Ultrasound integrated backscatter tissue characterization of remote myocardial infarction in human subjects

The variation of integrated backscatter in human hearts in differing ultrasonic transthoracic views

It has been shown previously that in normal subjects the interventricular septum imaged in the long-axis view (LAX) and the left ventricular posterior wall imaged in both the LAX and the short-axis view (SAX) exhibit cyclic variation of integrated backscatter (IB) throughout the cardiac cycle, with maximum values occurring at end diastole (ED) and minimum at end systole (ES). The ability to demonstrate this cyclic variation within these myocardial regions in only two ultrasonic views has limited the potential clinical utility of an IB imaging system. To determine whether clinically useful information on the variation of IB is available from different myocardial regions in different ultrasonic views, we measured ED to ES variation of IB from the parasternal and apical views in normal subjects with a radiofrequency acquisition technique. Two independent clinical observers analyzed ED to ES variation of IB from 14 normal volunteers (mean age 32 +/- 6 years; range 21 to 45 years) in reconstructed two-dimensional ultrasonic images obtained from the parasternal LAX and SAX and apical two-chamber (2C) and four-chamber (4C) views. ED to ES variation of IB was measured from manually traced regions of interest (ROI) within the myocardium. These ROIs were chosen interactively and were located within the midposterior wall and the midanteroseptum in LAX views; within the midposterior wall, midanteroseptum, midseptum, and midlateral wall in SAX views; within the midseptum and the midlateral wall in 4C views; and within the midinferior wall and the midanterior wall 2C views. In all analyzed ROIs within the parasternal and apical views, ED to ES variation of IB was found. We have shown that the maximum magnitude of IB was at ES within the midseptum and in 10 out of 14 volunteers in the midanteroseptum measured from SAX views, the midanterior wall from 2C views, and the midlateral wall from 4C views. The rest of the ROIs analyzed exhibited the maximum value of IB cyclic variation at ED. We have confirmed that the ED to ES variation of IB is present not only when measured from the two standard parasternal views but also from the two apical views in all analyzed myocardial walls, and the minimum of this cyclic variation was not always coincident with ES nor the maximum with ED.

Validation of transthoracic myocardial ultrasonic tissue characterization: comparison of transthoracic and open-chest measurements of integrated backscatter

To investigate whether myocardial integrated backscatter (IB) can be measured through the chest wall, myocardial IB parameters were measured in five adult mongrel dogs with a newly developed IB imaging system capable of measurements of myocardial IB relative to backscatter from the blood. There was no significant difference in the calibrated myocardial IB between the closed chest and the open chest conditions either in the septum or in the posterior wall if a 2.5- or 3.5-MHz frequency transducer was used. There was no significant difference in the magnitude of cyclic variation in IB between the closed chest and the open chest conditions independent of the frequency of the transducer used. These data suggest that we can accurately measure not only the magnitude of cyclic variation in IB but also the calibrated myocardial IB through the chest wall with a 2.5- or 3.5-MHz frequency transducer. Such data may validate measurements of myocardial IB parameters through the chest wall even in humans.

Acoustic propagation properties of normal, stunned, and infarcted myocardium. Morphological and biochemical determinants

BACKGROUND

Identification of viable but stunned myocardium remains a major problem. Since stunned myocardium results in impairment of myocardial function without any structural damage and infarcted myocardium causes major structural disruption, we postulated that acoustic properties could distinguish between the two insults.

METHODS AND RESULTS

Anesthetized open-chest dogs underwent a total occlusion of the left anterior descending coronary artery for 15 minutes (stunned, n = 7) and 90 minutes (infarcted, n = 8), followed by reperfusion for 3 hours. Circumflex coronary artery perfusion territory (n = 15) served as normal control tissue. Regions of myocardium were quantitatively evaluated with a scanning laser acoustic microscope operating at 100 MHz and a research ultrasound system operating at 4 to 7 MHz. Four ultrasonic parameters were determined: attenuation coefficient (an index of loss per unit distance), speed of propagation, a spatial variation of propagation speed called the heterogeneity index (HI), and ultrasonic backscatter at 5 MHz (IBR5). Myocardial water, lipid, and protein contents of normal, stunned, and infarcted myocardium were also determined. The attenuation coefficient of normal myocardium (179 +/- 20 dB/cm) was significantly greater than that of stunned (136 +/- 7 dB/cm, P < .001) and infarcted (130 +/- 8 dB/cm, P < .001) myocardium. The propagation speed of normal myocardium (1597 +/- 6 m/s) was similar to that of stunned (1600 +/- 6 m/s) and significantly higher than that of infarcted (1575 +/- 7 m/s, P < .001) myocardium. The HI for specimen thicknesses of 75 to 100 microns showed an increase of 33% between normal (5.0 +/- 0.8 m/s) and stunned (7.5 +/- 2.3 m/s, P < .05) myocardium. However, for the infarcted myocardium (5.8 +/- 2.0 m/s), the HI was essentially the same as that of the normal myocardium (5.0 +/- 0.8 m/s). The IBR5 of normal (-47.1 +/- 1.0 dB) was not significantly different from that of stunned myocardium (-46.8 +/- 0.9 dB). The IBR5 of infarcted myocardium (-42.4 +/- 1.0 dB) was significantly greater than that of normal myocardium. Myocardial water and protein contents were similar in the normal and stunned myocardium. Water content in the infarcted myocardium (80.8 +/- 2%) was significantly greater (P < .05) than in the normal (72.7 +/- 1.3%), and protein content of 18.5 +/- 0.7% was significantly lower (P < .05) than the normal (21.4 +/- 0.8%). Lipid content was increased in the stunned (8.5 +/- 0.5%) and virtually absent in the infarcted myocardium (0.8 +/- 0.3%) compared with normal (5.5 +/- 0.6%).

CONCLUSIONS

We conclude that acoustic propagation properties can identify stunned and infarcted myocardium and may be related to biochemical/morphological differences.

When can Doppler be used in place of integrated backscatter as a measure of scattered ultrasound intensity?

The purpose of this work was to determine under what circumstances the intensity of Doppler audio signals can be used as a substitute for the more direct and complex measure of ultrasonic backscatter (integrated backscatter) which requires radio-frequency ultrasound signals. Using a rotating rubber disk phantom and a microbubble echo-contrast flow phantom, we have shown that the intensity of audio Doppler signals is independent of the constraints typically associated with Doppler ultrasound (velocity and angle), but like integrated backscatter depends on the transmit intensity, gain of the ultrasound receiver, attenuation and the nature of the scatterers. Using Doppler ultrasound for backscatter measurements is ideally suited for the expected application of the technique: the assessment of echo contrast in cardiac chambers, blood vessels and tissue perfusion (i.e., any flow system). Compared to integrated backscatter, the Doppler audio method has the advantage of using standard clinical ultrasound machines, requires less sophisticated data storage and processing equipment and the positioning system for the region of interest (the Doppler sample volume) is built into all pulsed-wave Doppler machines. Further, the low-velocity filter removes all nonmoving scatterers (like the intense echoes from heart valves and the walls of blood vessels), thus allowing study of only those echoes originating from the blood pool. This combination of features is what attracted us to the Doppler method for quantitating ultrasonic backscatter in flow systems.

Myocardial tissue characterization by autocorrelation of two-dimensional ultrasonic backscatter

To evaluate a novel method for determining the spatial distribution of echocardiographic information based on the two-dimensional autocorrelation function, echocardiographic images were obtained from specific regions of interest from 10 healthy volunteers, seven patients with genetically defined hypertrophic cardiomyopathy, and nine patients with pressure-overload hypertrophy. The wavelength of distinct peaks from the two-dimensional autocorrelation of the images was compared between groups of patients and demonstrated a significant decrease in the mean length scale associated with distinct secondary correlation peaks in patients with hypertrophic cardiomyopathy or pressure-overload hypertrophy compared with healthy volunteers (p = 0.0009). With a discriminating wavelength of 3.3 mm, the sensitivity of this technique for detecting abnormal myocardium was 84% with a specificity of 89%. This study suggests that ultrasonic tissue characterization based on the two-dimensional autocorrelation function may have potential for distinguishing normal from abnormal myocardium and provides a rationale for textural approaches to ultrasonic tissue characterization.

Analysis of transmural trend of myocardial integrated ultrasound backscatter for differentiation of hypertrophic cardiomyopathy and ventricular hypertrophy due to hypertension

OBJECTIVES

This study was undertaken to differentiate hypertrophic cardiomyopathy from hypertensive hypertrophy using a newly developed M-mode format integrated backscatter imaging system capable of calibrating myocardial integrated backscatter with the power of Doppler signals from the blood.

BACKGROUND

Myocardial integrated ultrasound backscatter changes in patients with hypertrophic cardiomyopathy; however, it is unknown whether ultrasound myocardial tissue characterization may be useful in differentiating hypertrophic cardiomyopathy from hypertensive hypertrophy.

METHODS

Calibrated myocardial integrated backscatter and its transmural gradient were measured in the septum and posterior wall in 31 normal subjects, 13 patients with hypertensive hypertrophy and 22 patients with hypertrophic cardiomyopathy. The gradient in integrated backscatter was determined as the ratio of calibrated integrated backscatter in the endocardial half to that in the epicardial half of the myocardium.

RESULTS

Cyclic variation of integrated backscatter was smaller and calibrated myocardial integrated backscatter higher in patients with hypertrophied hearts than in normal subjects, but there were no significant differences in either integrated backscatter measure between patients with hypertensive hypertrophy and those with hypertrophic cardiomyopathy. Transmural gradient in myocardial integrated backscatter was present only in patients with hypertrophic cardiomyopathy (5.0 +/- 1.8 dB [mean +/- SD] for the septum; 1.2 +/- 1.6 dB for the posterior wall).

CONCLUSIONS

Hypertrophic cardiomyopathy and ventricular hypertrophy due to hypertension can be differentiated on the basis of quantitative analysis of the transmural gradient in integrated backscatter.

Quantitative myocardial ultrasonic integrated backscatter measurements during contrast injections

We and others have shown that normal myocardium exhibits 4 to 5 dB diastolic-to-systolic cyclic variation (CV) of integrated backscatter. To investigate the effect of intramyocardial contrast on integrated backscatter, we injected 5% sonicated albumin, containing microbubbles in the range of 5 microns in diameter, into the left atrium in nine open-chest dogs. The dogs were anesthetized and placed in the right lateral decubitus position on a specially designed table with a cutout allowing ultrasound imaging from below. Ultrasonic data was obtained from the right precordium by use of a prototype M-mode integrated backscatter system implemented in a commercially available two-dimensional system. Usable data were obtained in eight of nine dogs. Integrated backscatter increased up to 13 dB after contrast injections. There was a significantly decreased CV of integrated backscatter during myocardial contrast in all eight dogs. The mean level of CV of integrated backscatter for the eight dogs decreased from 4.7 dB (530 beats analyzed) without contrast to 2.8 dB during contrast (436 beats analyzed). There was a trend to greater CV at higher levels of contrast. Septal excursion, as measured by M-mode echocardiography simultaneously with integrated backscatter by the same ultrasound beam, was similar with and without contrast (mean 8.2 vs 8.3 mm). Thus left atrium contrast injection produces quantitatively measurable integrated backscatter effects. Cyclic variation of integrated backscatter decreases with contrast. However, at higher contrast levels the decrease tends to be smaller. These effects should be considered during quantitative tissue characterization and myocardial contrast studies.

Quantitative Echocardiography, Part 1

The quantitative applications of echocardiography in clinical practice are well known. The measurements of cardiac chamber dimensions, wall thickness, and overall performance have been uniformly adopted. An important emerging ultrasound modality known as tissue characterization of the myocardium has evolved from experimental studies to clinical investigation. The ability to quantitate myocardial acoustic properties by the measurement of integrated backscatter (in decibels) provides direct assessment of myocardial structural characteristics and contractile performance, to complement conventional two-dimensional imaging of ventricular wall motion and wall thickening. Despite the considerable amount of work that has been done, there are several areas of research that need to be further investigated before widespread clinical use of these techniques is possible. Specifically, absolute values of myocardial backscatter are not yet obtainable with the current instrumentation; only the relative change in backscatter during the cardiac cycle (cyclic variation) has been defined and employed in clinical studies. This review summarizes the principles of tissue characterization and the results of several clinical studies, specifically those carried out in patients with coronary artery disease and cardiomyopathies.

Quantification of ventricular remodeling in the tight-skin mouse cardiomyopathy with acoustic microscopy

To determine the role of ultrasonic tissue characterization for the detection of changes in myocardial architecture associated with cardiomyopathy, acoustic microscopy was performed on the hearts of 4- to 6-month-old tight-skin mice [TSK/+, C57-B10.D2 (58B)/SN strain], a model of cardiomyopathy characterized by diffuse interstitial fibrosis. Ultrasonic backscatter was measured from excised segments of left ventricular free walls of five TSK mice and five sex- and age-matched normal controls with a 50 MHz broad band focused piezoelectric transducer operated in a saline-filled water tank at room temperature. Forty-nine radio frequency (RF) lines were digitized from each specimen at 2 ns/sample. Power spectral analysis of RF data was performed and mean integrated backscatter (IB) computed. The TSK group demonstrated greater IB (-53.6 +/- 0.6 dB, n = 5) than did the control group (-56.6 +/- 0.7 dB, n = 5; p < 0.02). Myocardial collagen content determined by hydroxyproline assay increased by 11% in the TSK group (2.54 +/- 0.08 microgram/mg dry wt, n = 5) over that in controls (2.28 +/- 0.07 microgram/mg dry wt, n = 5; p < 0.05). A significant linear relationship was observed between myocardial hydroxyproline concentration and IB (r = 0.74; p < 0.02). Thus, ultrasonic tissue characterization permits sensitive detection of modest changes in the extent of interstitial fibrosis that accompany tissue remodeling in the early stages of cardiomyopathy.

Automated, on-line quantification of left ventricular dimensions and function by echocardiography with backscatter imaging and lateral gain compensation

To provide on-line quantification of left ventricular cavity dimensions and function by echocardiography 60 control subjects and 10 patients with cardiac dysfunction were studied. A novel, ultrasound imaging system was used which was developed to detect and track, in real time, ventricular endocardial blood boundaries based on quantitative assessment of acoustic properties of tissue. In addition, lateral gain compensation, a robust and novel image enhancement procedure, was used to provide instantaneous measurement and display of cavity areas and functional indexes on a beat-by-beat basis within regions of interest drawn around the blood pool cavity. In control subjects, short-axis end-diastolic area averaged 13.1 +/- 3.7 cm2 (SD), end-systolic area 5.9 +/- 2.7 cm2, and fractional area change 55.6 +/- 11.2%. Apical views yielded corresponding values of 23.8 +/- 4.5 cm2, 15.5 +/- 3.4 cm2 and 34.7 +/- 7.8%. Instantaneous peak rate of cavity area change approximated 50 cm2/s in systole and 60 cm2/s in diastole in each view. Serial measurements of area and functional index were reproducible over intervals of 2 to 3 weeks. Patients with dilated ventricles exhibited average apical view area values of 49.1 +/- 6.1 cm2 and 43.1 +/- 4.9 cm2 in diastole and systole with a fractional area change of 12.2 +/- 3.0%. Thus, results with on-line echocardiographic backscatter imaging-assisted automated edge detection are reproducible and capable of delineating cardiac dysfunction conveniently, promptly and serially at the bedside.

Detection of unique transmural architecture of human idiopathic cardiomyopathy by ultrasonic tissue characterization

BACKGROUND

Noninvasive approaches to the evaluation of idiopathic cardiomyopathy are limited. Recent work from our laboratory has used quantitative ultrasound to define the three-dimensional structure of normal human myocardium and the myocardial remodeling associated with infarction. Our goal was to define the role of ultrasonic tissue characterization for detection of specific alterations in the three-dimensional transmural architecture of idiopathic dilated cardiomyopathy.

METHODS AND RESULTS

We measured frequency-dependent backscatter from 22 cylindrical biopsy specimens from nine explanted fixed hearts of patients who underwent heart transplantation for idiopathic cardiomyopathy, seven specimens from normal portions, and 12 specimens of infarcted tissue from six explanted fixed human hearts. Consecutive transmural levels from each specimen were insonified with a 5-MHz broadband transducer. The dependence of apparent (uncompensated for attenuation) backscatter, B(f), on frequency (f) was computed from radiofrequency (rf) data as: magnitude of B(f)2 = afn, where n is an index that reflects in part the size of the dominant scatterers in myocardial tissue. Myofiber diameter and percentage fibrosis were determined at each transmural level for each specimen. For cardiomyopathic tissue, the frequency dependence of backscatter (n) increased progressively from epicardial to endocardial (0.02 +/- 0.37 to 1.01 +/- 0.12, p less than 0.05) levels in conjunction with a progressive decrease in myofiber diameter (29.5 +/- 0.9 to 21.4 +/- 0.6 microns, p less than 0.0001). In contrast, in tissue from areas of infarction, the frequency dependence decreased progressively from epicardium to endocardium (0.91 +/- 0.20 to 0.23 +/- 0.21, p less than 0.05) in conjunction with a progressive increase in the percentage of fibrosis (23.5 +/- 9.4% to 54.5 +/- 4.9%, p less than 0.005). Normal tissue exhibited no significant transmural trend for frequency dependence, myofiber diameter, or percentage fibrosis.

CONCLUSIONS

These data indicate the presence of a heterogenous transmural distribution of scattering structures associated with human idiopathic cardiomyopathy and myocardial infarction that may be detected by ultrasonic tissue characterization. The divergence of these transmural trends for frequency dependence of backscatter reflects distinct mechanisms of structural heterogeneity for different pathological processes that comprise a transmural gradation of cell size and fibrosis for idiopathic cardiomyopathy and infarction, respectively.

Estimation of left ventricular cavity area with an on-line, semiautomated echocardiographic edge detection system

BACKGROUND

Automated edge detection of endocardial borders in echocardiograms provides objective, reproducible estimation of cavity area; however, most methods have required off-line analysis. A recently developed prototype echocardiographic imaging system permits real-time automated edge detection during imaging and thus, the potential for measurement of cyclic changes in cavity area and the assessment of left ventricular function on-line. Our purpose was to compare measurements of endocardial area manually traced from conventional echocardiograms with those obtained with the real-time automated edge detection system in normal subjects.

METHODS AND RESULTS

Two training sets of images were used to establish optimal methods of gain setting; the settings were then evaluated in a test set of images. In the high-gain training group (n = 8 subjects, 119 images), gain settings were adjusted sufficiently high to display at least 90% of the endocardial border. Manually drawn and real-time area measurements correlated at r = 0.92, but manually drawn areas were underestimated by computer. In the low-gain training group (n = 7 subjects, 104 images), gain settings were adjusted sufficiently low to avoid cavity clutter despite the presence of dropout of endocardial edges. Manually drawn and real-time areas again correlated (r = 0.79), but manually drawn areas were overestimated by computer. In the intermediate-gain test group (n = 7 subjects, 105 images), gain settings were balanced between maximal endocardial definition (greater than or equal to 90%) and minimal cavity clutter (less than or equal to 1 cm2). Manually drawn and real-time areas correlated at r = 0.91 for the group, and r ranged from 0.94 to 0.99 in individual subjects. Interobserver variability was 9.5% for manually traced areas and 10.6% for real-time area measurements.

CONCLUSIONS

Real-time on-line automated edge detection provides accurate estimation of manually drawn cavity areas. Although the method is gain dependent, measurements are reproducible. The system should have clinical application in settings in which measurements of left ventricular function are important.

Quantitative ultrasonic imaging: tissue characterization and instantaneous quantification of cardiac function

Quantitative myocardial tissue characterization is being developed to complement and expand conventional echocardiography by delineating the physical state of myocardium under diverse pathophysiologic conditions. Real-time quantitative integrated backscatter imaging has already been applied to patients with ischemic heart disease, hypertrophic cardiomyopathy, and cardiac allograft rejection in clinical investigations performed in the United States, Europe, and Japan. A recently introduced modification of imaging processing algorithms employed for characterization of tissue facilitates automatic detection of endocardial-blood interfaces and on-line quantification of ventricular size and function. Further progress and anticipated developments in quantitative ultrasonic imaging will undoubtedly augment the clinical applications of tissue characterizations based on myocardial integrated backscatter for improved diagnosis, elucidation of pathophysiology, and assessment of cardiac function.

Abnormal myocardial acoustic properties in diabetic patients and their correlation with the severity of disease

Although patients with diabetes mellitus may be afflicted by cardiomyopathy, its prevalence and nature are controversial. Studies have shown that fibrosis alters the acoustic properties of the heart in animals and humans and that the changes are detectable by cardiac tissue characterization with ultrasound. The present study was performed to characterize myocardial acoustic properties in patients with insulin-dependent diabetes to determine whether ultrasound tissue characterization could detect changes potentially indicative of occult cardiomyopathy. The magnitude of cyclic variation of myocardial ultrasound integrated backscatter and its phase delay with respect to the onset of the cardiac cycle in the septum and posterior wall of the left ventricle were measured in 54 patients with diabetes who had no overt cardiac disease. Conventional echocardiography documented normal ventricular systolic function in 96%. As compared with results in age-matched patients without diabetes studied previously, cyclic variation of integrated backscatter was reduced (4.6 +/- 0.8 vs. 3.6 +/- 1.4 dB; p less than 0.001). In addition, delay was significantly increased (0.86 +/- 0.09 vs. 0.99 +/- 0.15). The primary analysis of the data focused on differences among the diabetic patients. Reduction of cyclic variation of backscatter was greatest in patients with diabetes who had neuropathy (3.2 +/- 1.0 dB; p less than 0.001) as was the increase in delay (1.04 +/- 0.16, p less than 0.001 vs. values in patients without neuropathy). Retinopathy and nephropathy were associated with abnormal myocardial acoustic properties as well. Thus, abnormalities that may reflect fibrosis or other occult cardiomyopathic changes in diabetic patients without overt heart disease are readily detectable by myocardial tissue characterization with ultrasound and parallel the severity of noncardiac diabetic complications.

Reproducibility of quantitative backscatter echocardiographic imaging in normal subjects

Cyclic backscatter variation is useful in differentiating normal from ischemic and myopathic myocardium; however, there are few data on the reproducibility of clinical cyclic variation measurements. Therefore, a study using 2-dimensional and M-mode backscatter imaging was performed in 20 normal male subjects by 2 observers at an initial session and by 1 of the observers after 1 week. Cyclic variation on M-mode was calculated as the difference between the end-diastolic backscatter and the backscatter at the nadir. Two-dimensional determinations of backscatter were made using a single frame at end-diastole and one at end-systole. The cyclic change was the difference between backscatter measured in the end-diastolic and end-systolic frames. There were no statistically significant differences in analysis of variance among the grouped repeated measurements in either the interventricular septum or the posterior left ventricular wall. At the initial session, cyclic backscatter variation in the posterior wall using M-mode techniques was 5.9 +/- 1.8 dB (SD). The cyclic change in backscatter in the septal wall, using the 2-dimensional technique, was 4.3 +/- 2.4 dB. In the posterior wall, the cyclic change in backscatter was 5.7 +/- 1.7 dB. Pairwise observer correlations between repeated measurements ranged from -0.48 to 0.45. Thus, although there were no significant differences in group means on repeat measurements, repeated measurements in individual subjects were not reliably reproduced because of limited independent sampling of backscatter measurements at only 2 points in the heart cycle. Increased independent sampling and measurement from a backscatter waveform throughout the cardiac cycle may improve reproducibility of measurements.

On-line assessment of ventricular function by automatic boundary detection and ultrasonic backscatter imaging

To provide an approach suitable for on-line analysis of ventricular function, a conventional two-dimensional ultrasound imaging system was modified to detect and track blood-tissue interfaces in real time based on their quantitative acoustic properties. This modification permitted on-line display of the left ventricular cavity area, fractional area change, volumes and ejection fraction on a beat by beat basis. Images were obtained from 54 patients and 12 normal subjects with broad ranges of ventricular dimensions and systolic function. On-line measurements of cavity areas were compared with off-line measurements of cavity areas (analysis of videotaped conventional images). Left ventricular cavity areas measured on-line from short-axis views correlated closely with off-line views as did areas from apical views. On-line fractional area change correlated well with ejection fraction calculated off-line. More than 70% of patients could be studied adequately with the approach developed. Thus, automatic boundary detection based on quantitative assessment of tissue acoustic properties permits on-line quantitation of ventricular cavity areas and indexes of function.

Structural remodeling of human myocardial tissue after infarction. Quantification with ultrasonic backscatter

BACKGROUND

Remodeling of myocardial tissue after infarction may culminate in the development of either a well-healed scar or a thin, expanded heart wall segment that predisposes to ventricular aneurysm formation, congestive heart failure, or ventricular tachycardia. The three-dimensional architecture of mature human infarct tissue and the mechanisms that determine it have not been elucidated. We have previously shown that quantitative ultrasonic backscatter can be used to define the transmural organization of human myofibers in the normal ventricular wall by measuring the dependence of backscatter on the angle of insonification, or ultrasonic anisotropy. We propose that measurement of ultrasonic anisotropy of backscatter may permit quantitative characterization of the transmural architecture of tissue from areas of myocardial infarction and facilitate identification of fundamental mechanisms of remodeling of the ventricular wall.

METHODS AND RESULTS

We measured integrated backscatter in 33 transmural sections from 12 cylindrical biopsy specimens (1.4-cm diameter) sampled from central regions of mature infarction in six explanted fixed human hearts. Tissue samples were insonified in two-degree steps around their entire circumference at successive transmural levels with a 5-MHz broad-band piezoelectric transducer. Backscatter radio frequency data were gated from the center of each specimen, and spectral analysis was performed on the gated radio frequency for the computation of integrated backscatter. Histological morphometric analysis was performed on each specimen for determination of the predominant fiber orientation and the percentage of tissue infarcted at consecutive transmural levels. The average percentage of tissue infarcted for all transmural levels was 49 +/- 3% (range, 13-80%). Histological attributes varied from patchy fibrosis to extensive confluent zones of scar tissue. The angle-averaged integrated backscatter for all transmural levels in infarct tissue was approximately 5 dB greater than that previously measured in normal tissue in our laboratory (-48.3 +/- 0.5 versus -53.4 +/- 0.4 dB, infarct versus normal). Marked anisotropy of backscatter was observed in tissue from areas of infarction and was characterized by a sinusoid-like dependence on the angle of insonification at each transmural level. Insonification perpendicular to infarct fibers yielded values for integrated backscatter 14.8 +/- 0.5 dB greater than those for insonification parallel to these fibers. Juxtaposition of the sinusoid-like anisotropy functions from all consecutive transmural levels demonstrated a progressive shift in the orientation of scar tissue elements from epicardial to endocardial levels of 14.6 +/- 1.5 degrees/mm of tissue. The transmural shift in fiber orientation per millimeter of tissue from the area of infarction exceeded that previously measured for normal tissue (9.2 +/- 0.7 degrees/mm) by 59%. This marked augmentation in angular shift per millimeter of tissue results from a generalized structural rearrangement (or reorientation) of fibers across the entire ventricular wall in the infarct zone that we hypothesize is determined in part by dynamic mechanical forces, imposed by the surrounding functional normal tissue, that tether the "infarcted" tissue.

CONCLUSIONS

Myocardial tissue from areas of myocardial infarction manifests substantial anisotropy of ultrasonic scattering that may be useful for quantitative characterization of the alignment and overall three-dimensional anatomic organization of mature infarct scars.

Ultrasonic backscatter tissue characterization in cardiac diagnosis

Ultrasonic tissue characterization has shown the potential to yield information about structural and functional properties of cardiovascular tissue. The development of real-time two-dimensional integrated backscatter imaging has made feasible clinical investigations of ultrasonic tissue characterization, including detection of stunned myocardium in patients with acute ischemia, recognition of remote infarction, detection of cardiac allograft rejection, and study of diffuse myocardial involvement with systemic diseases such as diabetes mellitus. Technical improvements and scientific advances in the understanding of the interaction between ultrasound and tissue may open an even wider range of clinical applications. Even in its present, relatively preliminary form, tissue characterization appears to have the potential for clinical application. Additional clinical experience will stimulate refinements and increases in the diagnostic power of this promising approach.

Three-dimensional characterization of human ventricular myofiber architecture by ultrasonic backscatter

Normal human left ventricular architecture comprises a highly aligned array of cardiac myofibers whose orientation depends on transmural location. This study was designed to determine whether measurement of integrated backscatter could be used detect the progressive transmural shift of myofiber alignment that occurs from epicardium to endocardium in human ventricular wall segments. Integrated backscatter was measured at 32 transmural levels in seven cylindrical biopsy specimens (1.4 cm diam) sampled from normal regions of six explanted fixed human hearts by insonification of samples at 180 independent angles in 2 degrees steps around their entire circumference with a 5-MHz broadband piezoelectric transducer. Histologic analysis was performed to determine fiber orientation. Integrated backscatter varied approximately as a sinusoidal function of the angle of insonification at each transmural level. Greater integrated backscatter was observed for insonification perpendicular as compared with parallel to fibers (difference = 14.5 +/- 0.6 dB). Ultrasonic analysis revealed a progressive transmural shift in fiber orientation of approximately 9.2 +/- 0.7 degrees/mm of tissue. Histologic analysis revealed a concordant shift in fiber orientation of 7.9 +/- 0.8 degrees/mm of tissue. Thus, human myocardium manifests anisotropy of ultrasonic scattering that may be useful for characterization of the intramural fiber alignment and overall three-dimensional organization of cardiac myofibers.

Estimation of myocardial infarct size with ultrasonic tissue characterization

BACKGROUND

Ultrasonic tissue characterization (UTC) can distinguish normal from infarcted myocardium. Infarcted myocardium shows an increase in integrated backscatter and loss of cardiac cycle-dependent variation in backscatter. The cyclic variation of backscatter is closely related to regional myocardial contractile function; the latter is a marker of myocardial ischemia. The present study was designed to test the hypothesis that intramural cyclic variation of backscatter can map and estimate infarct size.

METHODS AND RESULTS

Transmural myocardial infarction was produced in 12 anesthetized, open-chest dogs by total occlusion of the left anterior descending coronary artery for 4 hours. A real-time ultrasonic tissue characterization instrument, which graphically displays integrated backscatter Rayleigh 5, cardiac cycle-dependent variation, and patterns of cyclic variation in backscatter, was used to map infarct size and area at risk of infarction. Staining with 2,3,4-triphenyltetrazolium chloride (TTC) and Patent Blue Dye was used to estimate infarct size and the area at risk, respectively. The ratio of infarct size to area at risk of infarction determined with UTC correlated well with that determined with TCC (r = 0.862, y = 23.7 +/- 0.792x). Correlation coefficients for infarct size and area at risk were also good (r = 0.736, y = 12.3 +/- 737x for infarct size and r = 0.714, y = 5.80 +/- 1.012x for area at risk). However, UTC underestimated both infarct size and area at risk.

CONCLUSIONS

Ultrasonic tissue characterization may provide a reliable, noninvasive method to estimate myocardial infarct size.

The effect of frequency on the magnitude of cyclic variation of backscatter in dogs and implications for prompt detection of acute myocardial ischemia

The magnitude of cyclic variation of integrated backscatter was measured for 3-4, 4-5, 5-6, 6-7, and 7-8 MHz. In ten normal dogs, the magnitude of cyclic variation (M) was found to increase with ultrasonic frequency in an approximately linear fashion. The least squares linear fit to the data yielded M=2.5 dB+0.24f dB/MHz where f is the ultrasonic frequency (MHz). The potential frequency dependence of detection of the immediate consequences of myocardial ischemia was investigated. Acute ischemic injury was induced in each of seven dogs by ligation of a coronary artery. The magnitude of cyclic variation of integrated backscatter was measured in regions of myocardium supplied by this artery before and after ligation. Ischemic myocardium was clearly differentiable from normal myocardium in all five frequency bands. The magnitude of cyclic variation of integrated backscatter demonstrated substantial recovery upon reperfusion. The results offer promise for the detection of ischemia in humans using clinical imaging systems.

Diagnosis of recent myocardial infarction with quantitative backscatter imaging: preliminary studies

Acute myocardial ischemia and chronic myocardial infarction may be recognized with ultrasound tissue characterization techniques because of myocardial acoustic changes caused by reduced perfusion and/or collagen deposition. Our purpose was to study the acoustic properties of recent myocardial infarction when the predominating pathologic finding was myocardial edema and leukocytic infiltration. We used a new quantitative backscatter imaging system to study 18 patients 9 +/- 5 days after myocardial infarction (eight patients with anteroseptal myocardial infarction and 10 with inferior myocardial infarction) and 20 normal subjects. The cyclic variation of relative integrated backscatter (end-diastolic minus end-systolic) was calculated from on-line measurements. Standard parasternal long- and short-axis and apical four- and two-chamber views were obtained. In the anteroseptal myocardial infarction group, the cyclic variation of relative integrated backscatter was lower in the septum (1.5 +/- 1.6 dB) than in the posteroinferior wall (3.2 +/- 1.2 dB); however, the sample size of only three patients (of eight patients imaged) in the latter group prevented statistical comparison. The cyclic variation of relative integrated backscatter in the infarcted septum was less than the measurement obtained in the septum of the control group (4.3 +/- 2.4 dB, p less than 0.05). In the inferior infarction group, the cyclic variation of integrated backscatter in the posteroinferior wall (1.8 +/- 1.7 dB) was not significantly different from the measurement obtained in the septum (3.7 +/- 3.6 dB); however, the cyclic variation in the posteroinferior wall was significantly less than that obtained in the control group posteroinferior wall (5.7 +/- 1.7 dB, p less than 0.05).(ABSTRACT TRUNCATED AT 250 WORDS)

Ultrasonic integrated backscatter two-dimensional imaging: evaluation of M-mode guided acquisition and immediate analysis in 55 consecutive patients

We have shown previously that cardiac cycle-dependent integrated backscatter characterizes the physical state of myocardium in patients with ischemic heart disease and cardiomyopathy. In the present study the clinical applicability of M-mode guided two-dimensional integrated backscatter imaging was defined in evaluation of 55 nonselected patients. The mean amplitude of cyclic variation of integrated backscatter in normal segments (long-axis view) was as follows: basal septum, 4.2 +/- 1.3 dB (mean +/- SD; n = 27), mid-septum, 4.5 +/- 1.0 dB (n = 26), basal posterior, 4.8 +/- 1.0 dB (n = 30), and mid-posterior, 4.8 +/- 1.2 decibels (n = 27). The respective mean delay values (R wave to nadir) were as follows: 0.89 +/- 0.09, 0.84 +/- 0.09, 0.86 +/- 0.09, and 0.85 +/- 0.12. At least one cardiac cycle could be analyzed fully in 62% of patients. Limitations included technically difficult two-dimensional echocardiography, inadequate M-line orientation, technically remediable errors, or poor quality integrated backscatter images. In abnormal segments (n = 13) cyclic variation was reduced and delay was prolonged (1.2 +/- 1.1 dB and 1.21 +/- 1.1, respectively). Intraobserver and interobserver variability for amplitude measurements were modest, with respective correlation coefficients of r = 0.93; r = 0.72. The findings demonstrate that M-mode--assisted integrated backscatter is a practical approach for characterization of regional myocardial properties promptly and at the bedside in a large portion of patients with cardiac disease.

Serial measurement of integrated ultrasonic backscatter in human cardiac allografts for the recognition of acute rejection

Cyclic variation of integrated ultrasonic backscatter (IB) was noninvasively measured in the septum and left ventricular posterior wall using a quantitative IB imaging system to assess the alterations in the acoustic properties of myocardium associated with acute cardiac allograft rejection. The study population consisted of 23 cardiac allograft recipients and 18 normal subjects. In each cardiac allograft recipient, one to eight (mean, four) IB studies were performed, each within 24 hours of right ventricular endomyocardial biopsy performed for rejection surveillance. The magnitude of the cyclic variation of IB in the posterior wall was 5.9 +/- 0.9 dB in normal subjects and 6.2 +/- 1.3 dB in the cardiac allograft recipients without previous or current histological evidence of acute rejection (n = 17, p = NS vs. normal subjects). The magnitude of cyclic variation of IB in the septum was 4.8 +/- 1.1 dB in normal subjects and 3.8 +/- 2.0 dB in the cardiac allograft recipients (n = 15, p = NS vs. normal subjects). A significant decrease in the septal IB measure was observed in cardiac allograft recipients with left ventricular hypertrophy (wall thickness of at least 13 mm) (2.6 +/- 1.7 dB, n = 8, p less than 0.05 vs. normal subjects). IB studies were done before and during moderate acute rejection in 11 recipients (14 episodes). During moderate acute cardiac rejection, the magnitude of the cyclic variation in IB decreased from 6.7 +/- 1.3 to 5.1 +/- 1.4 dB in the posterior wall (n = 14, p less than 0.05) and from 4.2 +/- 2.1 dB to 2.9 +/- 1.8 dB in the septum (n = 12, p less than 0.05). These data suggest 1) the magnitude of the cyclic variation in IB of the septum is different in cardiac allografts with cardiac hypertrophy and normal subjects, possibly reflecting regionally depressed myocardial contractile performance and 2) acute cardiac rejection in humans is accompanied by an alteration in the acoustic properties of the myocardium. This change is detectable by serial measurement of the magnitude of the cyclic variation in IB, both in the septum and in the posterior wall.

Ultrasonic tissue characterization with a real time integrated backscatter imaging system in normal and aging human hearts

Experimental studies have shown that variation in the magnitude of integrated ultrasonic backscatter during the cardiac cycle represents acoustic properties of myocardium that are affected by pathologic processes; however, there are few clinical studies using integrated backscatter. Forty subjects without cardiovascular disease (aged 22 to 71 years, mean 41) were studied with use of a new M-mode format integrated backscatter imaging system to characterize the range of cyclic variation of integrated backscatter in normal subjects. Cyclic variation in integrated backscatter was noted in both the septum and the posterior wall in all subjects. The magnitude of the cyclic variation of integrated backscatter and the interval from the onset of the QRS wave of the electrocardiogram to the minimal integrated backscatter value were measured using an area of interest of variable size for integrated backscatter sampling and a software resident in the ultrasound scanner. The magnitude of cyclic variation was larger for the posterior wall than for the septum (6.3 +/- 0.8 versus 4.9 +/- 1.3 dB, p less than 0.01). The interval to the minimal integrated backscatter value was 328 +/- 58 ms for the septum and 348 +/- 42 ms for the posterior wall (p = NS). There was a weak correlation between the magnitude of cyclic variation of integrated backscatter and subject age for the posterior wall (r = -0.47, p less than 0.01), but this was not significant for the septum (r = -0.21) (partially because of inability to exclude specular septal echoes) and septal endocardium.(ABSTRACT TRUNCATED AT 250 WORDS)

Ultrasonic tissue characterization of human hypertrophied hearts in vivo with cardiac cycle-dependent variation in integrated backscatter

Integrated ultrasonic backscatter (IB) is a noninvasive measure of the acoustic properties of myocardium. Previous experimental studies have indicated that altered acoustic properties of the myocardium are reflected by the magnitude of variation of IB during the cardiac cycle. In our study, cardiac cycle-dependent variation of IB was noninvasively measured using a quantitative IB imaging system in 12 patients with uncomplicated pressure-overload hypertrophy and 13 patients with hypertrophic cardiomyopathy. Sixteen normal subjects served as a control. The magnitude of cardiac cycle-dependent variation of IB for the posterior wall was 6.0 +/- 0.9 dB in normal subjects, 5.7 +/- 0.8 dB in the patients with uncomplicated pressure-overload hypertrophy, and 6.7 +/- 2.1 dB in the patients with hypertrophic cardiomyopathy. There were no significant differences among any of these groups. In contrast, the magnitude of cardiac cycle-dependent variation of IB for the septum was significantly smaller in the patients with uncomplicated pressure-overload hypertrophy (2.8 +/- 1.3 dB) and in the patients with hypertrophic cardiomyopathy (3.1 +/- 2.3 dB) than in normal subjects (4.9 +/- 1.0 dB). The magnitude of cardiac cycle-dependent variation of IB was smaller as the wall-thickness index increased (r = -0.53, p less than 0.01, n = 82 for all data). This IB measure also correlated with percent-systolic thickening of the myocardium (r = 0.67, p less than 0.01, n = 82). Thus, alteration in the magnitude of cardiac cycle-dependent variation of IB was observed in hypertrophic hearts and showed apparent regional myocardial differences.

Ultrasonic tissue characterization with integrated backscatter. Acute myocardial ischemia, reperfusion, and stunned myocardium in patients

We have previously shown in studies of experimental animals that myocardium exhibits a cardiac cycle-dependent variation of integrated backscatter that reflects regional myocardial contractile performance and that is blunted promptly after arterial occlusion and recovers after reperfusion. To define the clinical utility of ultrasonic tissue characterization with integrated backscatter for detection of acute myocardial infarction and reperfusion, 21 patients (14 men and seven women) were studied in the cardiac care unit within the first 24 hours (mean time, 11.3 hours; range, 3.5-23.8 hours) after the onset of symptoms indicative of acute myocardial infarction with conventional two-dimensional and M-mode echocardiography and with analysis of integrated backscatter. The magnitude of cyclic variation of integrated backscatter was measured from several sites within acute infarct regions and normal regions remote from the infarct zone for each patient. The average magnitude of cyclic variation among all patients (n = 21) was 4.8 +/- 0.5 dB in normal regions compared with 0.8 +/- 0.3 dB in infarct regions (p less than 0.05) within the first 24 hours after the onset of symptoms. Among the patients who had two studies, 15 (mean, 7.1 days; range, 2-31 days for second study) underwent coronary arteriography to define vessel patency. In patients with vessels with documented patency (n = 10), the magnitude of cyclic variation in infarct regions increased over time from 1.3 +/- 0.6 to 2.5 +/- 0.5 dB from the initial to final study (p less than 0.05). Patients with occluded infarct-related arteries (n = 5) exhibited no significant recovery of cyclic variation (0.3 +/- 0.3-0.6 +/- 0.3 dB). A blinded analysis of standard two-dimensional echocardiographic images revealed no significant recovery of wall thickening in either group over the same time intervals. Ultrasonic tissue characterization promptly detects acute myocardial infarction and may delineate potential beneficial effects of coronary artery reperfusion manifest by restoration of cyclic variation of integrated backscatter in the presence of severe wall motion abnormalities.

Early identification with ultrasonic integrated backscatter of viable but stunned myocardium in dogs

It has been shown that canine and human hearts exhibit a cardiac cycle-dependent variation of integrated backscatter (cyclic variation) that reflects intrinsic regional contractile performance. To determine whether ultrasound tissue characterization can identify viable though stunned myocardium before recovery of regional wall thickening, transient ischemic injury was produced in eight open chest dogs for 15 min followed by reperfusion for 2 h. Cyclic variation and wall thickening were measured before ischemia, at 15 min after the onset of ischemia and at selected intervals after the onset of reperfusion from multiple sites within the ischemic zone with a novel combined two-dimensional and M-mode acquisition system. Cyclic variation and wall thickening were computed from digitized M-mode integrated backscatter images with an algorithm developed and validated for this purpose. Magnitude and "delay" of cyclic variation and wall thickening were compared. Delay represents the degree of synchrony of regional cyclic variation or wall thickening with global ventricular mechanical systole. Baseline cyclic variation and wall thickening magnitudes were 3.8 +/- 0.2 dB and 37 +/- 1.4%, respectively. With ischemia, cyclic variation and wall thickening decreased to 1.7 +/- 0.2 dB and 17 +/- 2%, respectively (p less than 0.05, compared with baseline). Cyclic variation recovered to baseline levels within 20 min after reperfusion (3.3 +/- 0.4 dB, p = NS). Wall thickening remained depressed for 2 h after the onset of reperfusion (23 +/- 2%, p less than 0.05 compared with baseline). Delay of cyclic variation in a unitless ratio expressed as delay (in milliseconds) divided by the QT interval (in milliseconds) increased from 0.87 +/- 0.03 at baseline to 1.10 +/- 0.12 with ischemia, a change consistent with mild asynchrony, and returned to baseline (0.95 +/- 0.07, p = NS compared with baseline) within 20 min after reperfusion. Delay of wall thickening was 0.88 +/- 0.02 at baseline, increased to 1.23 +/- 0.09 with ischemia and remained significantly increased 2 h after reperfusion (1.07 +/- 0.05, p less than 0.05 compared with baseline). Recovery time constants for cyclic variation and wall thickening with reperfusion reflected earlier restoration of cyclic variation (8.1 min) than of wall thickening (420.5 min). Thus, cyclic variation recovers before wall thickening with reperfusion. Its analysis appears to provide a useful index of the presence of viable and potentially salvageable tissue in regions of stunned myocardium that is independent of wall thickening.