that have been proposed,41 the PDF model and its parameters provide an efficient quantitative method by which the entire Doppler E- and A-wave contour can be numerically expressed. This method not only provides numerically unique values for the PDF parameters but also generates standard errors for each parameter that reflect the noise content of the signal and the sensitivity of the contour to changes in the PDF parameter under consideration.21

The work of Meisner27 indicates that when physiological model parameters are altered to reflect changes associated with hypertension (ie, slowed relaxation, stiffer ventricle, etc), the atrioventricular gradient and hence the E-wave contour must change with respect to control, even if we choose parameters in such a way as to keep the peak of the E-wave constant. One obtains similar results in considering the effect of parameter variation in physiological models of transmitral flow due to Thomas et al28 and Sun et al.29

An appreciation for the complexity of physiological models is provided by the recent work of Sun et al.29 Their model used an electric analog model of the circulatory system with 10 simultaneous (nonlinear) differential equations. Without including effects related to the pericardium, thorax, lung, and other anatomic variables, more than 40 parameters related to properties of the circulation, atria, valves, and ventricles were required. As in comparable modeling approaches,27 28 the large number of model parameters eliminated the possibility of unique parameter determination by fitting model-predicted flow to the clinical Doppler contour.

The general limitation of the nonlinear models, due to their incorporation of a large number of parameters, is that they are not invertible in the mathematical sense. In other words, only if one specifies all of a model's parameters a priori can one obtain a model-generated transmitral flow velocity profile, ie, an E-wave. The opposite, that is to say, the solution of the inverse problem, is not possible. Using numerical methods we have compared physiological models from studies by Meisner27 and Thomas et al28 and have shown that one cannot use a clinically recorded Doppler E-wave contour as the sole input to these models and uniquely determine the model's parameters.42

One advantage of the MBIP strategy is that it provides unique model parameters directly from the clinical Doppler data; the cost however is that the kinematic parameters reflect the effects of a lumped set of physiological parameters that determine, for example, the E-wave contour's amplitude. Because the amplitude (versus the width or rate of decay) of the E-wave is dependent on a host of physiological parameters in a complex manner,43 we do not expect a single PDF parameter such as c or k to correspond to a single physiological variable.

Certain indexes derived or constructed from PDF parameters have physiologically intuitive kinematic analogues. The parameters' kinematic role, their relationship to the Doppler contour, and indexes derived from them have been previously discussed.20 21 22 23 24 Examples include the following: the initial displacement xo (cm) of the oscillator spring, the equivalent of the E-wave time-velocity integral (cm); the energy (ergs) stored in the oscillator before release (1/2 kxo2), a measure of stored elastic strain to power rapid filling; the maximum force (dynes) in the spring (kxo), which, per unit area, is the analogue of the maximum atrioventricular pressure gradient during rapid filling. Other purely kinematic parameters include the quantity ß=c2–4 mk. The quantity ß<0, ß=0, or ß>0 corresponds to underdamped, critically damped, or overdamped motion for the oscillator. Underdamped kinematics characterizes E-waves observed in normal hearts, where recoil forces dominate over damping; critically damped motion corresponds to velocity profiles seen in older and generally healthy hearts, a balance between recoil forces and damping; and overdamped kinematics characterize Doppler contours in association with abnormalities of relaxation, where damping forces dominate over recoil forces. Severe overdamped kinematics, where a high E-wave peak decays slowly, is seen in mitral stenosis.44

A result of particular interest concerns indexes derived from the E-wave alone. It is reassuring that E-wave acceleration is different between the groups as determined by the MBIP method because a discernible shorter acceleration time in hypertensive subjects compared with control subjects is consistent with the physiological interpretation that hypertensive hearts initiate their filling faster and are stiffer (ie, the contours have larger values of k and c). Among conventional indexes computed numerically, only the acceleration time attained significance between groups (P=.0032). For other MBIP-derived indexes, 1/2 kxo2, c, and c2–4mk achieved a significance of P=.0015, P=.0002, and P=.0393, respectively. The ability of model-based indexes to differentiate among groups using the E-wave as the only input is explained by the requirement that the method fit the entire contour rather than only a single point (E-wave peak) or an interval determined by two points (deceleration time).

Comparison to Other Clinical Studies
The group selected for study, comprised of older patients with echocardiographic LV hypertrophy and hypertension, represent a well-defined clinical set of patients who are known to have diastolic dysfunction.1 2 3 4 5 6 7 8 9 10 45 46 47

Our overall results are in concert with the observations of Pearson et al45 who carried out echocardiographic evaluation of elderly subjects with isolated systolic hypertension as part of the Systolic Hypertension in the Elderly Program. The mean ages for their subjects (71 years) and the subjects in our study (71 years) were the same, as was the observation of significant differences in A-wave amplitude between hypertensive subjects and control subjects (P=.02 in their study). Similarly, we observed no differences in echocardiographic indexes of ventricular systolic performance. Because a comparison of E-wave amplitude, acceleration time, or deceleration time was not included in their analysis, a direct comparison to this study is not possible; however, E/A was found to be lower in the hypertensive group compared with normal subjects in their study, P=.0124. They concluded that elderly subjects with well-preserved systolic function and a high incidence of ventricular hypertrophy have abnormal diastolic filling in comparison with age-matched control subjects.

In an earlier study of systolic and diastolic flow abnormalities in 17 patients with hypertensive hypertrophic cardiomyopathy of the elderly, Pearson et al46 not only observed increased peak A-wave amplitudes but also observed elevated velocities in the LV outflow tract corresponding to significant systolic intraventricular gradients. In their study significant differences in shortening fraction were noted between groups, in comparison to our study in which indexes of systolic function were not significantly different. Hence, their study represents a group somewhat different from the group under consideration in our study because their study included subjects with more profound cardiac manifestations of hypertension, including smaller end-diastolic dimension, greater shortening fraction, and intraventricular pressure gradients. In this more pathological group, no conventional Doppler-based E-wave index differentiated between groups.

In a digitized M-mode, apexcardiographic, and Doppler study of 60 patients with anatomic LV hypertrophy, including patients with hypertrophic myopathy, valvular or subvalvular aortic stenosis, and 6 patients with systemic hypertension, Shapiro and Gibson47 concluded that abnormalities of selected Doppler indexes do not reflect a single underlying abnormality. They proposed four possibilities: prolongation of isovolumic relaxation, incoordination during isovolumic relaxation, reduced rate of rapid filling, and increased A-wave amplitude likely caused by increased passive stiffness. A method of Doppler contour analysis that could differentiate among these proposed possibilities awaits development.

Filling patterns in LV hypertrophy were characterized using Doppler and acoustic quantification by Chenzbraun and colleagues.48 Thirty patients with LV hypertrophy due to asymmetrical or concentric hypertrophy were compared with 16 healthy non–age-matched control subjects. For the Doppler parameters considered, significant differences were observed between the hypertensive and normal group except for the rapid filling contribution (E-wave integral/total filling integral). Because of a lack of age-matched control subjects, the value of specific E-wave parameter comparison would be limited. It was concluded that acoustic quantification complements the Doppler method and combined use may improve the diagnosis of diastolic abnormalities.

Therefore, relative to the conventional approaches in other studies, the PDF formalism–based MBIP method used in this study is unique. The results of the present study demonstrate alterations of the E-wave contour due to hypertension elicited by MBIP. Thus, the differences in MBIP parameters reflect abnormal diastolic function in patients with hypertension. This novel method provides a new clinical tool for systematic quantitative analysis of Doppler contours in pathophysiological states known to alter diastolic function (diabetes, ischemic syndromes, restrictive/constrictive physiology, transplant rejection, and CHF among others). It also permits Doppler-based assessment of response to pharmacological therapy directed at diastolic dysfunction. The method suggests that derivation of clinically relevant noninvasive imaging–based diastolic function indexes stemming from kinematic physiological modeling is an attainable goal.

Study Limitations
Prolongation of isovolumic relaxation is an important and sensitive marker of diastolic dysfunction in patients with LV hypertrophy. One potential limitation of our approach is that because the PDF model is kinematic, and because its prediction for flow begins at the onset of mitral valve opening, the isovolumic relaxation phase is not an explicit component of the model. However, the consequences of delayed relaxation, which manifest as alteration of the E-wave contour relative to control,25 26 27 result in alteration of the PDF parameters relative to control. Therefore, PDF can account for prolongation of isovolumic relaxation to the extent that prolonged relaxation results in an E-wave contour that is different than control. This is most easily seen in the damping parameter c (see Table 3Up).

A minor limitation of this study relates to sample size. Although, the number of patients in each group is modest, our groups are comparable to those in other studies. Moreover, the Doppler contours in this study are typical of those observed in other studies1 2 3 4 5 6 7 8 9 10 and are also characteristic of what we encounter in clinical practice in a hypertensive group of subjects.


*    Selected Abbreviations and Acronyms
 
E/A = early peak transmitral flow to late transmitral flow velocity ratio
LV = left ventricular
MBIP = model-based image processing
PDF = parametrized diastolic filling


*    Acknowledgments
 
This study was supported in part by the Whitaker Foundation, Washington, DC, and the National Institutes of Health (R0-1 HL54179). Abigail L. Manson McGuire is a Compton Scholar, Physics Department, Washington University School of Arts and Sciences. We thank the staff of the Echocardiography Laboratory: Peggy Brown, Galena Dub, and Tinoa Terry for their assistance in clinical data acquisition and Scott Nudelman for assistance in manuscript preparation.

Received January 20, 1997; first decision February 20, 1997; accepted April 8, 1997.


*    References
up arrowTop
 up arrowAbstract
 up arrowIntroduction
 up arrowMethods
 up arrowResults
 up arrowDiscussion
 *References
 

  1. Hanrath P, Mathey DG, Siegert R, Bleifeld W. Left ventricular relaxation and filling pattern in different forms of left ventricular hypertrophy: an echocardiographic study. Am J Cardiol. 1980;45:15-23.[Medline] [Order article via Infotrieve]
  2. Fouad FM, Tarazi RC, Gallagher JH, MacIntyre WJ, Cook SA. Abnormal left ventricular relaxation in hypertensive patients. Clin Sci. 1980;59(suppl 6):411S-414S.
  3. Kitabatake A, Inoue M, Asao M, Tanouchi J, Masuyama T, Abe H, Morita H, Senda S, Matsuo H. Transmitral blood flow reflecting diastolic behavior of the left ventricle in health and disease: a study by pulsed Doppler technique. Jpn Circ J. 1982;46:92-102.[Medline] [Order article via Infotrieve]
  4. Hartford M, Wikstrand J, Wallentin I, Ljungman S, Wilhelmsen L, Berglund G. Diastolic function of the heart in untreated primary hypertension. Hypertension. 1984;6:329-338.[Abstract/Free Full Text]
  5. Fouad FM, Slominski JM, Tarazi RC. Left ventricular diastolic function in hypertension: relation to left ventricular mass and systolic function. J Am Coll Cardiol. 1984;3:1500-1506.[Abstract]
  6. Papademetriou V, Gottdiener JS, Fletcher RD, Freis ED. Echocardiographic assessment by computer-assisted analysis of diastolic left ventricular function and hypertrophy in borderline or mild systemic hypertension. Am J Cardiol. 1985;56:546-550.[Medline] [Order article via Infotrieve]
  7. Smith VE, Schulman P, Karimeddini M, White WB, Meeran MK, Katz AM. Rapid left ventricular filling in left ventricular hypertrophy II: pathologic hypertrophy. J Am Coll Cardiol. 1985;5:869-874.[Abstract]
  8. Phillips RA, Coplan NL, Krakoff LR, Yeager K, Ross RS, Gorlin R, Goldman ME. Doppler echocardiographic analysis of left ventricular filling in treated hypertensive patients. J Am Coll Cardiol. 1987;9:317-322.[Abstract]
  9. Lorell BH, Grossman W. Cardiac hypertrophy: the consequences for diastole. J Am Coll Cardiol. 1987;9:1189-1193.[Medline] [Order article via Infotrieve]
  10. Douglas PS, Berko B, Lesh M, Reichek N. Alterations in diastolic function in response to progressive left ventricular hypertrophy. J Am Coll Cardiol. 1989;13:461-467.[Abstract]
  11. Nishimura RA, Housmans PR, Hatle LK, Tajik AJ. Assessment of diastolic function of the heart: background and current applications of Doppler echocardiography, 1. physiologic and pathophysiologic features. Mayo Clin Proc. 1989;64:71-81.[Medline] [Order article via Infotrieve]
  12. Nishimura RA, Able MD, Hatle LK, Tajik AJ. Assessment of diastolic function of the heart: background and current applications of Doppler echocardiography, 2. Clinical studies. Mayo Clin Proc. 1989;64:181-204.[Medline] [Order article via Infotrieve]
  13. Plotnick GD. Changes in diastolic function: difficult to measure, harder to interpret. Am Heart J. 1989;118:637-641.[Medline] [Order article via Infotrieve]
  14. LeWinter MM, Eivind ESP, Slinker BK. Influence of the pericardium and ventricular interaction on diastolic function. In: Gaasch W, LeWinter ML, eds. Left Ventricular Diastolic Dysfunction and Heart Failure. Philadelphia, PA: Lea & Febiger; 1994:103-117.
  15. Keren G, Sherez J, Megidish R, Levitt B, Laniado S. Pulmonary venous flow pattern; its relationship to cardiac dynamics: a pulsed Doppler echocardiographic study. Circulation. 1985;71:1105-1112.[Abstract/Free Full Text]
  16. Keren G, Meisner JS, Sherez J, Yellin EL, Laniado S. Interrelationship of mid-diastolic mitral motion, pulmonary venous flow and transmitral flow. Circulation. 1986;74:36-44.[Abstract/Free Full Text]
  17. Kuechere HF, Kusumoto F, Muhiudeen IA, Calahan MK, Schiller NB. Pulmonary venous flow patterns by transesophageal pulsed Doppler echocardiography: relation to parameters of left ventricular systolic and diastolic function. Am Heart J. 1991;122:1683-1693.[Medline] [Order article via Infotrieve]
  18. Masuyama T, Lee J-M, Yamamoto K, Tanouchi J, Hori M, Kamada T. Analysis of pulmonary venous flow velocity patterns in hypertensive hearts: its complementary value in the interpretation of mitral flow velocity patterns. Am Heart J. 1992;124:983-994.[Medline] [Order article via Infotrieve]
  19. Hall AF, Bettlach JA, Nudelman SP, Kovács SJ. Manufacturer and machine setting based variation of parameters resulting from model-based image processing of echocardiographic transmitral Doppler velocity profiles. Proceedings of the 1996 IEEE Ultrasonics Symposium, San Antonio TX. 1996:1205-1210.
  20. Kovács SJ, Barzilai B, Pérez JE. Evaluation of diastolic function with Doppler echocardiography: the PDF formalism. Am J Physiol. 1987;252:H178-H187.[Abstract/Free Full Text]
  21. Hall AF, Kovács SJ. Comparison of audio and video data sources for quantitative analysis of echocardiographic Doppler velocity profiles. Proceedings of the 16th Annual International Conference, IEEE Engineering in Medicine & Biology Society, Baltimore, Md; 1994:662-663.
  22. Hall AF, Kovács SJ Jr. Processing parameter effects on the robustness of the solution to the `Inverse Problem' of diastole from Doppler echocardiographic data. Proceedings of the 15th Annual International Conference, IEEE Engineering in Medicine & Biology Society, San Diego, Calif; 1993:385-387.
  23. Hall AF, Kovács SJ. Automated method for characterization of diastolic transmitral Doppler velocity contours: early rapid filling. Ultrasound Med Biol. 1994;20:107-116.[Medline] [Order article via Infotrieve]
  24. Hall AF, Aronovitz JA, Nudelman SP, Kovács SJ. Automated method for characterization of diastolic transmitral Doppler velocity contours: late atrial filling. Ultrasound Med Biol. 1994;20:859-869.[Medline] [Order article via Infotrieve]
  25. Hall AF, Kovács SJ. Model-based image processing of transmitral Doppler velocity profiles: automated determination of E-wave data set start point. Proccedings of the 17th Annual International Conference, IEEE Engineering in Medicine & Biology Society, Montreal, Canada, 1995:604.
  26. Hall AF, Kovács SJ. Model-based image processing of Doppler velocity profiles: toward automation. Proceedings of the IEEE Ultrasonics, Ferroelectrics and Frequency Control Conference, Seattle, Wash. 1995:1541-1544.
  27. Meisner JS. Left atrial role in left ventricular filling: dog and computer studies. PhD thesis, Albert Einstein College of Medicine; 1986.
  28. Thomas JD, Newell JB, Choong CYP, Weyman AE. Physical and physiological determinants of transmitral velocity: numerical analysis. Am J Physiol. 1991;91:H1718-H1731.
  29. Sun Y, Sjoberg BJ, Ask P, Loyd D, Wranne B. Mathematical model that characterizes transmitral and pulmonary venous flow velocity patterns. Am JPhysiol. 1995;268(1, pt 2):H476-H489.
  30. Schiller NB, Shah PM, Crawford M, Demaria A, Devereux R, Feigenbaum H, Gutgesell H, Reichek N, Sahn D, Schnittger I, Silverman NH, Tajik AJ. Recommendations for quantitation of the left ventricle by two-dimensional echocardiography. J Am Soc Echocardiogr. 1989;2:358-367.[Medline] [Order article via Infotrieve]
  31. Hammermeister KE, Warbasse JR. The rate of change of left ventricular volume in man, II: diastolic events in health and disease. Circulation. 1974;49:739-747.[Abstract/Free Full Text]
  32. Gaasch WH, Cole S, Quinones MA, Alexander JK. Dynamic determinants of left ventricular diastolic pressure volume relations in man. Circulation. 1975;51:317-323.[Abstract/Free Full Text]
  33. Choong CY. Left ventricle: diastolic function-its principles and evaluation. In: Weyman N, ed. Principles and Practice of Echocardiography. Philadelphia, PA: Lea & Febiger; 1994:721-780.
  34. Courtois MA, Kovács SJ, Ludbrook PA. Transmitral pressure-flow velocity relation. Importance of regional pressure gradients in the left ventricle during diastole. Circulation. 1988;78:661-671.[Abstract/Free Full Text]
  35. Choong CY, Abascal VM, Thomas JD, Guerrero JL, McGlew S, Weyman AE. Combined influence of ventricular loading and relaxation on the transmitral flow velocity profile in dogs measured by Doppler echocardiography. Circulation. 1988;78:672-683.[Abstract/Free Full Text]
  36. Choong CY, Herrmann HC, Weyman AE, Fifer MA. Preload dependence of Doppler-derived indexes of left ventricular diastolic function in humans. J Am Coll Cardiol. 1987;10:800-808.[Abstract]
  37. Sartori MP, Quinones MA, Kuo LC. Relation of Doppler derived LV filling parameters to age and radius/thickness ratio in normal and pathologic states. Am J Cardiol. 1987;59:1179-1182.[Medline] [Order article via Infotrieve]
  38. Appleton CP. Influence of incremental changes in heart rate on mitral flow velocity: assessment in lightly sedated, conscious dogs. J Am Coll Cardiol. 1991;17:227-236.[Abstract]
  39. Miyatake K, Okamoto M, Kinoshita N, Owa M, Nakasone I, Sakakibara H, Nimura Y. Augmentation of atrial contribution to left ventricular inflow with aging as assessed by intracardiac Doppler flowmetry. Am J Cardiol. 1984;53:586-589.[Medline] [Order article via Infotrieve]
  40. Kuo LC, Quinones MA, Rokey, Sartori M, Aninader EG, Zoghbi WA. Quantification of atrial contribution to left ventricular filling by pulsed Doppler echocardiography and the effect of age in normal and diseased hearts. Am J Cardiol. 1987;59:1174-1178.[Medline] [Order article via Infotrieve]
  41. Little WC, Downes TR. Clinical evaluation of left ventricular diastolic performance. Prog Cardiovasc Dis. 1990;32:273-290.[Medline] [Order article via Infotrieve]
  42. Nudelman SP, Manson AL, Hall AF, Kovács SJ. Comparison of diastolic filling models and their fit to transmitral Doppler contours. Ultrasound Med Biol. 1995;21:989-999.[Medline] [Order article via Infotrieve]
  43. Douglas PS. Complex dependence of diastolic function indexes on systolic performance in the canine left ventricle. J Appl Cardiol. 1990;5:291-297.
  44. Kovács SJ, Hall AF. Mitral valve resistance to inertiance ratio determination by Doppler echocardiography. J Am Coll Cardiol. 1994;23(suppl):192A. Abstract.
  45. Pearson AC, Gudipati CR, Nagelhaut D, Sear J, Cohen JD, Labovitz AJ. Echocardiographic evaluation of cardiac structure and function in elderly subjects with isolated systolic hypertension. J Am Coll Cardiol. 1991;17:422-430.[Abstract]
  46. Pearson AC, Gudipati CR, Labovitz AJ. Systolic and diastolic flow abnormalities in elderly patients with hypertensive hypertrophic cardiomyopathy. J Am Coll Cardiol. 1988;12:989-995.[Abstract]
  47. Shapiro LM, Gibson DG. Patterns of diastolic dysfunction in left ventricular hypertrophy. Br Heart J. 1988;59:438-445.[Abstract/Free Full Text]
  48. Chenzbraun A, Pinto FJ, Popylisen S, Schnittger I, Popp RL. Filling patterns in left ventricular hypertrophy: a combined acoustic quantification and Doppler study. J Am Coll Cardiol. 1994;23:1179-1185.[Abstract]



This article has been cited by other articles:


Home page
 Am. J. Physiol. Heart Circ. Physiol.Home page
M. M. Riordan, E. P. Weiss, T. E. Meyer, A. A. Ehsani, S. B. Racette, D. T. Villareal, L. Fontana, J. O. Holloszy, and S. J. Kovacs
The effects of caloric restriction- and exercise-induced weight loss on left ventricular diastolic function
Am J Physiol Heart Circ Physiol, March 1, 2008; 294(3): H1174 - H1182.
[Abstract] [Full Text] [PDF]

Home page
 Am. J. Physiol. Heart Circ. Physiol.Home page
L. Shmuylovich and S. J. Kovacs
E-wave deceleration time may not provide an accurate determination of LV chamber stiffness if LV relaxation/viscoelasticity is unknown
Am J Physiol Heart Circ Physiol, June 1, 2007; 292(6): H2712 - H2720.
[Abstract] [Full Text] [PDF]

Home page
 Am. J. Physiol. Heart Circ. Physiol.Home page
M. M. Riordan and S. J. Kovacs
Absence of diastolic mitral annular oscillations is a marker for relaxation-related diastolic dysfunction
Am J Physiol Heart Circ Physiol, June 1, 2007; 292(6): H2952 - H2958.
[Abstract] [Full Text] [PDF]

Home page
 CirculationHome page
S. Fukushima, A. Varela-Carver, S. R. Coppen, K. Yamahara, L. E. Felkin, J. Lee, P. J.R. Barton, C. M.N. Terracciano, M. H. Yacoub, and K. Suzuki
Direct Intramyocardial But Not Intracoronary Injection of Bone Marrow Cells Induces Ventricular Arrhythmias in a Rat Chronic Ischemic Heart Failure Model
Circulation, May 1, 2007; 115(17): 2254 - 2261.
[Abstract] [Full Text] [PDF]

Home page
 Am. J. Physiol. Heart Circ. Physiol.Home page
C. S. Chung, A. Strunc, R. Oliver, and S. J. Kovacs
Diastolic ventricular-vascular stiffness and relaxation relation: elucidation of coupling via pressure phase plane-derived indexes
Am J Physiol Heart Circ Physiol, November 1, 2006; 291(5): H2415 - H2423.
[Abstract] [Full Text] [PDF]

Home page
 J. Appl. Physiol.Home page
L. Shmuylovich and S. J. Kovacs
Load-independent index of diastolic filling: model-based derivation with in vivo validation in control and diastolic dysfunction subjects
J Appl Physiol, July 1, 2006; 101(1): 92 - 101.
[Abstract] [Full Text] [PDF]

Home page
 J. Appl. Physiol.Home page
M. M. Riordan and S. J. Kovacs
Quantitation of mitral annular oscillations and longitudinal "ringing" of the left ventricle: a new window into longitudinal diastolic function
J Appl Physiol, January 1, 2006; 100(1): 112 - 119.
[Abstract] [Full Text] [PDF]

Home page
 Am. J. Physiol. Heart Circ. Physiol.Home page
C. S. Chung, M. Karamanoglu, and S. J. Kovacs
Duration of diastole and its phases as a function of heart rate during supine bicycle exercise
Am J Physiol Heart Circ Physiol, November 1, 2004; 287(5): H2003 - H2008.
[Abstract] [Full Text] [PDF]

Home page
 Am. J. Physiol. Heart Circ. Physiol.Home page
Y. Y. Li, T. Kadokami, P. Wang, C. F. McTiernan, and A. M. Feldman
MMP inhibition modulates TNF-alpha transgenic mouse phenotype early in the development of heart failure
Am J Physiol Heart Circ Physiol, March 1, 2002; 282(3): H983 - H989.
[Abstract] [Full Text] [PDF]

Home page
 J. Appl. Physiol.Home page
J. B. Lisauskas, J. Singh, A. W. Bowman, and S. J. Kovacs
Chamber properties from transmitral flow: prediction of average and passive left ventricular diastolic stiffness
J Appl Physiol, July 1, 2001; 91(1): 154 - 162.
[Abstract] [Full Text] [PDF]

Home page
 Proc. Natl. Acad. Sci. USAHome page
Y. Y. Li, Y. Q. Feng, T. Kadokami, C. F. McTiernan, R. Draviam, S. C. Watkins, and A. M. Feldman
Myocardial extracellular matrix remodeling in transgenic mice overexpressing tumor necrosis factor alpha can be modulated by anti-tumor necrosis factor alpha therapy
PNAS, November 7, 2000; 97(23): 12746 - 12751.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow Request Permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Kovács, S. J.
Right arrow Articles by Hall, A. F.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Kovács, S. J.
Right arrow Articles by Hall, A. F.