Characterisation of atherosclerotic plaque by spectral analysis of intravascular ultrasound: an in vitro methodology

Quantitative characterization of the colorectal cancer in a rabbit model using high-frequency endoscopic ultrasound

PURPOSE

Colonoscopy accompanied with biopsy works as the routine endoscopic strategy for the diagnosis of colorectal cancer (CRC) in clinic; however, the colonoscopy is limited to the tissue surface. During the last decades, enabling technologies are emerging to complement with the colonoscopy for better administration of CRC. The conventional low-frequency (<12 MHz) endoscopic ultrasound (EUS) guided fine-needle aspiration (FNA) has been widely used to assess the lesion penetration. With the high-frequency ultrasound transducer (>20 MHz), EUS allows more precise visualization of the colorectal abnormalities. In order to achieve the accurate detection or in situ characterization of the colorectal lesions, the EUS diagnosis needs more patho-physiological related information in the micro-structural or molecular level. Quantitative ultrasound (QUS) technique, which could extract the micro-structural information from the ultrasound radio-frequency (RF) signal, is promising for the non-invasive tissue characterization. To date, the knowledge of the high-frequency endoscopic QUS for the CRC characterization has not been fully determined.

METHODS

In this work, to our best knowledge, it is the first application of the QUS technique based on a customized high-frequency EUS system (30.5 MHz center frequency) to characterize the colorectal malignancies in a VX2 rabbit CRC model. To eliminate the response from the ultrasound electronic system and transducer, the ultrasound signals from colon tissue were calibrated. And, the resulting quasi-liner ultrasound spectra were fit by the linear regression test. As a result, three spectral parameters, including the slope (k), intercept (I) and Midband Fit (M), were obtained from the best-fit line. The three spectral parameters were compared between the malignant tissue regions and adjacent normal tissue regions of the colon tissue specimen ex vivo. The independent t-test was conducted between the three parameters from the normal and malignant group. The statistical method of Fisher Linear Discriminant (FLD) was used to explore the linear combinations of the three parameters, so as to provide more tissue micro-structural features than the single parameter alone. The three FLD values were derived from three different combinations among k, I and M. The threshold was selected from the statistical analysis to optimize the differentiation criteria between the malignant and the normal tissues. The color-coded images were used to display the local FLD values and combined with the EUS B-mode image.

RESULTS AND CONCLUSIONS

The mean Midband Fit (M) and intercept (I) showed significant differences between the malignant and normal tissue regions. The statistical analysis showed that there were significant differences in all the mean FLD values of the spectral parameter combinations (kI, kM and IM) (t test, P < 0.05). And, the combined image result from the B-mode image and color-coded image could visually correlate with the histology result. In conclusion, the high-frequency endoscopic QUS technique was potential to be used as a complementary method to distinguish the colorectal malignancies by leveraging its morphological and micro-structural ultrasound information.

Delineation of atherosclerotic plaque using subharmonic imaging filtering techniques and a commercial intravascular ultrasound system

The ability to delineate atherosclerotic plaque from the surrounding tissue using custom-developed subharmonic imaging (SHI) digital filtering techniques was investigated in vivo using a commercially available system. Atherosclerosis was induced in the aorta of two Watanabe Heritable Hyperlipidemic rabbits following which injections of an ultrasound contrast agent (UCA) Definity (Lantheus Medical Imaging, N Billerica, Massachusetts) were administered. Imaging was performed using a Galaxy intravascular ultrasound (IVUS) scanner (Boston Scientific, Natick, Massachusetts) equipped with an Atlantis® SR Pro Imaging Catheter (Boston Scientific). Four preliminary band-pass filters were designed to isolate the subharmonic signal (from surrounding tissue) and applied to the radio-frequency (RF) data. Preliminary filter performances were compared in terms of vessel-tissue contrast-to-tissue ratio (CTR) and visual examination. Based on preliminary results, a subharmonic adaptive filter and a stopband (SB) filter were designed and applied to the RF data. Images were classified as fundamental, SHI, and SB. Four readers performed qualitative analysis of 168 randomly selected images (across all three imaging modes). The images were scored for overall image quality, image noise, plaque visualization, and vessel lumen visualization. A Wilcoxon signed-rank test was used to compare the scores followed by intraclass correlation (ICC) evaluation. Quantitative analysis was performed by calculating the CTRs for the vessel-to-plaque and vessel-to-tissue (compared using a paired student's t test). Qualitative analysis showed SHI and SB to have significantly less image noise relative to the fundamental mode (p < 0.001). Fundamental mode scored significantly higher than SHI and SB for the remaining three categories. ICC showed mixed results among reader evaluation for delineation of plaque. However, quantitatively, SHI produced the best vessel-plaque CTR.

Probabilistic inverse problem to characterize tissue-equivalent material mechanical properties

The understanding of internal processes that affect the changes of consistency of soft tissue is a challenging problem. An ultrasound-monitoring Petri dish has been designed to monitor the evolution of relevant mechanical parameters during engineered tissue formation processes in real time. A better understanding of the measured ultrasonic signals required the use of numerical models of the ultrasound-tissue interactions. The extraction of relevant data and its evolution with sufficient sensitivity and accuracy is addressed by applying well-known signal processing techniques to both the experimental and numerically predicted measurements. In addition, a stochastic model-class selection formulation is used to rank which of the proposed interaction models are more plausible. The sensitivity of the system is verified by monitoring a gelation process.

Normalization and backscatter spectral analysis of human carotid arterial data acquired using a clinical linear array ultrasound imaging system

The risk of plaque rupture in carotid atherosclerotic disease is associated more closely with the composition of plaque rather than the severity of stenosis. The constituents of plaque can be determined from ultrasonic spectral parameters obtained from normalized backscatter tissue data. Calibration of the data is done using echoes off a specular reflector which removes the system response of an ultrasound transducer, Terason (Teratech Corporation), from the backscatter data. A reference spectrum study is used to compare specular reflectors based on time domain (echo) and frequency domain (power spectrum, centroid and parabola test) analysis. Nylon and a tissue-mimicking phantom (velocity = 1560 m/s, slope of attenuation = 0.7 dB/cm MHz) have an intermediate acoustic impedance with respect to water and appear good choices as specular reflectors for clinical ultrasound imaging scanners compared to Plexiglas and other higher reflecting materials. A tissue-mimicking phantom is used to correct for attenuation in plaque, diffraction and saturation of electronics of the ultrasound scanner. Autoregressive power spectrum estimation methods are used to extract spectral parameters (spectral slope, y-intercept, midband fit, maximum and minimum power with corresponding frequencies, and integrated backscatter) from calibrated tissue data and linear and quadratic discriminant rules developed for classification of carotid arterial plaque. Regions of interest (n = 64; 64 samples x 8 scan lines with 30 MHz sampling frequency) consisting of 48 fibrous-fibrofatty (Class 1), 11 thrombus-necrotic core (Class 2), and 5 dense calcium (Class 3) areas selected for analysis show that fibrosis can be differentiated from necrosis and calcification. The quadratic discriminant rule identified necrosis with a lower misclassification rate (9.1%) than the linear discriminant rule (18.2%).

Relationship between ultrasonic attenuation, size and axial strain parameters for ex vivo atherosclerotic carotid plaque

Many ultrasonic parameters, primarily related to attenuation and scatterer size, have been used to characterize the composition of atherosclerotic plaque tissue. In this study, we combine elastographic (axial strain ratio) and ultrasonic tissue characterization parameters, namely the attenuation coefficient and a scattering parameter associated with an "equivalent" scatterer size to delineate between fibrous, calcified, and lipidic plaque tissue. We present results obtained from 44 ex vivo atherosclerotic plaque specimens obtained after carotid endarterectomy on human patients. Our results in the frequency range 2.5 - 7.5 MHz indicate that softer plaques (with higher values of the strain ratio) are usually associated with larger equivalent scatterer size estimates (200 - 500 microm) and lower values of the attenuation coefficient slope (<1 dB/cm/MHz). On the other hand, stiffer plaques (with lower strain ratio values) are associated with smaller equivalent scatterer size estimates (100 - 200 microm) and higher values of the attenuation coefficient slope (1 - 3 dB/cm/MHz). These results indicate that ultrasonic tissue characterization and strain parameters have the potential to differentiate between different plaque types. These parameters can be estimated from radio-frequency data acquired under in vivo conditions and may help the clinician decide on appropriate interventional techniques.

Simulation and validation of arterial ultrasound imaging and blood flow

We reviewed the simulation and validation of arterial ultrasound imaging and blood flow assessment. The physical process of ultrasound imaging and measurement is complex, especially in disease. Simulation of physiological flow in a phantom with tissue equivalence of soft tissue, vessel wall and blood is now achievable. Outstanding issues are concerned with production of anatomical models, simulation of arterial disease, refinement of blood mimics to account for non-Newtonian behavior and validation of velocity measurements against an independent technique such as particle image velocimetry. String and belt phantoms offer simplicity of design, especially for evaluation of velocity estimators, and have a role as portable test objects. Electronic injection and vibrating test objects produce nonphysiologic Doppler signals, and their role is limited. Computational models of the ultrasound imaging and measurement process offer considerable flexibility in their ability to alter multiple parameters of both the propagation medium and ultrasound instrument. For these models, outstanding issues are concerned with the inclusion of different tissue types, multilayer arteries, inhomogeneous tissues and diseased tissues.

Ultrasound kidney image analysis for computerized disorder identification and classification using content descriptive power spectral features

The objective of this work is to classify few important kidney categories by characterizing the tissues of kidney region using the unique power spectral features with ultrasound as imaging modality. The images are acquired from male and female subjects of age 45 +/- 15 years. Three kidney categories namely normal, medical renal diseases and cortical cyst are considered for the analysis. The acquired images are initially preprocessed to retain the pixels-of-interest. The proposed features depend on the spatial distribution of spectral components in the kidney region. A set of power spectral features P(T)(W1), P(T)(W2), P(T-W12)R1, P(T-W12)R2, P(T-W1d)R3 and P(T-W1d)R4 are estimated at the specific cut-off frequencies omega(rc1), and omega(rc2) in the spectrum and by considering global mean total power. The results obtained show that the features are highly content descriptive and provide discrete range of values for each kidney category. Such isolated feature values facilitate to identify the kidney categories objectively which may be used as a secondary observer. The proposed method and features also explores the possibility of implementing computer-aided diagnosis system exclusively for US kidney images.

Update: aptamers as novel radiopharmaceuticals: their applications and future prospects in diagnosis and therapy

The production of biomaterials with the capacity to bind tightly and specifically to cell surface receptors of malignant cells can greatly benefit cancer diagnosis and treatment. Whereas antibodies have the ability to specifically recognize some tumor cell makers, their large size and immunogenecity markedly limit their value. The development of nuclease-resistant oligonucleotide agents, termed aptamers, offers an alternative to antibodies as targeting, diagnostic, and delivery agents. Using the systematic evolution of ligands by exponential enrichment (SELEX) methodology or other variations, one can select specific sequences that have appropriate binding affinities and specificities against clinically relevant markers from large libraries of oligonucleotide ligands. Aptamers have been found to bind their targets with high specificity and with dissociation constants in the subnanomolar or picomolar range. However, the possibility for the selected aptamers to be developed as targeting agents for diagnostic imaging or targeted radiotherapy purposes has yet to be realized. Peptide-coupling reactions between amino and carboxylic groups offer the possibility of labeling the aptamers with a number of chelators that, coupled with appropriate radionuclides, would generate novel targeted radiopharmaceuticals for the diagnosis and therapy of disease. The unparalleled combinatorial chemical diversity, small size, and modification ability of aptamers is expected to meet the criteria for robust, generic drug discovery technology and open new horizons for the development of future radiopharmaceuticals.

Physical properties of tissues relevant to arterial ultrasound imaging and blood velocity measurement

A review was undertaken of physical phenomena and the values of associated physical quantities relevant to arterial ultrasound imaging and measurement. Arteries are multilayered anisotropic structures. However, the requirement to obtain elasticity measurements from the data available using ultrasound imaging necessitates the use of highly simplified constitutive models involving Young's modulus, E. Values of E are reported for healthy arteries and for the constituents of diseased arteries. It is widely assumed that arterial blood flow is Newtonian. However, recent studies suggest that non-Newtonian behavior has a strong influence on arterial flow, and the balance of published evidence suggests that non-Newtonian behavior is associated primarily with red cell deformation rather than with aggregation. Hence, modeling studies should account for red cell deformation and the shear thinning effect that this produces. Published literature in healthy adults gives an average hematocrit and high-shear viscosity of 0.44 +/- 0.03 and 3.9 +/- 0.6 mPa.s, respectively. Published data on the acoustic properties of arteries and blood is sufficiently consistent between papers to allow compilation and derivation of best-fit equations summarizing the behavior across a wide frequency range, which then may be used in future modeling studies. Best-fit equations were derived for the attenuation coefficient vs. frequency in whole arteries (R(2) = 0.995), plasma (R(2) = 0.963) and blood with hematocrit near 45% (R(2) = 0.999), and for the backscatter coefficient vs. frequency from blood with hematocrit near 45% (R(2) = 0.958).

Emerging approaches for imaging vulnerable plaques in patients

The diagnosis of vulnerable plaques, which have the propensity to develop atherothrombosis, remains an elusive goal in clinical medicine. The most accepted features of vulnerable plaques, such as a large lipid core, increased inflammatory milieu and thin fibrous caps, have been well characterized through pathological studies. The ability to image a vulnerable plaque in susceptible patients would theoretically result in useful prognostic information that can be used to either monitor or treat patients at risk more aggressively. Several invasive techniques, such as integrated backscatter, virtual histology, palpography, optical coherence tomography and thermal heterogeneity, have been validated ex vivo and are now being evaluated in clinical studies. Non-invasive techniques, such as nuclear imaging, show promise in identifying increased metabolic activity and characteristic features of vulnerable plaques in patients. Natural history and intervention studies will need to be performed to determine whether identifying and treating vulnerable plaques will lead to improved clinical outcomes.

Laser Doppler anemometry measurements of the shear stresses on ultrasonic contrast agent microbubbles attached to agar

Ultrasonic contrast agents are currently being developed to target and bind to specific areas of interest such as atheromous plaque. A microbubble has been developed in-house which can be targeted to attach to specific cell-lines. To assess the feasibility of using the microbubble in vivo, the shear stresses which the bound microbubbles can withstand need to be known. A flow chamber was developed for use with intravascular ultrasound (IVUS) and laser Doppler anemometry (LDA). Biotin was incorporated into the microbubble shells and streptavidin was used to attach them to agar. IVUS at 40 MHz was then used to image the attached microbubbles under steady flow at a range of flow rates from 75 to 480 mL min(-1) through a flow area of 9 mm(2). LDA was employed to find high resolution velocity profiles of the flow in the chamber at a selection of these flow rates and the shear stresses on the bubbles were calculated. The bubbles were found to remain attached to the agar for shear stresses of up to 3.4 Pa. This compares with mean physiological arterial shear stresses of less than 1.5 Pa for pulsatile flow.

Peripheral application of intravascular ultrasound virtual histology

Intravascular ultrasound (IVUS) provides direct depiction of coronary artery anatomy. Traditional use of this tomographic imaging modality has been in determination of geometric measurements of an artery, such as lumen or plaque size. However, by analyzing the backscattered or radiofrequency (RF) data it is possible to glean information on the composition of plaques. This chapter describes the theory of spectral analysis and its application clinical practice.

Carotid plaque spaces relate to symptoms and ultrasound scattering

We have previously found that spectral analysis of ultrasound (US) can discriminate in vitro plaques from asymptomatic and symptomatic (transient ischemic attack within previous 4 weeks) patients. That study found no differences in percentages of lipid or thrombus between the two groups by optical microscopy/planimetry. The present study was to find out if another feature from the microscopy could show a difference. The number and size of spaces resulting from cell death or new blood vessels were measured to see if they related to symptoms or could help explain US differentiation. Twelve plaques from each group were examined by optical microscopy. The sizes and concentrations of two kinds of spaces, endothelial lined (vessel spaces) and unlined (tissue spaces), were correlated both with symptoms and also with the US tissue characterization scores from the previous study. Symptomatic vs. asymptomatic plaques showed a higher concentration and a larger size: 0.87 vs. 0.21 per mm(2) (p < 0.005) and 154 vs. 110 microm (p < 0.02). A discriminant function of spaces with symptoms as dependent variables correctly identified 91.7% of the plaques (p < 0.001). The concentration in plaques previously designated by US as true positive plaques or true negative was 1.21 vs. 0.22 per mm(2) (p < 0.005). spaces were increased in plaques of symptomatic patients and were related to UTC scores. Both the lined and unlined spaces were useful as predictors.

Classification of arterial plaque by spectral analysis in remodelled human atherosclerotic coronary arteries

We aimed to characterise and to identify the predominant plaque type in vivo using unprocessed radiofrequency (RF) intravascular ultrasound (US) backscatter, in remodelled segments of human atherosclerotic coronary arteries. A total of 16 remodelled segments were identified using a 30-MHz intravascular ultrasound (IVUS) scanner in vivo. Of these, 9 segments were classified as positively remodelled (>1.05 of the total vessel area in comparison with the proximal and distal reference segments) and 7 as negatively remodelled (<0.95 of reference segment area). Spectral parameters (maximum power, mean power, minimum power and power at 30 MHz) were determined and plaque type was defined as mixed fibrous, calcified or lipid-rich. Positively remodelled segments had a larger total vessel area (16.5 +/- 1.1 mm2 vs. 8.7 +/- 0.9 mm2, p<0.01) and plaque area (7.3 +/- 1.1 mm2 vs. 4.4 +/- 0.8 mm2, p=0.05) than negatively remodelled segments. Both positively and negatively remodelled segments had a greater percentage of fibrous plaque (p<0.01) than calcified or lipid-rich plaque. Comparing positively and negatively remodelled segments, there was no significant difference between the proportion of fibrous, calcified or lipid-rich plaque. We have been able to characterise and to identify plaque composition in vivo in human atherosclerotic coronary arteries. Our data suggest that remodelled segments are predominantly composed of fibrous plaque, as identified by RF analysis, although plaque composition is similar, irrespective of the remodelling type.

Characterizing vulnerable plaque features with intravascular elastography

BACKGROUND

In vivo detection of vulnerable plaques is presently limited by a lack of diagnostic tools. Intravascular ultrasound elastography is a new technique based on intravascular ultrasound and has the potential to differentiate between different plaques phenotypes. However, the predictive value of intravascular elastography to detect vulnerable plaques had not been studied.

METHODS AND RESULTS

Postmortem coronary arteries were investigated with intravascular elastography and subsequently processed for histology. In histology, a vulnerable plaque was defined as a plaque consisting of a thin cap (<250 microm) with moderate to heavy macrophage infiltration and at least 40% of atheroma. In elastography, a vulnerable plaque was defined as a plaque with a high strain region at the surface with adjacent low strain regions. In 24 diseased coronary arteries, we studied 54 cross sections. In histology, 26 vulnerable plaques and 28 nonvulnerable plaques were found. Receiver operator characteristic analysis revealed a maximum predictive power for a strain value threshold of 1.26%. The area under the receiver operator characteristic curve was 0.85. The sensitivity was 88%, and the specificity was 89% to detect vulnerable plaques. Linear regression showed high correlation between the strain in caps and the amount of macrophages (P<0.006) and an inverse relation between the amount of smooth muscle cells and strain (P<0.0001). Plaques, which are declared vulnerable in elastography, have a thinner cap than nonvulnerable plaques (P<0.0001).

CONCLUSIONS

Intravascular elastography has a high sensitivity and specificity to detect vulnerable plaques in vitro.

Toward virtual biopsy through an all fiber optic ultrasonic miniaturized transducer: a proposal

The present generation of devices based on opto-acoustic and acousto-optic conversion lets us foresee the possibility of realizing complete miniaturized transmitting-receiving transducers, able to generate and detect wideband ultrasounds by laser light. In the present paper, a miniaturized ultrasonic transducer entirely based on fiber optic technology is proposed. Such a device springs from the conjunction between our research, which has produced a highly efficient fiber optic opto-acoustic source, with the results obtained by other researchers concerning the realization of an ultrasonic receiver based on optical interferometry. Making use of the thermo-elastic effect for ultrasound generation, a source of ultrasound can be obtained by coupling a fiber optic to pulsed laser, if a film capable of absorbing laser light is placed onto fiber end. Starting from these remarks, we propose an efficient opto-acoustic source, able to generate pressure pulses with amplitude of the order of 10(4) Pa and bandwidth extending up to 40 MHz and beyond by using graphite materials as absorbing film. This solution makes use of a low-power pulsed laser as optical source possible. An ultrasonic receiving element was realized placing a Fabry-Perot cavity over the tip of a fiber optic. The cavity thickness modulation induced by ultrasonic beam is detected by an interferometer optical technique. We have realized a prototype of a receiving device that exhibits a sensitivity comparable with that of piezoelectric devices (10-100 nV/Pa) and an almost flat bandwidth extending up to 20 MHz or more. The extreme miniaturization of the resulting ultrasonic transducer, together with its wide ultrasonic frequency bandwidth, is the first step toward ultrasonic tissue biopsy. In this paper, before discussing the problem of constructing a complete ultrasonic transducer composed by a transmitter and receiver, the results carried out in these fields during the last decade are reviewed.

Thrombus characterization with intravascular ultrasound: potential to predict successful thrombolysis

PURPOSE

To define analysis methods using radiofrequency (RF) data from intravascular ultrasound (IVUS) to characterize thrombus so that the success or failure of thrombolysis can be predicted.

METHODS

Experimentation was done in 2 phases: first, 40 clots created from platelet-rich (n=20) and platelet-poor (n=20) plasma were imaged with 20 and 30-MHz IVUS probes. The digitized RF signals were analyzed to determine the attenuation and slope of attenuation characteristics as a reference standard for the second part of the study. In this phase, 20 perfusion trials were performed with 10 platelet-rich and 10 platelet-poor samples in an in vitro arterial perfusion model; the composition of the samples was blinded during interrogation with IVUS. Alteplase (0.1 mg) was then infused directly into the thrombus for 1 hour; restoration of flow and amount of lysis were quantified.

RESULTS

Significant differences were detected for the integrated attenuation value (p<0.001) between platelet-poor and platelet-rich clots interrogated with a 30-MHz probe, but no differences were noted using the 20-MHz catheter. However, both transducers were capable of differentiating platelet-rich and platelet-poor clots using a slope of attenuation analysis, which was successful in identifying all 20 (100%) of the blinded clots in the in vitro model. The speed and completeness of lysis were significantly higher (p<0.05) for platelet-rich (78%) than platelet-poor (21%) clots.

CONCLUSIONS

IVUS preprocessed imaging is capable of differentiating various types of thrombus. This information can be used to predict the success or failure of pharmacological lysis techniques.

Processing radio frequency ultrasound images: a robust method for local spectral features estimation by a spatially constrained parametric approach

Spectral estimation is a major component in studies aiming at characterizing biological tissues through the analysis of backscattered radio frequency (RF) ultrasonic signals and images. However, conventional spectral estimation techniques yield a well-known trade-off between spatial resolution and variance. The backscattered signals are stochastic by nature, so short-term local analysis results in a high variance of the estimates, which cannot efficiently be reduced through conventional spatial averaging. We address this issue by describing a spectral estimation technique that reduces the variance of the estimates (by smoothing the local estimates in spectrally homogeneous regions) while preserving spectral discontinuities (i.e., the smoothing is not performed across regions with different spectral contents). The proposed approach is set in a Bayesian framework and is based on local autoregressive (AR) estimation, constrained by smoothness priors. These smoothness priors are introduced through a Markov random field in which the associated potential functions are nonquadratic, allowing thereby to preserve discontinuity. The method is validated on simulated RF images and tested on echocardiographic images acquired in vivo. The results are compared to the estimates provided by the conventional Burg technique. These results clearly demonstrate the ability of the proposed approach to improve spectral estimation in terms of variance reduction and discontinuity detection.

Dynamic noise correction for IVUS quantitative volume blood flow: methods and numerical validation

In recent years, a new method to measure transverse blood flow based on the decorrelation of the radio-frequency (RF) signals, has been developed. Transverse blood flow estimation may be influenced by noise. In this paper, we investigated a new correlation-based method for noise correction. The decorrelation characteristics of transverse blood flow using an intravascular ultrasound (US) or IVUS array catheter were studied by means of computer modeling. Blood was simulated as a collection of randomly located point scatterers; moving this scattering medium transversely across the acoustical beam represented flow. Parabolic blood flow was simulated. Additive noise was added to the RF signals at a given signal-to-noise ratio (SNR). Next, a new method to dynamically estimate and suppress the decorrelation due to noise is presented. The decorrelation due to noise was estimated from the correlation coefficients from flowing blood obtained at increasing time lags. The correlation graphs are corrected for the decorrelation due to noise, leaving the decorrelation due to blood flow. The method shows promise to estimate and correct the correlation coefficients for noise.

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.

Decorrelation-based blood flow velocity estimation: effect of spread of flow velocity, linear flow velocity gradients, and parabolic flow

In recent years, a new method to measure transverse blood flow, based on the decorrelation of the radio frequency (RF) signals has been developed. In this paper, we investigated the influence of nonuniform flow on the velocity estimation. The decorrelation characteristics of transverse blood flow using an intravascular ultrasound (IVUS) array catheter are studied by means of computer modeling. Blood was simulated as a collection of randomly located point scatterers; moving this scattering medium transversally across the acoustical beam represented flow. First-order statistics were evaluated, and the signal-to-noise ratio from the signals were measured. The correlation coefficient method was used to present the results. Three velocity profiles were simulated: random spread of blood-flow velocity, linear blood-flow velocity gradient, and parabolic blood-flow. Radio frequency and envelope signals were used to calculate the decorrelation pattern. The results were compared to the mean decorrelation pattern for plug blood-flow. The RF signals decorrelation patterns were in good agreement with those obtained for plug blood flow. Envelope decorrelation patterns show a close agreement with the one for plug blood flow. For axial blood flow, there is a discrepancy between decorrelation patterns. The results presented here suggest that the decorrelation properties of an IVUS array catheter for measuring quantitative transverse blood flow probably will not be affected by different transverse blood-flow conditions.

In vitro acoustic characterisation of four intravenous ultrasonic contrast agents at 30 MHz

The acoustic properties of four ultrasonic contrast agents (Optison, Definity, SonoVue and Sonazoid) were studied at 30 MHz using a Boston Scientific ClearView Ultra intravascular ultrasound (US) scanner modified to allow access to the unprocessed US data. A range of contrast agent concentrations were studied using either saline or glucose as the diluent of choice. Mean backscatter power was measured over regions-of-interest (ROI) at distances of 1, 1.5, 2, 3, 4 and 5 mm from the centre of the intravascular probe and normalised to the US data collected from a standard glass reflector. For all of the agents, the mean backscatter power at 30 MHz varied in a linear manner with concentration between 0.01 million microbubbles/mL and 1 million microbubbles/mL. Furthermore, for two of the agents, mean backscatter enhancement was detectable at concentrations as low as 2 microbubbles/sample volume.

Effects of transducer position on backscattered intensity in coronary arteries

Acute myocardial infarction is a frequent cause of sudden death, and is typically initiated by the rupture of coronary artery plaques. The likelihood and severity of rupture are influenced by the plaque structures and components. Radiofrequency (RF) intravascular ultrasound (US) (IVUS-RF) measurements extend current IVUS imaging techniques and may eventually enable the in vivo identification of these features. However, IVUS-RF measurements are affected by the transducer's instantaneous position in the vessel. Specifically, backscattered intensity (BI) decreases as either the distance between the tissue and the transducer increases, or as the beam's angle of incidence on the tissue increases. IVUS-RF data were acquired from seven disease-free coronary arteries in vitro. The 0-dB level for BI was defined as the peak intensity of the reflection from a stainless-steel flat reflector at each distance. The baseline BI measured in adventitial tissue was -32.5 dB (at 0 degrees, 0 mm) with angle and distance dependencies of -0.172 dB/ degrees and -3.37 dB/mm. In contrast, the BI from combined intima and media was -38.2 dB with dependencies of -0.111 dB/ degrees and -4.46 dB/mm (p < 0.05 for all three parameters). Acknowledging and compensating for these effects may allow IVUS-RF to develop into a rapidly deployable tool for the clinical detection of vulnerable plaques and to monitor coronary artery disease progression and regression.

Assessing spectral algorithms to predict atherosclerotic plaque composition with normalized and raw intravascular ultrasound data

Spectral analysis of backscattered intravascular ultrasound (IVUS) data has demonstrated the ability to characterize plaque. We compared the ability of spectral parameters (e.g., slope, midband fit and y-intercept), computed via classic Fourier transform (CPSD), Welch power spectrum (WPSD) and autoregressive (MPSD) models, to classify plaque composition. Data were collected ex vivo from 32 human left anterior descending coronary arteries. Regions-of-interest (ROIs), selected from histology, comprised 64 collagen-rich, 24 fibrolipidic, 23 calcified and 37 calcified-necrotic regions. A novel quantitative method was used to correlate IVUS data with corresponding histologic sections. Periodograms of IVUS samples, identified for each ROI, were used to calculate spectral parameters. Statistical classification trees (CT) were computed with 75% of the data for plaque characterization. The remaining data were used to assess the accuracy of the CTs. The overall accuracies for normalized spectra with CPSD, WPSD and MPSD were, respectively, 84.7%, 85.6% and 81.1% (training data) and 54.1%, 64.9% and 37.8% (test data). These numbers were improved to 89.2%, 91.9% and 89.2% (training) and 62.2%, 73% and 59.5% (test) when the calcified and calcified-necrotic regions were combined for analysis. Most CTs misclassified a few fibrolipidic regions as collagen, which is histologically acceptable, and the unnormalized and normalized spectra results were similar.

Impact of coronary plaque morphology as assessed by IVUS computer-aided analysis on mechanisms of balloon angioplasty and stenting

This study was performed in order to quantitate structural coronary plaque modifications after balloon angioplasty and stenting and to evaluate the impact of plaque morphology on the mechanisms of lumen enlargement during angioplasty. Plaque morphology was studied by computer-aided analysis of 60 cross-sectional intravascular ultrasound (IVUS) images of the target lesion in 20 patients undergoing percutaneous coronary angioplasty. Based on a computer-aided video densitometry classification of plaque morphology, three groups of plaques were defined based on the slope value of a fifth polynomial regression of the plaque gray-level distribution. In groups A and B, balloon angioplasty provided significant increases in lumen area (P < 0.0001) and vessel area (P < 0.05) without a reduction in plaque area; neither parameter increased in group C. In group A, stenting was associated with an additional lumen enlargement (P < 0.0001) due to plaque reduction (P < 0.05). In groups B and C, stenting further increased lumen area (P < 0.0001) by improving vessel area (P < 0.001) but without plaque reduction. Balloon angioplasty and stenting provided a significant decrease in plaque area in group A as compared to groups B (P < 0.05) and C (P < 0.01). Finally, vessel area improvement was greater in group B than in groups A (P < 0.01) and C (P < 0.05). The mechanisms underlying lumen enlargement after coronary angioplasty are highly dependent on plaque morphology as defined by an IVUS computer-aided analysis and may differ between balloon angioplasty and stenting.

Decorrelation characteristics of transverse blood flow along an intravascular array catheter: effects of aggregation of red blood cells

A method to measure transverse blood flow, based on the correlation between consecutive radiofrequency (RF) signals, has been introduced. This method was validated for an intravascular (IVUS) rotating single element catheter. Currently, we are implementing the method for an IVUS array transducer catheter. The decorrelation characteristics during transverse blood flow using the IVUS array catheter were investigated using computer modeling. Before this, blood was simulated as a collection of randomly located point scatterers and, by moving this scattering medium transversely across the acoustical beam, blood flow was simulated. This paper presents a more realistic scattering media by simulating aggregates of red blood cells (RBCs) as strings of point scatterers. Three configurations of aggregates of RBCs were simulated. First, aggregates of RBCs were strings with different lengths and parallel to the catheter axis. Second, the strings were with a fixed length and angles of plus or minus 45 degrees with respect to the catheter axis. Third, the strings were with different lengths and random angles ranging from -45 degrees to + 45 degrees. The decorrelation characteristics for these configurations of aggregates of RBCs were investigated and compared with point scatterers. For the aggregates of RBCs parallel to the catheter axis, the decorrelation rate became slower when the aggregate length was increased. RBC aggregations with fixed and random lengths and angles resulted in a decorrelation rate that approaches the decorrelation pattern from point scatterers. Results suggests that the presence of aggregates of RBCs will probably not affect the measurements of transverse blood flow using a decorrelation-based method and an IVUS array catheter.

Delineation of the extracellular determinants of ultrasonic scattering from elastic arteries

Elastic arteries consist of three primary components: elastin fibers, extracellular collagen matrix and smooth muscle cells. However, the relative contribution of elastin and collagen fibers to overall ultrasonic scattering from an intact arterial wall is poorly understood. To define the principal source of extracellular scattering from the medial layer of elastic arteries, canine ascending aortas (n = 10) were excised, fixed and sectioned for insonification. Subsequently, aortic specimens were restudied after treatment to dissolve all tissue components except extracellular collagen matrix (n = 5) and elastin fibers (n = 5). Histological staining revealed very few elastin fibers and sparse intact collagen in collagen-isolated and elastin-isolated tissues, respectively. Integrated backscatter, attenuation and backscatter coefficients differentiated these two treated tissues. The backscatter coefficient for elastin-isolated tissue demonstrated a fivefold increase over collagen-isolated tissue, suggesting that elastin fibers represent a primary scattering component within elastic arteries, and the collagen fibers may provide a secondary component of scattering.

Classification of arterial plaque by spectral analysis of in vitro radio frequency intravascular ultrasound data

To test whether radio-frequency analysis of coronary plaques predicts the histological classification, r.f. data were collected using a 30 MHz intravascular ultrasound scanner. Two hundred ninety-nine regions-of-interest from eight postmortem coronary arteries were selected and identified by histology as falling into one of seven different tissue types. These are loose fibrous tissue (n = 78), moderate fibrous tissue (n = 27), dense fibrous tissue (n = 33), microcalcification (n = 14), calcified plaque (n = 55), lipid/fibrous mixture (n = 51) and homogeneous areas of lipid pool (n = 29). On the basis of a previous study, four spectral parameters were calculated for each of the regions-of-interest: maximum power (dB), mean power (dB), spectral slope (dB/MHz) over the bandwidth 18-35 MHz and the intercept of the spectral slope with the 0 Hz axis (dB). A minimum-distance classifier using the Mahalanobis (1948) distance was applied to the data. Following resubstitution of the training data into the classifier, the total correctly classified was 54%. The data were reclassified using three broader tissue groups: (1) calcified plaque, (2) lipid pool and (3) a mixed fibrous category, incorporating loose fibrous tissue, moderate fibrous tissue, dense fibrous tissue, lipid/fibrous mixture and microcalcification. The total correctly classified was 86%. Using "leave-one-out" cross-validation, the classification rates were 48% for seven tissue subgroups and 83% for three broader categories of tissue type.

Decorrelation characteristics of transverse blood flow along an intravascular array catheter

In recent years, a new method to measure transverse blood flow based on the decorrelation of radio frequency (RF) signals has been introduced. In this paper, we investigated the decorrelation characteristics of transverse blood flow measurement using an intravascular ultrasound (IVUS) array catheter by means of computer modeling. Blood was simulated as a collection of randomly located point scatterers. Moving this scattering medium transversally across the acoustical beam represented flow. First-order statistics were evaluated, and the signal-to-noise ratio from the signals was measured. The correlation coefficient method was used to present the results. The decorrelation patterns for RF and for RF-envelope signals were studied. The decorrelation patterns from the RF signals were in good agreement with those obtained from theoretical beam profiles. This agreement suggests that the decorrelation properties of an IVUS array catheter for measuring quantitative transverse blood flow can be assessed by measuring the ultrasound beam. A line of point scatterers, moved transversally across the acoustical beam (line spread function), can determine this decorrelation behaviour.

High-resolution three-dimensional in vivo imaging of atherosclerotic plaque

The internal structure of atherosclerotic-plaque lesions may be a useful predictor of which lesions will rupture and cause sudden events such as heart attack or stroke. With lipid and flow suppression, we obtained high-resolution, three-dimensional (3D) images of atherosclerotic plaque in vivo that show the cap thickness and core size of the lesions. 3D GRASE was used because it provides flexible T(2) contrast and good resistance to off-resonance artifacts. While 2D RARE has similar properties, its resolution in the slice-select direction, which is important because of the irregular geometry of atherosclerotic lesions, is limited by achievable slice-excitation profiles. Also, 2D imaging generally achieves lower SNR than 3D imaging because, for SNR purposes, 3D image data is averaged over all the slices of a corresponding multislice 2D dataset. Although 3D RARE has many of the advantages of 3D GRASE, it requires a longer scan time because it uses more refocusing pulses to acquire the same amount of data. Finally, cardiac gating is an important part of our imaging sequence, but can make the imaging time quite long. To obtain reasonable scan times, a 2D excitation pulse was used to restrict the field of view. Magn Reson Med 42:762-771, 1999.

Developments in cardiovascular ultrasound. Part 2: Arterial applications

Many of the changes resulting from arterial disease can be measured, using Doppler ultrasound for measurement of blood velocity and B-scan imaging for measurement of tissue structure and composition. Wall thickness, the degree of arterial narrowing and plaque volume can be measured using B-scan imaging, and 3D ultrasound can be used to improve the accuracy of measurements of plaque volume and for improved visualisation of complex arterial geometries. Measurement of the dynamic properties of the arterial wall permits estimation of wall elasticity and plaque motion. From the Doppler signal, measurements of blood velocity are used to estimate the degree of arterial narrowing and volumetric flow, although measurement errors can be large. Wall shear stress can be estimated by measuring the velocity gradient at the vessel wall. The problems of inadequate spatial resolution and interference from overlying tissue are largely removed when intravascular systems are used, and these have superior capability in the assessment of arterial structure and tissue composition. However, measurement of quantities relating to blood flow is more difficult using the intravascular approach, as the indwelling cather disturbs the blood flow pattern, and currently, assessment of flow and vessel cross-section are not performed at the same site.

Characterisation of coronary atherosclerotic morphology by spectral analysis of radiofrequency signal: in vitro intravascular ultrasound study with histological and radiological validation

OBJECTIVE

To determine whether spectral analysis of unprocessed radiofrequency (RF) signal offers advantages over standard videodensitometric analysis in identifying the morphology of coronary atherosclerotic plaques.

METHODS

97 regions of interest (ROI) were imaged at 30 MHz from postmortem, pressure perfused (80 mm Hg) coronary arteries in saline baths. RF data were digitised at 250 MHz. Two different sizes of ROI were identified from scan converted images, and relative amplitudes of different frequency components were analysed from raw data. Normalised spectra was used to calculate spectral slope (dB/MHz), y-axis intercept (dB), mean power (dB), and maximum power (dB) over a given bandwidth (17-42 MHz). RF images were constructed and compared with comparative histology derived from microscopy and radiological techniques in three dimensions.

RESULTS

Mean power was similar from dense fibrotic tissue and heavy calcium, but spectral slope was steeper in heavy calcium (-0.45 (0.1)) than in dense fibrotic tissue (-0.31 (0.1)), and maximum power was higher for heavy calcium (-7.7 (2.0)) than for dense fibrotic tissue (-10.2 (3.9)). Maximum power was significantly higher in heavy calcium (-7.7 (2.0) dB) and dense fibrotic tissue (-10.2 (3.9) dB) than in microcalcification (-13.9 (3.8) dB). Y-axis intercept was higher in microcalcification (-5.8 (1.1) dB) than in moderately fibrotic tissue (-11.9 (2.0) dB). Moderate and dense fibrotic tissue were discriminated with mean power: moderate -20.2 (1.1) dB, dense -14.7 (3.7) dB; and y-axis intercept: moderate -11.9 (2.0) dB, dense -5.5 (5.4) dB. Different densities of fibrosis, loose, moderate, and dense, were discriminated with both y-axis intercept, spectral slope, and mean power. Lipid could be differentiated from other types of plaque tissue on the basis of spectral slope, lipid -0.17 (0.08). Also y-axis intercept from lipid (-17.6 (3.9)) differed significantly from moderately fibrotic tissue, dense fibrotic tissue, microcalcification, and heavy calcium. No significant differences in any of the measured parameters were seen between the results obtained from small and large ROIs.

CONCLUSION

Frequency based spectral analysis of unprocessed ultrasound signal may lead to accurate identification of atherosclerotic plaque morphology.

Intravascular ultrasound elastography in human arteries: initial experience in vitro

Intravascular elastography is a new technique to obtain the local mechanical properties of the vessel wall and its pathology using intravascular ultrasound (IVUS). Knowledge of these mechanical properties may be useful for guiding interventional procedures. An experimental set-up is described for assessment of the strain data of arteries. Using a 30-MHz IVUS catheter, radio frequency data are acquired with a custom-made high-performance data acquisition system. High-resolution, local tissue displacement estimation by cross-correlation is followed by computation of local strain. An algorithm that uses a priori knowledge of the correlation coefficient function was applied to filter the obtained strain data. With this experimental set-up, intravascular elastograms containing 400 angles/revolution with a radial resolution of 200 microns can be produced. The feasibility of intravascular elastography with this experimental set-up is demonstrated using two diseased human femoral arteries. Qualitative comparison of the elastograms with the echograms and the histology demonstrates the potential of intravascular elastography to obtain mechanical information from the vessel wall and from plaque.

Characterisation of red and white thrombus by intravascular ultrasound using radiofrequency and videodensitometric data-based texture analysis

Visual assessment of intravascular ultrasound (IVUS) video images cannot reliably identify thrombus. We examined if texture analysis of radiofrequency (r.f.) data or videodensitometric data (VD) could distinguish thrombi of different ages and cell compositions. Whole human blood (red clot = RC), platelet-rich plasma (white clot = WC) and plasma (n = 6/group) were imaged at 4 and 24 h with 30 MHz IVUS transducers. At 4 h, VD- and r.f.-based analyses revealed significant differences between RC and WC with variance (VD red 26.4 +/- 2.5, white 33.9 +/- 7.8; r.f. red 1.4 +/- 0.5, white 4.9 +/- 1.3), kurtosis (VD red 0.29 +/- 0.9, white 0.23 +/- 0.3) and skewness (VD red 0.23 +/- 0.13, white 0.35 +/- 0.52; r.f. red 0.06 +/- 0.01, white -0.06 +/- 0.05). Also mean grey-level from both data sets was higher in RC (VD 134.8 +/- 18.0; r.f. -13.3 +/- 1.2) than in WC (VD 105.3 +/- 17.4, r.f. 16.5 +/- 2.2) (p < 0.01). With increasing time, variance increased in WC (5.5 +/- 1.5 at 24 h) and decreased in RC (0.9 +/- 0.3.3 at 24 h). The more heterogeneous structure of WC may be distinguished from that of RC using texture analysis of either VD or r.f.-signals.