Ultrasound myocardial integrated backscatter signal processing: frequency domain versus time domain

Modeling of errors in Nakagami imaging: illustration on breast mass characterization

Nakagami imaging is an attractive tissue characterization method, as the parameter estimated at each location is related to properties of the tissues. The application to clinical ultrasound images is problematic, as the estimation of the parameters is disturbed by the presence of complex structures. We propose to consider separately the different aspects potentially affecting the value of the Nakagami parameters and quantify their effects on the estimation. This framework is applied to the classification of breast masses. Quantitative parameters are computed on two groups of ultrasound images of benign and malignant tumors. A statistical analysis of the result indicated that the previously observed difference between average values of the Nakagami parameters is explained mostly by estimation errors. In the future, new methods for reliable computation of Nakagami parameters need to be developed, and factors of error should be considered in studies using Nakagami parameters.

Ultrasound image segmentation and tissue characterization

Ultrasound image segmentation deals with delineating the boundaries of structures, as a step towards semi-automated or fully automated measurement of dimensions or for characterizing tissue regions. Ultrasound tissue characterization (UTC) is driven by knowledge of the physics of ultrasound and its interactions with biological tissue, and has traditionally used signal modelling and analysis to characterize and differentiate between healthy and diseased tissue. Thus, both aim to enhance the capabilities of ultrasound as a quantitative tool in clinical medicine, and the two end goals can be the same, namely to characterize the health of tissue. This article reviews both research topics, and finds that the two fields are becoming more tightly coupled, even though there are key challenges to overcome in each area, influenced by factors such as more open software-based ultrasound system architectures, increased computational power, and advances in imaging transducer design.

LV segmentation through the analysis of radio frequency ultrasonic images

LV segmentation is often an important part of many automated cardiac diagnosis strategies. However, the segmentation of echocardiograms is a difficult task because of poor image quality. In echocardiography, we note that radio-frequency (RF) signal is a rich source of information about the moving LV as well. In this paper, first, we will investigate currently used, important RF derived parameters: integrated backscatter coefficient (IBS), mean central frequency (MCF) and the maximum correlation coefficients (MCC) from speckle tracking. Second, we will develop a new segmentation algorithm for the segmentation of the LV boundary, which can avoid local minima and leaking through uncompleted boundary. Segmentations are carried out on the RF signal acquired from a Sonos7500 ultrasound system. The results are validated by comparing to manual segmentation results.

Measurements of the anisotropy of ultrasonic attenuation in freshly excised myocardium

Echocardiography requires imaging of the heart with sound propagating at varying angles relative to the predominant direction of the myofibers. The degree of anisotropy of attenuation can significantly influence ultrasonic imaging and tissue characterization measurements in vivo. This study quantifies the anisotropy of attenuation of freshly excised myocardium at frequencies typical of echocardiographic imaging. Results show a significantly larger anisotropy than previously reported in specimens of locally unidirectional myofibers. Through-transmission radio frequency-based measurements were performed on specimens from 12 ovine and 12 bovine hearts. Although ovine hearts are closer in size to human, the larger size of bovine hearts offers the potential for specimens in which myofibers are more nearly unidirectionally aligned. The attenuation coefficient increased approximately linearly with frequency. The mean slope of attenuation with frequency was 3-4 times larger for propagation parallel than for perpendicular to the myofibers. At perpendicular insonification, slopes between ovine and bovine myocardium were approximately equal. However, attenuation in bovine specimens was larger for angles approaching parallel. The difference in results for parallel appears consistent with what might be expected from increased myofiber curvature associated with smaller lamb hearts. Quantitative knowledge of anisotropy of attenuation may be useful in understanding mechanisms underlying the interaction of ultrasound with myocardium.

A parametric imaging approach for the segmentation of ultrasound data

When an ultrasonic examination is performed, a segmentation tool would often be very useful, either for the detection of pathologies, the early diagnosis of cancer or the follow-up of the lesions. Such a tool must be both reliable and accurate. However, because of the relatively reduced quality of ultrasound images due to the speckled texture, the segmentation of ultrasound data is a difficult task. We have previously proposed to tackle the problem using a multiresolution Bayesian region-based algorithm. For computation time purposes, a multiresolution version has been implemented. In order to improve the quality of the segmentation, we propose to perform the segmentation not only from the envelope image but to combine more information about the properties of the tissues in the segmentation process. Several acoustical parameters have thus been computed, either directly from the images or from the radio-frequency (RF) signal. In a previous study, two parametric images were involved in the segmentation process. The parameter represented the integrated backscatter (IBS) and the mean central frequency (MCF), which is a measurement related to the attenuation of ultrasound waves in the media. In this study, parameters representative of the scattering conditions in the tissue are evaluated in the multiparametric segmentation process. They are extracted from the K-distribution (alpha,b) and the Nakagami distribution (m,Omega) and are related to the local density of scatterers (alpha,m), the size of the scatterers (b) and the backscattering properties of the medium (Omega). The acoustical features are calculated locally on a sliding window. This procedure allows to built parametric mapping representing the particular characteristics of the medium. To test the influence of the acoustical parameters in the segmentation process, a set of numerical phantoms has been computed using the Field software developed by J.A. Jensen. Each phantom consists in two regions with two different acoustical properties: the density of scatterers and the scattering amplitude. From both the simulated RF signals and envelope images, the parameters have been computed; their relevance to represent a particular characteristic of the medium is evaluated. The segmentation has been processed for each phantom. The ability of each parameter to improve the segmentation results is validated. A agar-gel phantom has also been created, in order to test the accuracy of the parameters in conditions closer to the in vivo ultrasound imaging. This phantom contains four inclusions with different concentrations of silica. A B&K ultrasound device provides the RF data. The quantification of the segmentation quality is based on the rate of correctly classified pixels and it has been computed for all the parameters either from the field images or the phantom images. The large improvement in the segmentation results obtained reveals that the multiparametric segmentation scheme proposed in this study can be a reliable tool for the processing of noisy ultrasound data.

Ultrasonic myocardial tissue characterization in patients with hypertrophic cardiomyopathy and pressure-overloaded hypertrophy by backscattered energy temporal analysis

This study measured integrated backscatter (IB) values in the subendocardium and subepicardium of patients with hypertrophy using the newly developed Backscattered Energy Temporal Analysis (BETA) system, and evaluated the differences of acoustic properties according to etiology. Twenty-one patients with hypertrophic cardiomyopathy (HCM), 16 with pressure-overloaded hypertrophy (POH), and 21 controls were studied. M-mode formatted IB images were obtained using BETA and the region of interest (ROI), automatically divided into epicardial and endocardial halves of the myocardium, was placed in the ventricular septum and posterior wall. Values for the cyclic variation of IB (CVIB) in the entire ROI and in each half of the ROI were obtained. CVIB significantly decreased in the ventricular septum in HCM and POH compared with normal subjects, but there were no significant differences between HCM and POH. In the posterior wall, the CVIB was less in the subendocardium than in the epicardium in POH, and was also less than in normal subjects (7.0+/-1.7 dB vs 8.6+/-1.9 dB and 8.8+/-2.1 dB, p<0.05, respectively). Separate ultrasonic tissue characterization of the subendocardium and subepicardium provides further etiological information of various heart diseases.

Discrimination of early/intermediate and advanced/complicated coronary plaque types by radiofrequency intravascular ultrasound analysis

Radiofrequency intravascular ultrasound (IVUS-RF) analysis, as an extension of conventional IVUS imaging, may provide more accurate plaque discrimination. Thirty-two autopsy atherosclerotic coronary arteries were investigated. Corresponding sectors in different plaques were matched by histologic and RF analysis. Histologic analysis utilized the American Heart Association plaque classification. The backscattered ultrasound RF signal was analyzed by fast-Fourier transform, providing the underlying frequency components of its power spectrum. The normalized backscattered signal power (in decibels [dB]) for frequencies between 15.3 and 40.3 MHz was then measured for plaque discrimination. Advanced/complicated plaque types showed a higher signal power at all frequencies than early/intermediate lesion types (p <0.001 to p = 0.005). Discrimination of advanced/complicated lesion types was best at 15.3 MHz, with a cut-off point of 2.5 dB (sensitivity 93%, specificity 79%), and second best at 17.6 MHz (sensitivity 87%, specificity 71%, cut-off point 1.9 dB). With conventional IVUS, plaque discrimination was weaker; the best sensitivity for diagnosing early/intermediate lesion types was reached for "soft plaque" (sensitivity 63%, specificity 73%). Compared with conventional IVUS, IVUS-RF can discriminate between advanced/complicated and early/intermediate coronary atherosclerotic lesions with relatively high sensitivity and specificity in vitro.

Is there a change in myocardial nonlinearity during the cardiac cycle?

The distortion of a sound wave during propagation results in progressive transfer of the energy from fundamental to higher harmonics, and is dependent on the nonlinearity of the medium. We studied if relative changes in acoustical nonlinearity occur in healthy myocardium during the cardiac cycle. Radiofrequency data were acquired from transthoracic echocardiography (2.5 and 3.5 MHz), parasternal long axis view, from five dogs and nine healthy volunteers. Integrated backscatter was calculated after filtering for fundamental (FIB) and second harmonic frequencies (SHIB), from a region in the posterior myocardial wall. The results suggest that there is little difference between the SHIB and FIB, although there were large variations between individuals. The maximal changes in nonlinearity, as estimated by SHIB/FIB ratio, mostly occurred during systole. SHIB presented similar cyclic variation with FIB (p = NS). Further studies are necessary to separate the role of myocardial nonlinearity, attenuation, propagating distance, or acoustical properties of the blood. The results are important in further tissue characterization studies employing second harmonic data.

A multiparametric and multiresolution segmentation algorithm of 3-D ultrasonic data

An algorithm devoted to the segmentation of 3-D ultrasonic data is proposed. The algorithm involves 3-D adaptive clustering based on multiparametric information: the gray-scale intensity of the echographic data, 3-D texture features calculated from the envelope data, and 3-D tissue characterization information calculated from the local frequency spectra of the radio-frequency signals. The segmentation problem is formulated as a Maximum A posterior (MAP) estimation problem. A multi-resolution implementation of the algorithm is proposed. The approach is tested on simulated data and on in vivo echocardiographic 3-D data. The results presented in the paper illustrate the robustness and the accuracy of the proposed approach for the segmentation of ultrasonic data.

Are changes in myocardial integrated backscatter restricted to the ischemic zone in acute induced ischemia? An in vivo animal study

Integrated backscatter (IB) from a myocardial region, calculated from radiofrequency echocardiographic data, has been proposed as a useful parameter for investigating changes in myocardial tissue induced by ischemia. In 10 closed-chest dogs, 5 minutes of myocardial ischemia was induced by either a proximal occlusion of the circumflex coronary artery (CX) (5 dogs), resulting in extensive ischemia in the posterior wall, or by occluding the distal CX vessel (5 dogs) to produce a small localized ischemic zone in the posterior wall. High-resolution digital radiofrequency data from the whole left ventricular myocardium, in the imaging plane during one complete heart cycle, were acquired with a whole-image real-time acquisition approach. Regions in the septum and posterior wall (both ischemic tissue and, in the case of distal occlusions, tissue surrounding the ischemic zone) were chosen for analysis, and IB and cyclic variation (CV) of IB were calculated. Post occlusion, an increase in mean IB values was found in the ischemic segment. However, an increase in CV was also observed in the peri-ischemic zone for the distal CX occlusion and in the septum after proximal CX occlusion. These findings show that changes in CV are not restricted to the ischemic zone but may also occur in distal myocardium. This may be explained by changes in the regional contractile state and loading conditions of the "normal" myocardium, which are altered in response to the distal ischemia.

Evaluation of an experimental system for the in vitro assessment of ultrasonic contrast agents

The ultrasonic properties of microbubble contrast agents need to be fully understood if reproducible images and quantitative results are to be produced. Additional aspects of the physical and chemical environment into which the contrast agents are introduced also need to be taken into account, and their effect on contrast agent performance evaluated. A setup that provides an accurate and reproducible data-acquisition system is presented and evaluated in this paper. The linear range of this system is assessed, as well as its accuracy and precision. A new approach to the investigation of contrast agents, based on normalised backscatter, is discussed. Also, a common technique of degassing, widely used in other areas, is described and evaluated to determine its appropriateness to contrast agent studies.

Feasibility of in vivo intravascular ultrasound tissue characterization in the detection of early vascular transplant rejection

BACKGROUND

Unprocessed ultrasound radiofrequency (RF) signal analysis has been shown to distinguish different tissue structures more reliably than gray-scale interpretation of conventional ultrasound images.

METHODS AND RESULTS

The objective of this study was to test the feasibility of in vivo intravascular ultrasound (IVUS) RF signal analysis in an animal model of allograft rejection. Six cynomolgus monkeys underwent transplantation of 3-cm aortic allograft segments distal to the renal arteries from immunologically mismatched donors. IVUS imaging with a 30-MHz system was performed 84 to 105 days after the operation. RF signals were acquired from cross sections of the recipient and the allograft aortas in real time with a digitizer at 500 MHz with 8-bit resolution. Sixty-five cross sections and 68 regions of interest (31 in host aorta and 37 in allograft) were analyzed in the adventitial layer with a total number of 8568 vectors processed. For each region of interest, a weighted-average attenuation was calculated on the basis of the attenuation and length for each individual vector. Histological examination was performed at every cross section imaged by IVUS. When the gray-scale images of conventional IVUS scored by an independent observer were compared, no distinction between adventitia of the native aorta and allograft was possible. Analysis of the average RF backscatter power also showed no significant difference (70.32+/-3.55 versus 70.72+/-3.38 dB). However, the average attenuation of allografts was significantly lower than that of the host aortas (2.64+/-1.38 versus 4.02+/-1.16 dB/mm, P<0.001). Histology demonstrated a marked adventitial inflammatory response in all allografts, with no inflammation observed in the host aortas.

CONCLUSIONS

In vivo IVUS tissue characterization can be performed during routine imaging. In this model of transplant vasculopathy, RF attenuation measurements were more sensitive than visual or quantitative gray-scale analysis.

Robustness of integrated backscatter for myocardial tissue characterization

Integrated backscatter (IB) has been used for ultrasonic tissue characterization. To assess the potential variables in IB measurements, we performed both theoretical simulations and in vitro phantom measurements. First, we simulated data in which the scatterer position randomly was varied. IB values for the resulting images were calculated. Second, RF data from a tissue-mimicking phantom were acquired. Third, an adapted imaging approach, based on phase insensitivity, was evaluated. For both the simulations and phantom measurements, IB showed a standard deviation of +/-20%. These large deviations can be explained by variations in interference of signals and are not related to the state of the tissue. Small deviations in position of the scatterers resulted in important variations in IB. They must be taken into account and may limit the use of IB in cardiological applications. An improvement potentially can be obtained using phase insensitivity in new ultrasound processing schemes.

Influences of ultrasonic machine settings, transducer frequency and placement of region of interest on the measurement of integrated backscatter and cyclic variation

Integrated backscatter and its cyclic variation are potentially important parameters to discriminate normal from diseased myocardium. Cyclic variation of integrated backscatter is expected to be independent of machine settings. Backscatter images of swine hearts were taken using a two-dimensional backscatter system while acoustic power was varied at different time gain control (TGC) settings. Cyclic variation was measured in vivo with various acoustic power and TGC settings using different transducer frequencies. Three different regions were analyzed. For any given TGC setting, the relationship between acoustic power and integrated backscatter in vitro was linear only over a narrow range. In vivo, cyclic variation was present at all regions studied in both long- and short-axis views. However, lower acoustic power (< 15 dB) and TGC (< 20 dB), or excessive settings of acoustic power (> 35 dB) and TGC (> 50 dB), produced minimal cyclic variation. Appropriate acoustic power (20-35 dB) and TGC (30-50 dB) produced larger and more consistent cyclic, variation at the posterior region of the left ventricle. These data indicate that each region has specific, appropriate machine settings to maximize the magnitude of cyclic variation.

Influence of data processing on cyclic variation of integrated backscatter and wall thickness in stunned porcine myocardium

This study was performed to investigate the relationship between the cyclic variation of integrated backscatter and myocardial wall thickening in stunned myocardium. Different definitions of cyclic variation were evaluated to be able to compare with other studies. Ultrasound data were acquired from 10 open-chested Yorkshire pigs (25-33 kg) at baseline, during regional ischemia and during 30 min of stunning, using a broadband ultrasound transducer (3-7 MHz) sutured directly upon the left ventricular myocardial wall. Cyclic variation of integrated backscatter and myocardial wall thickening were calculated using three definitions obtained from the literature. Independent of the definition, cyclic variation of wall thickness and integrated backscatter were blunted during acute ischemia and returned transiently to or above baseline during the first minute of reperfusion, followed by a gradual decrease to a level under baseline during stunning. An early return of the cyclic variation of the integrated backscatter was not observed in pigs, independent of the data processing used. The relationship between integrated backscatter and wall thickness was maintained.

New technology in echocardiography II: imaging techniques

Images

Quantitative ultrasonic diagnosis of silicone breast implant rupture: an in vitro feasibility study

Ultrasound is useful in evaluating the integrity of silicone breast implants. However, extensive operator experience is required. A system for computer-assisted diagnosis is being developed to reduce operator dependence. Feasibility was examined by measuring the ultrasonic properties of breast implants in vitro. Silicone gels from 45 explanted implants (26 intact and 19 ruptured) were placed in sealed acoustic test chambers and 60 RF A-lines were acquired from each. Velocity of sound, attenuation and integrated backscatter (IB) were estimated. Receiver operating characteristic (ROC) analysis was performed. The mean speed of sound was 1060 m/s +/- 50.1 m/s in intact implants and 1115 m/s +/- 74.3 m/s in ruptured ones (p < 0.003). Differences in attenuation were not statistically significant. The mean IB was -83.9 dB +/- 7.94 dB in intact and -77.2 dB +/- 9.07 dB in ruptured implants (p < 0.006). The area under the ROC curve (Az) was 0.70 and 0.73 for IB and velocity, respectively, while combining the two yielded Az = 0.81. Changes in speed of sound and IB, with changes in integrity of breast implants, have been demonstrated in vitro. The results indicate the potential for quantitative assessment of silicone breast implants performed in vivo.

Influence of preload, afterload, and contractility on myocardial ultrasonic tissue characterization with integrated backscatter

Influence of hemodynamic changes in preload, afterload and contractility on myocardial integrated backscatter (IB) was studied in 26 adult mongrel dogs by measuring myocardial IB calibrated with the backscatter from the blood during volume infusion (preload alteration), during aortic constriction (afterload alteration), and during dobutamine or propranolol infusion (contractility alteration). Changes in preload, afterload or contractility did not significantly affect the calibrated myocardial IB either in the septum or in the posterior wall. Changes in preload and afterload did not affect the magnitude of cyclic variation in IB. However, dobutamine produced a significant increase in the magnitude of cyclic variation in IB and propranolol significantly decreased the magnitude of cyclic variation in IB. These data indicated that the calibrated myocardial IB is independent of preload, afterload and contractility, and that the magnitude of cyclic variation in IB is influenced by contractility. We may estimate static (related to histological changes such as fibrosis, edema, necrosis, and so on) and dynamic (related to myocardial contraction such as sarcomere length, muscle fiber orientation, and so on) properties of the myocardium more precisely using myocardial IB calibrated with the backscatter from the blood in addition to the magnitude of cyclic variation in IB.

A computer model for simulating ultrasonic scattering in biological tissues with high scatterer concentration

Scattering of ultrasonic waves by biological tissues at different scatterer concentrations is investigated using one- and two-dimensional computer simulation models. The backscattered power as a function of scatterer concentrations is calculated using two types of incident waves, a Gaussian shaped pulse and a continuous wave (CW). The simulation results are in good agreement with the Percus-Yevick packing theory within the scatterer concentrations, from 0% to 100% in one-dimensional (1D) space, and 0% to 46% in two-dimensional (2D) space. In all cases, the simulation results from a pulsed incident wave show a much smaller standard deviation (SD) than those from an incident CW. The simulation can serve as a useful tool to verify scattering theories, simulate different experimental conditions, and to investigate the interaction between the scatterer properties and the scattering of ultrasonic waves. More importantly, the 2D simulation procedure serves as an initial step toward the final realization of a true three-dimensional (3D) simulation of ultrasonic scattering in biological tissues.

A comparison of methods used to calculate ultrasonic myocardial backscatter in the time domain

The cyclic variation of ultrasonic integrated backscatter from the left ventricular posterior wall and interventricular septum of the heart is now well documented in the literature, with minimal values occurring at end-systole and maximal at end-diastole. However, little work has been performed to date to determine whether cyclic cardiac variation of other, more easily derived, backscatter parameters exists. In this study, 20 baseline, epicardial, long-axis cardiac-cycle sequences were obtained from eight open-chest pigs, yielding a total of 285 ultrasonic frames of RF data, which were analysed off-line on a Sun workstation. In addition, the video data from these studies was digitised and collected from each scan. Five backscatter parameters, calculated in the time-domain, including (1) the average integrated backscatter; (2) the average of the backscatter power from the log-compressed data; (3) the square of the average of the uncompressed radio frequency (RF) amplitude; (4) the square of the average of the RF amplitude from the log-compressed data; and (5) the square of the average grey-scale video data, were analysed and their variation throughout the cardiac cycle correlated against that obtained from integrated backscatter measurements. The backscatter values obtained were referenced to a gel calibration phantom widely used for 2-D calibration studies. Significant (p < 0.001), good correlation existed between the four backscatter parameters derived from the RF data. Reduced correlation was obtained between the video backscatter parameter and those derived from the RF data. Maximum cyclic variation between systole and diastole was measured from the integrated backscatter parameter and minimum from the three log-compressed data sets.