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.

Quantitative assessment of myocardial ultrasound tissue characterization through receiver operating characteristic analysis of Bayesian classifiers.

OBJECTIVES: This work proposes a self-consistent assessment methodology for quantitative evaluation of any combination of diagnostic features, with the immediate goal of quantitatively assessing the discriminating power in diabetic patients of features derived from ultrasound backscatter from myocardium. BACKGROUND: Four features from analysis of left ventricular myocardial ultrasound backscatter have previously been shown to be sensitive to potentially cardiomyopathic changes in patients with insulin-dependent diabetes mellitus who have no overt heart disease. The measured features were significantly different between such patients and normal control subjects, as well as among groups of such patients with and without systemic complications of the disease. The quantitative discriminating potential of the features was not assessed. METHODS: Multivariate classifier functions were constructed and analyzed by using the methodology of the receiver operating characteristic curve, which allows quantitative assessment of the discriminating power of these features, alone or in combination. The area under the receiver operating characteristic curve--the true positive rate averaged over all false positive rates--was used as a summary measure of performance. RESULTS: In distinguishing patients with insulin-dependent diabetes mellitus from normal control subjects, the most discriminating combination of ultrasound features for the detection of such changes in these patients yielded receiver operating characteristic curves with area measures of approximately 0.80; for such patients with retinopathy the measure increased to 0.90. This performance is comparable to that of many commonly used diagnostic tests. CONCLUSIONS: A self-consistent set of evaluation methodologies has quantitatively demonstrated the sensitivity of four ultrasound backscatter features to otherwise latent changes in myocardial structure that accompany the evolution of insulin-dependent diabetes mellitus. The results are remarkable in themselves and suggest the potential of the features for the general field of cardiac ultrasound tissue characterization.

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

To determine whether quantitative ultrasound tissue characterization differentiates normal myocardial regions from segments of remote infarction, 32 consecutive patients with a diagnosis of previous myocardial infarction were evaluated. Images were obtained in real time with a modified two-dimensional ultrasound system capable of providing continuous signals in proportion to the logarithm of integrated backscatter along each A line. In 15 patients, adequate parasternal long-axis images that delineated both normal and infarct segments were obtained with standard time-gain compensation. Image data were analyzed to yield both magnitude and delay (electrocardiographic R wave to nadir normalized for the QT interval) of the cyclic variation of backscatter. Cyclic variation was present in 55 of 56 normal myocardial sites, averaging (mean +/- SEM) 3.2 +/- 0.2 dB in magnitude and exhibiting a mean normalized delay of 0.87 +/- 0.03. The magnitude of cyclic variation in infarct segments was significantly reduced to 1.1 +/- 0.2 dB (42 sites), and the delay was markedly increased to 1.47 +/- 0.12 (21 sites) (p less than 0.0001 for both). In 20 of 42 infarct sites, no cyclic variation was detectable. Thus, ultrasound tissue characterization quantitatively differentiated infarct segments from normal myocardium in patients with remote myocardial infarction.

Methods for estimation of statistical properties of envelopes of ultrasonic echoes from myocardium.

Several investigators have characterized various forms of heart disease from the statistical properties of envelopes of ultrasonic echos from myocardium. In particular, the mean-to-standard deviation ratio (MSR), skewness, and kurtosis of the envelope probability density function have been used for the detection of myocardial ischemia, infarction, reperfusion, and hypertrophy. In this paper, the effects of phenomena other than tissue acoustic properties upon estimates of statistical parameters are investigated. These include system characteristics (center frequency, bandwidth, beam width, etc.), sample volume dimensions, and tissue velocity. In myocardium, relatively small amounts of tissue are available for interrogation. It is shown that, under these limited data acquisition conditions, substantial systematic biases in the estimates of statistical parameters may occur. Analytic forms for errors in the envelope variance estimate are derived. Estimation of the envelope mean, variance, MSR, skewness, and kurtosis is investigated experimentally, using a commercial medical ultrasound scanner and a tissue-mimicking phantom.

High resolution ultrasonic backscatter coefficient estimation based on autoregressive spectral estimation using Burg's algorithm.

An autoregressive (AR) method for spectral estimation was applied toward the task of estimating ultrasonic backscatter coefficients from small volumes of tissue. High spatial resolution is desirable for generating images of backscatter coefficient. Data was acquired from a homogeneous tissue-mimicking phantom and from a normal human liver in vivo. The AR method was much more resistant to gating artifacts than the traditional DFT (discrete Fourier transform) approach. The DFT method consistently underestimated backscatter coefficients at small gate lengths. Therefore backscatter coefficient image formation will be quantitatively more meaningful if based on AR spectral estimation rather than the DFT. The autoregressive method offers promise for enhanced spatial resolution and accuracy in ultrasonic tissue characterization and nondestructive evaluation of materials.

Constrained reconstruction applied to 2-D chemical shift imaging.

The method of constrained reconstruction, previously applied to magnetic resonance imaging (MRI), is extended to magnetic resonance spectroscopy. This method assumes a model for the MR signal. The model parameters are estimated directly from the phase encoded data. This process obviates the need for the fast Fourier transform (FFT) (which often exhibits limited resolution and ringing artifact). The technique is tested on simulated data, phantom data, and data acquired from human liver in vivo. In each case, constrained reconstruction offers spatial resolution superior to that obtained with the FFT.

Measurements of phase velocity and group velocity in human calcaneus.

Ultrasonic velocity in calcaneus correlates highly with bone mineral density, which is a good predictor of osteoporotic fracture risk. Several commercial bone sonometers perform a velocity measurement based on the transit time of a broadband pulse to assess skeletal status. This approach is somewhat problematic, however, because several authors have reported ambiguities in measurements in calcaneus. Phase velocity is an alternative that may be less dependent on device spectral characteristics. In addition, dispersion (the frequency-dependence of phase velocity) is a fundamental property worth investigating to increase understanding of interaction between ultrasound and bone. To compare two group-velocity measurement methods and one phase-velocity measurement method, a polycarbonate sample (for method validation) and 24 human calcanei were investigated in vitro. Phase velocity in calcaneus at 500 kHz was 1511 m/s +/- 30 m/s (mean +/- standard deviation). Average phase velocity decreased approximately linearly with frequency (-18 m/s MHz). The two group velocity measurements were comparable and tended to be slightly lower than phase velocity. The magnitude of dispersion showed little correlation with bone mineral density.

Temperature dependence of ultrasonic attenuation in human calcaneus.

Ultrasonic attenuation in calcaneus has been shown to be a useful measurement for the diagnosis of osteoporosis. Several studies indicate that this measurement is affected by temperature fluctuations, although the fundamental causes for this are currently not well understood. To investigate this phenomenon, six defatted human calcanei were interrogated in vitro at six temperatures ranging from 10 degrees C to 40 degrees C. The temperature-related variation was -0.18 dB/cm MHz degrees C (95% confidence interval: -0.27 dB/cm MHz degrees C, -0.10 dB/cm MHz degrees C). This study reinforces the notion, advanced by other investigators, that temperature-related effects need to be taken into account when performing diagnostic measurements that require high precision (such as monitoring responses to drug intervention), will aid in the interpretation of in vivo experiments designed to investigate temperature-dependent precision limitations, will facilitate comparisons between in vitro studies normally carried out at room temperature with in vivo studies carried out at body temperature, and fills a gap in the compendium of measurements of temperature-dependences of acoustic properties of biologic tissues.

Assessment of bone density using ultrasonic backscatter.

The goal of this project was to investigate the utility of ultrasonic backscatter for the assessment of bone status. Ultrasound offers a low-cost, portable, nonionizing alternative or complement to common X-ray- or radioisotope (gamma ray)-based methods of bone densitometry. Ultrasonic backscatter may provide useful information not revealed by ultrasonic attenuation and sound-speed densitometers. Backscatter is sensitive to microstructural variations in acoustic impedance and should therefore provide information regarding architecture (which is related to fracture risk), as well as density. Ultrasonic backscatter at 2.25 MHz and CT bone densitometric data have been acquired from 10 healthy human volunteers. The degree of correlation between CT and ultrasonic backscatter is high (r = 0.87, p < 0.001). The envelope signal-to-noise ratio was 1.81 +/- 0.08 (mean +/- standard deviation). This suggests that the number of scatterers per resolution cell is large, the radiofrequency signal approximately obeys circular Gaussian statistics, and the envelope obeys Rayleigh statistics. These results indicate promise for ultrasonic backscatter as a substitute for or an adjunct to other ultrasonic measurements (attenuation and sound speed) and X-ray measurements for the assessment of bone status.

Relationships of ultrasonic backscatter with ultrasonic attenuation, sound speed and bone mineral density in human calcaneus.

Ultrasonic attenuation and sound speed have been investigated in trabecular bone by numerous authors. Ultrasonic backscatter has received much less attention. To investigate relationships among these three ultrasonic parameters and bone mineral density (BMD), 30 defatted human calcanei were investigated in vitro. Normalized broadband ultrasonic attenuation (nBUA), sound speed (SOS), and logarithm of ultrasonic backscatter coefficient (LBC) were measured. Bone mineral density was assessed using single-beam dual energy x-ray absorptiometry (DEXA). The correlation coefficients of least squares linear regressions of the three individual ultrasound (US) parameters with BMD were 0.84 (nBUA), 0.84 (SOS) and 0.79 (LBC). The 95% confidence intervals for the correlation coefficients were 0. 69-0.92 (nBUA), 0.68-0.92 (SOS) and 0.60-0.90 (LBC). The correlations among pairs of US variables ranged from 0.63-0.79. Variations in nBUA accounted for r(2) = 62% of the variations in LBC. Variations in SOS accounted for r(2) = 40% of the variations in LBC. These results suggest that ultrasonic backscattering properties may contain substantial information not already contained in nBUA and SOS. A multiple regression model including all three US variables was somewhat more predictive of BMD than a model including only nBUA and SOS.

Myocardial ultrasonic backscatter for characterization of ischemia and reperfusion: relationship to wall motion.

We have previously shown that cardiac cycle-dependent variation of integrated backscatter occurs in normal myocardium. To determine whether myocardial ischemia and reperfusion can be distinguished by real-time integrated backscatter imaging we performed 10 min balloon occlusion of the Left Anterior Descending (LAD) coronary artery followed by reperfusion in 10 closed-chest anesthetized dogs. Images were obtained at baseline, during occlusion, and up to 120 min after reperfusion. We measured the magnitude and delay of cyclic variation of integrated backscatter in segments with and without asynergy. Radiolabeled microspheres were used to verify both ischemia and reperfusion. Ischemic segments exhibited decreased magnitude and increased normalized delay of cyclic variation of integrated backscatter (from 3.3 +/- 0.3 dB to 1.4 +/- 0.2 dB, mean +/- SE; and from 0.95 +/- 0.03 to 1.67 +/- 0.15, respectively, all p less than or equal to 0.001). Reperfusion promptly restored the magnitude of cyclic variation toward normal. However, the delay of the cyclic variation was restored only partially. Wall motion analysis of the ischemic sites revealed persistent abnormalities throughout the reperfusion interval despite return to normal of the magnitude and delay of cyclic variation. Thus, real-time integrated backscatter imaging permits detection and differentiation of changes in myocardial acoustic properties indicative of ischemia and of subsequent reperfusion.

The effects of frequency-dependent attenuation and dispersion on sound speed measurements: applications in human trabecular bone.

Sound speed may be measured by comparing the transit time of a broadband ultrasonic pulse transmitted through an object with that transmitted through a reference water path. If the speed of sound in water and the thickness of the sample are known, the speed of sound in the object may be computed. To measure the transit time differential, a marker such as a zero-crossing, may be used. A sound speed difference between the object and water shifts all markers backward or forward. Frequency-dependent attenuation and dispersion may alter the spectral characteristics of the waveform, thereby distorting the locations of markers and introducing variations in sound-speed estimates. Theory is derived to correct for this distortion for Gaussian pulses propagating through linearly attenuating, weakly dispersive media. The theory is validated using numerical analysis, measurements on a tissue mimicking phantom, and on 24 human calcaneus samples in vitro. Variations in soft tissue-like media are generally not exceptionally large for most applications but can be substantial, particularly for high bandwidth pulses propagating through media with high attenuation coefficients. At 500 kHz, variations in velocity estimates in bone can be very substantial, on the order of 40 to 50 m/s because of the high attenuation coefficient of bone. In trabecular bone, the effects of frequency-dependent attenuation are considerable, and the effects of dispersion are negligible.

A stratified model to predict dispersion in trabecular bone.

Frequency-dependent phase velocity (dispersion) has previously been measured in trabecular bone by several groups. In contrast to most biologic tissues, phase velocity in trabecular bone tends to decrease with frequency. A stratified model, consisting of alternating layers of bone and marrow (in vivo) or water (in vitro), has been employed in an attempt to explain this phenomenon. Frequency-dependent phase velocity was measured from 300 to 700 kHz in 1) phantoms consisting of regularly spaced thin parallel layers of polystyrene sheets in water and 2) 30 calcaneus samples in vitro. For the polystyrene phantoms, the agreement between theory and experiment was good. For the calcaneus samples, the model has some limited usefulness (uncertainty of about 5%) in predicting average phase velocity. More importantly, the model seems to perform consistently well for predicting the frequency dependence of phase velocity in calcaneus.

Ultrasonic attenuation in human calcaneus from 0.2 to 1.7 MHz.

Ultrasonic attenuation has been demonstrated to be a useful measurement in the diagnosis of osteoporosis. Most studies have employed ultrasound in a range of frequencies from about 200 kHz-300 kHz to 600 kHz-1 MHz, and many have assumed a linear dependence of attenuation on frequency. In order to investigate the attenuation properties of human calcaneus at higher frequencies, 16 defatted human calcanea were interrogated in vitro using two matched pairs of transducers with center frequencies of 500 kHz and 2.25 MHz. The linear dependence of attenuation on frequency seems to extend up to at least 1.7 MHz. The correlation between attenuation coefficient and frequency from 400 kHz to 1.7 MHz was r = 0.999 (95% confidence interval, CI, = 0.998-1.00). The measurements suggest that some deviations from linear frequency dependence of attenuation may occur at lower frequencies (below 400 kHz), however.

The relationship between ultrasonic backscatter and bone mineral density in human calcaneus.

Backscatter and attenuation coefficients were measured from 24 human calcanei in vitro. The logarithm of the backscatter coefficient at 500 kHz showed moderate correlations with bone mineral density (r=0.81, 95% confidence interval: 0.59-0.91) and attenuation (r=0.79, 95% CI: 0.56-0.91). These results suggest that backscatter measurements may be useful in the diagnosis of osteoporosis.

Application of autoregressive spectral analysis to cepstral estimation of mean scatterer spacing.

The problem of estimation of mean scatterer spacing in an object containing regularly spaced structures is addressed. An autoregressive (AR) spectral estimation method is compared with a conventional fast Fourier transform (FFT)-based approach for this task. Regularly spaced structures produce a periodicity in the power spectrum of ultrasonic backscatter. This periodicity is manifested as a peak in the cepstrum. A phantom was constructed for comparison of the two methods. It contained regularly spaced nylon filaments. It also contained randomly positioned glass spheres that produced incoherent backscatter. In an experiment in which this target was interrogated using broadband ultrasound, the AR spectral estimate offered considerable improvement over the FFT when the analysis gate length was on the order of the structural dimension. Advantages included improved resolution, reduction in bias and variance of scatterer spacing estimates, and greater resistance to ringing artifacts. Data were also acquired from human liver in vivo. AR spectral estimates on human data exhibited a decreased dependence on gate length. These results offer promise for enhanced spatial resolution and accuracy in ultrasonic tissue characterization and nondestructive evaluation of materials.

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.

Ultrasonic scattering from cancellous bone: a review.

This paper reviews theory, measurements, and computer simulations of scattering from cancellous bone reported by many laboratories. Three theoretical models (binary mixture, Faran cylinder, and weak scattering) for scattering from cancellous bone have demonstrated some consistency with measurements of backscatter. Backscatter is moderately correlated with bone mineral density in human calcaneus in vitro (r(2) = 0.66 - 0.68). Backscatter varies approximately as frequency cubed and trabecular thickness cubed in human calcaneus and femur in vitro. Backscatter from human calcaneus and bovine tibia exhibits substantial anisotropy. So far, backscatter has demonstrated only modest clinical utility. Computer simulation models have helped to elucidate mechanisms underlying scattering from cancellous bones.