Regularized autoregressive analysis of intravascular ultrasound backscatter: improvement in spatial accuracy of tissue maps

Mapping adipose tissue in short-axis echocardiograms using spectral analysis

The number one cause of death in the United States is consistently cardiovascular disease (CVD). Studies have proven that the buildup of cardiac adipose tissue (CAT) around the heart is a biomarker of CVD. MRI is the gold standard for imaging CAT but is expensive and not widely available. Ultrasound is less expensive and portable, but the images are noisy, and it is difficult to identify or quantify CAT. The aim of this project is to use spectral analysis of raw radiofrequency (RF) ultrasound data as input for a machine learning classifier to automatically classify regions-of-interest (ROIs) around the heart as containing CAT or not. ROIs are labeled using corresponding MRI images of the same patients. A previous study used a random forest classifier with 9 spectral parameters as input to classify tissue types in echocardiograms. This project focuses on improving this classifier by experimenting with properties of the chosen ROIs. Experiments were performed independently varying the ROI circumference (length), width, the threshold CAT thickness used for labeling an ROI as CAT, and the signal level required for valid processing. Additional experiments explored the impact of the anatomical location of each ROI as an input. The addition of two parameters indicating the distance of each ROI from the two left and right myocardium intersections as well as the use of the optimal ROI parameters as determined from experimentation resulted in an accuracy of 75.5%. This demonstrates feasibility of this approach for identifying CAT around the heart and will lead to future work in estimating the thickness of fat in each ROI.

Spectral analysis of ultrasound radiofrequency backscatter for the identification of epicardial adipose tissue

Purpose: The coronary arteries are embedded in a layer of fat known as epicardial adipose tissue (EAT). The EAT influences the development of coronary artery disease (CAD), and increased EAT volume can be indicative of the presence and type of CAD. Identification of EAT using echocardiography is challenging and only sometimes feasible on the free wall of the right ventricle. We investigated the use of spectral analysis of the ultrasound radiofrequency (RF) backscatter for its potential to provide a more complete characterization of the EAT. Approach: Autoregressive (AR) models facilitated analysis of the short-time signals and allowed tuning of the optimal order of the spectral estimation process. The spectra were normalized using a reference phantom and spectral features were computed from both normalized and non-normalized data. The features were used to train random forests for classification of EAT, myocardium, and blood. Results: Using an AR order of 15 with the normalized data, a Monte Carlo cross validation yielded accuracies of 87.9% for EAT, 84.8% for myocardium, and 93.3% for blood in a database of 805 regions-of-interest. Youden's index, the sum of sensitivity, and specificity minus 1 were 0.799, 0.755, and 0.933, respectively. Conclusions: We demonstrated that spectral analysis of the raw RF signals may facilitate identification of the EAT when it may not otherwise be visible in traditional B-mode images.

Tissue classification in intercostal and paravertebral ultrasound using spectral analysis of radiofrequency backscatter

Paravertebral and intercostal nerve blocks have experienced a resurgence in popularity. Ultrasound has become the gold standard for visualization of the needle during injection of the analgesic, but the intercostal artery and vein can be difficult to visualize. We investigated the use of spectral analysis of raw radiofrequency (RF) ultrasound signals for identification of the intercostal vessels and six other tissue types in the intercostal and paravertebral spaces. Features derived from the one-dimensional spectrum, two-dimensional spectrum, and cepstrum were used to train four different machine learning algorithms. In addition, the use of the average normalized spectrum as the feature set was compared with the derived feature set. Compared to a support vector machine (SVM) (74.2%), an artificial neural network (ANN) (68.2%), and multinomial analysis (64.1%), a random forest (84.9%) resulted in the most accurate classification. The accuracy using a random forest trained with the first 15 principal components of the average normalized spectrum was 87.0%. These results demonstrate that using a machine learning algorithm with spectral analysis of raw RF ultrasound signals has the potential to provide tissue characterization in intercostal and paravertebral ultrasound.

Spectral Analysis of Ultrasound Radiofrequency Backscatter for the Detection of Intercostal Blood Vessels

Spectral analysis of ultrasound radiofrequency backscatter has the potential to identify intercostal blood vessels during ultrasound-guided placement of paravertebral nerve blocks and intercostal nerve blocks. Autoregressive models were used for spectral estimation, and bandwidth, autoregressive order and region-of-interest size were evaluated. Eight spectral parameters were calculated and used to create random forests. An autoregressive order of 10, bandwidth of 6 dB and region-of-interest size of 1.0 mm resulted in the minimum out-of-bag error. An additional random forest, using these chosen values, was created from 70% of the data and evaluated independently from the remaining 30% of data. The random forest achieved a predictive accuracy of 92% and Youden's index of 0.85. These results suggest that spectral analysis of ultrasound radiofrequency backscatter has the potential to identify intercostal blood vessels. (jokling@siue.edu) © 2018 World Federation for Ultrasound in Medicine and Biology.

A New Dynamic Complex Baseband Pulse Compression Method for Chirp-Coded Excitation in Medical Ultrasound Imaging

Chirp-coded excitation can increase the signal-to-noise ratio (SNR) without degrading the axial resolution. Effective pulse compression (PC) is important to maintain the axial resolution and can be achieved with radio frequency (RF) and complex baseband (CBB) data (i.e., and , respectively). can further reduce the computational complexity compared to ; however, suffers from a degraded SNR due to tissue attenuation. In this paper, we propose a new dynamic CBB PC method ( that can improve the SNR while compensating for tissue attenuation. The compression filter coefficients in the method are generated by dynamically changing the demodulation frequencies along with the depth. For PC, the obtained coefficients are independently applied to the in-phase and quadrature components of the CBB data. To evaluate the performance of the proposed method, simulation, phantom, and in vivo studies were conducted, and all three studies showed improved SNR, i.e., maximally 3.87, 7.41, and 5.75 dB, respectively. In addition, the measured peak range sidelobe level of the proposed method yielded lower values than the and , and it also derived a suitable target location, i.e., a <0.07-mm target location error, while maintaining the axial resolution. In an in vivo abdominal experiment, the method depicted brighter and clearer features in the hyperechoic region because highly correlated signals were produced by compensating for tissue attenuation. These results demonstrated that the proposed method can improve the SNR of chirp-coded excitation while preserving the axial resolution and the target location and reducing the computational complexity.

Experimental evaluation of spectral-based quantitative ultrasound imaging using plane wave compounding

Quantitative ultrasound (QUS) based on backscatter coefficient (BSC) estimation has shown potential for tissue characterization. Beamforming using plane wave compounding has advantages for echographic, Doppler, and elastographic imaging; however, to date, plane wave compounding has not been experimentally evaluated for the purpose of BSC estimation. In this study, two BSC-derived parameters (i.e., the BSC midband fit and intercept) were estimated from experimental data obtained using compound plane wave beamforming. For comparison, QUS parameters were also estimated from data obtained using both fixed focus and dynamic receive beamforming. An ultrasound imaging system equipped with a 9-MHz center frequency, 64-element array was used to collect data up to a depth of 45 mm. Two gelatin phantoms with randomly distributed 20-μm inclusions with a homogeneous scatterer concentration and a two-region scatterer concentration were used for assessing the precision and lateral resolution of QUS imaging, respectively. The use of plane wave compounding resulted in accurate QUS estimation (i.e., bias in the BSC parameters of less than 2 dB) and relatively constant lateral resolution (i.e., BSC midband fit 10% to 90% rise distance ranging between 1.0 and 1.5 mm) throughout a 45 mm field of view. Although both fixed focus and dynamic receive beamforming provided the same performance around the focal depth, the reduction in SNR away from the focus resulted in a reduced field of view in the homogeneous phantom (i.e., only 28 mm). The lateral resolution also degraded away from the focus, with up to a 2-fold and 10-fold increase in the rise distance at 20 mm beyond the focal depth for dynamic receive and fixed focus beamforming, respectively. These results suggest that plane wave compounding has the potential to improve the performance of spectral-based quantitative ultrasound over other conventional beamforming strategies.

Correlation of cognitive function with ultrasound strain indices in carotid plaque

Instability in carotid vulnerable plaque can generate cerebral micro-emboli, which may be related to both stroke and eventual cognitive abnormality. Strain imaging to detect plaque vulnerability based on regions with large strain fluctuations, with arterial pulsation, may be able to determine the risk of cognitive impairment. Plaque instability may be characterized by increased strain variations over a cardiac cycle. Radiofrequency signals for ultrasound strain imaging were acquired from the carotid arteries of 24 human patients using a Siemens Antares with a VFX 13-5 linear array transducer. These patients underwent standardized cognitive assessment (Repeatable Battery for the Assessment of Neuropsychological Status [RBANS]). Plaque regions were segmented by a radiologist at end-diastole using the Medical Imaging Interaction Toolkit. A hierarchical block-matching motion tracking algorithm was used to estimate the cumulated axial, lateral and shear strains within the imaging plane. The maximum, minimum and peak-to-peak strain indices in the plaque computed from the mean cumulated strain over a small region of interest in the plaque with large deformations were obtained. The maximum and peak-to-peak mean cumulated strain indices over the entire plaque region were also computed. All strain indices were then correlated with RBANS Total performance. Overall cognitive performance (RBANS Total) was negatively associated with values of the maximum strain and the peak-to-peak for axial and lateral strains, respectively. There was no significant correlation between the RBANS Total score and shear strain and strain indices averaged over the entire identified plaque for this group of patients. However, correlation of maximum lateral strain was higher for symptomatic patients (r = -0.650, p = 0.006) than for asymptomatic patients (r = -0.115, p = 0.803). On the other hand, correlation of maximum axial strain averaged over the entire plaque region was significantly higher for asymptomatic patients (r = -0.817, p = 0.016) than for symptomatic patients (r = -0.224, p = 0.402). The results reveal a direct relationship between the maximum axial and lateral strain indices in carotid plaque and cognitive impairment.

Intracranial-derived atherosclerosis assessment: an in vitro comparison between virtual histology by intravascular ultrasonography, 7T MRI, and histopathologic findings

BACKGROUND AND PURPOSE

Atherosclerotic plaque composition and structure contribute to the risk of plaque rupture and embolization. Virtual histology by intravascular ultrasonography and high-resolution MR imaging are new imaging modalities that have been used to characterize plaque morphology and composition in peripheral arteries.

MATERIALS AND METHODS

The objectives of this study were 1) to determine the correlation between virtual histology-intravascular ultrasonography and histopathologic analysis (reference standard) and 2) to explore the comparative results of 7T MR imaging (versus histopathologic analysis), both to be performed in vitro by use of intracranial arterial segments with atherosclerotic plaques. Thirty sets of postmortem samples of intracranial circulation were prepared for the study. These samples included the middle cerebral artery (n = 20), basilar artery (n = 8), and anterior cerebral artery (n = 2). Virtual histology-intravascular ultrasonography and 7T MR imaging were performed in 34 and 10 points of interest, respectively. The formalin-fixed arteries underwent tissue processing and hematoxylin-eosin staining. The plaques were independently categorized according to revised Stary classification after review of plaque morphology and characteristics obtained from 3 modalities. The proportion of fibrous, fibrofatty, attenuated calcium, and necrotic components in the plaques were determined in histology slides and compared with virtual histology-intravascular ultrasonography and MR imaging.

RESULTS

Of 34 points of interest in the vessels, 32 had atherosclerotic plaques under direct visualization. Plaques were visualized in gray-scale intravascular ultrasonography as increased wall thickness, outer wall irregularity, and protrusion. The positive predictive value of virtual histology-intravascular ultrasonography for identifying fibroatheroma was 80%. Overall, virtual histology-intravascular ultrasonography accurately diagnosed the type of the plaque in 25 of 34 samples, and κ agreement was 0.58 (moderate agreement). The sensitivity and specificity of virtual histology-intravascular ultrasonography readings for fibroatheroma were 78.9% and 73.3%, respectively. The overall sensitivity and specificity for virtual histology-intravascular ultrasonography were 73.5% and 96.6%, respectively. Plaques were identified in 7T MR imaging as increased wall thickness, luminal stenosis, or outer wall protrusion. The positive predictive value of 7T MR imaging for detecting fibrous and attenuated calcium deposits was 88% and 93%, respectively.

CONCLUSIONS

This in vitro study demonstrated that virtual histology-intravascular ultrasonography and high-resolution MR imaging are reliable imaging tools to detect atherosclerotic plaques within the intracranial arterial wall, though both imaging modalities have some limitations in accurate characterization of the plaque components. Further clinical studies are needed to determine the clinical utility of plaque morphology and composition assessment by noninvasive tests.

The vulnerable coronary plaque: update on imaging technologies

Several studies have been carried out on vulnerable plaque as the main culprit for ischaemic cardiac events. Historically, the most important diagnostic technique for studying coronary atherosclerotic disease was to determine the residual luminal diameter by angiographic measurement of the stenosis. However, it has become clear that vulnerable plaque rupture as well as thrombosis, rather than stenosis, triggers most acute ischaemic events and that the quantification of risk based merely on severity of the arterial stenosis is not sufficient. In the last decades, substantial progresses have been made on optimisation of techniques detecting the arterial wall morphology, plaque composition and inflammation. To date, the use of a single technique is not recommended to precisely identify the progression of the atherosclerotic process in human beings. In contrast, the integration of data that can be derived from multiple methods might improve our knowledge about plaque destabilisation. The aim of this narrative review is to update evidence on the accuracy of the currently available non-invasive and invasive imaging techniques in identifying components and morphologic characteristics associated with coronary plaque vulnerability.

Coronary plaque characteristics that indicate distal embolization during percutaneous coronary intervention in patients with stable angina-virtual histology intravascular ultrasound study

Distal embolization (DE) is a serious complication of percutaneous coronary intervention (PCI) in patients with stable angina. The purpose of this study was to evaluate the coronary plaque characteristics that indicate DE during PCI in patients with stable angina using virtual histology intravascular ultrasound (VH-IVUS). Three hundred and sixty-four consecutive stable angina patients who underwent PCI were enrolled in this study. The patients were divided into two groups as follows: patients exhibiting DE (DE group, n = 10) and patients without DE (non-DE group, n = 354). Coronary plaque compositions were assessed by VH-IVUS. The fibro-fatty (FF) ratio (28 ± 17 vs. 11 ± 9 %, p < 0.0001) was higher in the DE group compared with the non-DE group. The best cut-off value of FF ratio for prediction of DE was 20 %, with a sensitivity of 0.80 and a specificity of 0.81 (odds ratio; 17.1, 95 % confidence interval 3.56-82.5, p = 0.0004). Coronary plaques with a high FF ratio may be the predictor of indicating DE in patients with stable angina during PCI.

Virtual histology analysis of carotid atherosclerotic plaque: plaque composition at the minimum lumen site and of the entire carotid plaque

BACKGROUND

Virtual Histology intravascular ultrasound (VH IVUS) volumetric analysis (analysis of the entire plaque responsible for stenosis) has been used for carotid plaque diagnosis. Knowing the carotid plaque characteristics by analyzing the plaque composition only at the minimum lumen site will facilitate plaque diagnosis using VH IVUS.

PURPOSE

To detect the relationship between the VH IVUS volumetric analysis of the entire plaque responsible for carotid artery stenosis and the VH IVUS cross-section plaque analysis at the minimum lumen site.

METHODS

Forty-eight atherosclerotic cervical carotid stenoses in 45 consecutive patients were included in the study. VH IVUS was obtained during the carotid artery stenting procedure.

RESULTS

Simple regression analysis revealed that the volumetric proportion of each plaque type correlated significantly with the corresponding plaque-type area at the minimum lumen site. The adjusted coefficients of determination of the simple regression analyses were .782 (P < .001) for fibrous tissue, .741 (P < .001) for fibrofatty tissue, .864 (P < .001) for dense calcium, and .918 (P < .001) for necrotic core.

CONCLUSION

The plaque composition at the minimum lumen site represents the volumetric composition of the entire carotid plaque that causes atherosclerotic cervical carotid artery stenosis.

Thermal strain imaging: a review

Thermal strain imaging (TSI) or temporal strain imaging is an ultrasound application that exploits the temperature dependence of sound speed to create thermal (temporal) strain images. This article provides an overview of the field of TSI for biomedical applications that have appeared in the literature over the past several years. Basic theory in thermal strain is introduced. Two major energy sources appropriate for clinical applications are discussed. Promising biomedical applications are presented throughout the paper, including non-invasive thermometry and tissue characterization. We present some of the limitations and complications of the method. The paper concludes with a discussion of competing technologies.

Combined optical coherence tomography and intravascular ultrasound radio frequency data analysis for plaque characterization. Classification accuracy of human coronary plaques in vitro

This study was performed to characterize coronary plaque types by optical coherence tomography (OCT) and intravascular ultrasound (IVUS) radiofrequency (RF) data analysis, and to investigate the possibility of error reduction by combining these techniques. Intracoronary imaging methods have greatly enhanced the diagnostic capabilities for the detection of high-risk atherosclerotic plaques. IVUS RF data analysis and OCT are two techniques focusing on plaque morphology and composition. Regions of interest were selected and imaged with OCT and IVUS in 50 sections, from 14 human coronary arteries, sectioned post-mortem from 14 hearts of patients dying of non-cardiovascular causes. Plaques were classified based on IVUS RF data analysis (VH-IVUS^TM), OCT and the combination of those. Histology was the benchmark. Imaging with both modalities and coregistered histology was successful in 36 sections. OCT correctly classified 24; VH-IVUS 25, and VH-IVUS/OCT combined, 27 out of 36 cross-sections. Systematic misclassifications in OCT were intimal thickening classified as fibroatheroma in 8 cross-sections. Misclassifications in VH-IVUS were mainly fibroatheroma as intimal thickening in 5 cross-sections. Typical image artifacts were found to affect the interpretation of OCT data, misclassifying intimal thickening as fibroatheroma or thin-cap fibroatheroma. Adding VH-IVUS to OCT reduced the error rate in this study.

Development of a dual-modal tissue diagnostic system combining time-resolved fluorescence spectroscopy and ultrasonic backscatter microscopy

We report a tissue diagnostic system which combines two complementary techniques of time-resolved laser-induced fluorescence spectroscopy (TR-LIFS) and ultrasonic backscatter microscopy (UBM). TR-LIFS evaluates the biochemical composition of tissue, while UBM provides tissue microanatomy and enables localization of the region of diagnostic interest. The TR-LIFS component consists of an optical fiber-based time-domain apparatus including a spectrometer, gated multichannel plate photomultiplier, and fast digitizer. It records the fluorescence with high sensitivity (nM concentration range) and time resolution as low as 300 ps. The UBM system consists of a transducer, pulser, receiving circuit, and positioning stage. The transducer used here is 45 MHz, unfocused, with axial and lateral resolutions 38 and 200 microm. Validation of the hybrid system and ultrasonic and spectroscopic data coregistration were conducted both in vitro (tissue phantom) and ex vivo (atherosclerotic tissue specimens of human aorta). Standard histopathological analysis of tissue samples was used to validate the UBM-TRLIFS data. Current results have demonstrated that spatially correlated UBM and TR-LIFS data provide complementary characterization of both morphology (necrotic core and calcium deposits) and biochemistry (collagen, elastin, and lipid features) of the atherosclerotic plaques at the same location. Thus, a combination of fluorescence spectroscopy with ultrasound imaging would allow for better identification of features associated with tissue pathologies. Current design and performance of the hybrid system suggests potential applications in clinical diagnosis of atherosclerotic plaque.

Can periprocedural hypotension in carotid artery stenting be predicted? A carotid morphologic autonomic pathologic scoring model using virtual histology to anticipate hypotension

SUMMARY

Periprocedural hypotension, which frequently occurs during carotid artery stenting (CAS), is an important risk factor for complications such as stroke or death after CAS. To determine if a scoring model can be established to predict periprocedural hypotension (systolic blood pressure < or = 90 mm Hg) and prolonged periprocedural hypotension (requiring vasopressor for > 3 hours) in CAS, we conducted a prospective cohort study of patients undergoing interventional treatment of cervical carotid artery stenosis in an urban tertiary referral hospital from April 2006 to April 2007. Forty-eight stenotic lesions in 45 consecutive patients treated with CAS were included in the study. Multivariate analysis showed three independent risk factors of periprocedural hypotension; "fibrous plaque on Virtual Histology" (P = 0.029), "stenotic lesion involving both the common carotid artery and internal carotid artery on angiogram" (P = 0.004), and "patients without history of diabetes mellitus" (P = 0.020). Further, "distance between carotid bifurcation and point of minimum lumen size < or = 10 mm on angiogram" (P = 0.003) was an independent risk factor of prolonged periprocedural hypotension. Carotid morphologic autonomic pathologic score (carotid MAPS), determined by adding one point for each of those risk factors (total 0 to 4), had good discrimination for both periprocedural hypotension (area under receiver operating characteristic curve: ROC AUC = 0.876; SE 0.053) and prolonged periprocedural hypotension (ROC AUC = 0.811; SE 0.066). Carotid MAPS is useful for predicting periprocedural hypotension and prolonged periprocedural hypotension during CAS.

The virtual histology intravascular ultrasound appearance of newly placed drug-eluting stents

Intravascular ultrasound (IVUS) is used before and after intervention and at follow-up to assess the quality of the acute result as well as the long-term effects of stent implantation. Virtual histology (VH) IVUS classifies tissue into fibrous and fibrofatty plaque, dense calcium, and necrotic core. Although most interventional procedures include stent implantation, VH IVUS classification of stent metal has not been validated. In this study, the VH IVUS appearance of acutely implanted stents was assessed in 27 patients (30 lesions). Most stent struts (80%) appeared white (misclassified as "calcium") surrounded by red (misclassified as "necrotic core"); 2% appeared just white, and 17% were not detectable (compared with grayscale IVUS because of the software-imposed gray medial stripe). The rate of "white surrounded by red" was similar over the lengths of the stents; however, undetectable struts were mostly at the distal edges (31%). Quantitatively, including the struts within the regions of interest increased the amount of "calcium" from 0.23 +/- 0.35 to 1.07 +/- 0.66 mm(2) (p <0.0001) and the amount of "necrotic core" from 0.59 +/- 0.65 to 1.31 +/- 0.87 mm(2) (p <0.0001). Most important, because this appearance occurs acutely, it is an artifact, and the red appearance should not be interpreted as peristrut inflammation or necrotic core when it is seen at follow-up. In conclusion, acutely implanted stents have an appearance that can be misclassified by VH IVUS as "calcium with or without necrotic core." It is important not to overinterpret VH IVUS studies of chronically implanted stents when this appearance is observed at follow-up. A separate classification for stent struts is necessary to avoid these misconceptions and misclassifications.

Low adiponectin levels are associated with atherogenic dyslipidemia and lipid-rich plaque in nondiabetic coronary arteries

OBJECTIVE

The purpose of this study was to determine whether an association exists between adiponectin and plaque composition in human coronary arteries.

RESEARCH DESIGN AND METHODS

Adiponectin is an adipocyte-derived protein with antiatherogenic and insulin-sensitizing properties. To date, the relationship between adiponectin and plaque composition is unknown. Fasting blood samples were collected from 185 patients undergoing coronary angiography and intravascular ultrasound (IVUS). Plaque composition was categorized as fibrous, fibrofatty, necrotic core, or dense calcium and further classified as IVUS-derived adaptive or pathological intimal thickening, fibroatheroma, fibrocalcific, or thin cap fibroatheroma.

RESULTS

Adiponectin correlated with normalized plaque volume (r = -0.16, P = 0.025) and atheroma lipid content as measured by normalized fibrofatty volume (r = -0.19, P = 0.009). Low adiponectin levels were associated with IVUS-derived pathological intimal thickening (r = -0.18, P = 0.01). With increasing quartiles (Q) of adiponectin, the normalized volume of fibrofatty plaque decreased (P = 0.03), which was driven by reductions in the nondiabetic cohort (Q1 44.2 mm(3); Q2 28.2 mm(3); Q3 24.7 mm(3); and Q4 23.4 mm(3); P = 0.01). No similar association was present in diabetic patients. Low adiponectin levels were also associated with IVUS-derived pathological intimal thickening in nondiabetic (r = -0.20, P = 0.03) but not diabetic patients.

CONCLUSIONS

Low adiponectin levels are associated with atherogenic lipoproteins (elevated triglycerides, small dense LDL cholesterol, and low HDL cholesterol), increased plaque volume, lipid-rich plaque, and IVUS-derived pathological intimal thickening in the total cohort that was driven by the nondiabetic population, suggesting an antiatherogenic role in the early stages of lesion development.

Diagnosis and treatment of coronary vulnerable plaques

Thin-capped fibroatheroma is the morphology that most resembles plaque rupture. Detection of these vulnerable plaques in vivo is essential to being able to study their natural history and evaluate potential treatment modalities and, therefore, may ultimately have an important impact on the prevention of acute myocardial infarction and death. Currently, conventional grayscale intravascular ultrasound, virtual histology and palpography data are being collected with the same catheter during the same pullback. A combination of this catheter with either thermography capability or additional imaging, such as optical coherence tomography or spectroscopy, would be an exciting development. Intravascular magnetic resonance imaging also holds much promise. To date, none of the techniques described above have been sufficiently validated and, most importantly, their predictive value for adverse cardiac events remains elusive. Very rigorous and well-designed studies are compelling for defining the role of each diagnostic modality. Until we are able to detect in vivo vulnerable plaques accurately, no specific treatment is warranted.

Diagnostic accuracy of optical coherence tomography and intravascular ultrasound for the detection and characterization of atherosclerotic plaque composition in ex-vivo coronary specimens: a comparison with histology

BACKGROUND

Both intravascular ultrasound and optical coherence tomography have been purported to accurately detect and characterize coronary atherosclerotic plaque composition. The aim of our study was to directly compare the reproducibility and diagnostic accuracy of optical coherence tomography and intravascular ultrasound for the detection and characterization of coronary plaque composition ex vivo as compared with histology.

METHODS AND RESULTS

Intravascular ultrasound (20 MHz) and optical coherence tomography imaging was performed in eight heart specimens using motorized pullback. Standard histology using hematoxylin-eosin and van Gieson staining was performed on 4 mum thick slices. Each slice was divided into quadrants and accurately matched cross-sections were analyzed for the presence of fibrous, lipid-rich, and calcified coronary plaque using standard definitions for both intravascular ultrasound and optical coherence tomography and correlated with histology. After exclusion of 145/468 quadrants, we analyzed the remaining 323 quadrants with excellent image quality in each procedure. Optical coherence tomography demonstrated a sensitivity and specificity of 91/88% for normal wall, 64/88% for fibrous plaque, 77/94% for lipid-rich plaque, and 67/97% for calcified plaque as compared with histology. Intravascular ultrasound demonstrated a sensitivity and specificity of 55/79% for normal wall, 63/59% for fibrous plaque, 10/96% for lipid-rich plaque, and 76/98% for calcified plaque. Both intravascular ultrasound and optical coherence tomography demonstrated excellent intraobserver and interobserver agreement (optical coherence tomography: kappa=0.90, kappa=0.82; intravascular ultrasound: kappa=0.87, kappa=0.86).

CONCLUSION

Optical coherence tomography is superior to intravascular ultrasound for the detection and characterization of coronary atherosclerotic plaque composition, specifically for the differentiation of noncalcified, lipid-rich, or fibrous plaque.

In vivo relationship between compositional and mechanical imaging of coronary arteries. Insights from intravascular ultrasound radiofrequency data analysis

OBJECTIVE

We sought to explore in vivo the relation between mechanical and compositional properties of matched cross sections (CSs) using novel catheter-based techniques.

BACKGROUND

Intravascular ultrasound (IVUS) palpography allows the assessment of local mechanical tissue properties. Spectral analysis of IVUS radiofrequency data (IVUS-VH) is a tool to assess plaque morphology and composition.

METHODS AND RESULTS

Palpography analysis defined high- and low-strain regions. One hundred twenty-three CSs (27 vessels) were colocalized. The mean strain value was higher in CSs with necrotic core (NC) in contact with the lumen than in CSs with no NC contact with the lumen (1.03 +/- 0.5 vs 0.86 +/- 0.4, P = .06). Mean relative calcium (1.61 +/- 2.5% vs 0.25 +/- 0.7%, P = .001) and NC (15.64 +/- 10.6% vs 2.8 +/- 3.9%, P < .001) content were significantly higher in the CSs with NC in contact with the lumen, whereas the inverse was seen for the fibrotic component of the plaque (64.16 +/- 11.6% vs 75.75 +/- 13.7, P < .001). The sensitivity, specificity, positive predictive value, and negative predictive value of IVUS-VH to detect high strain were 75.0%, 44.4%, 56.3%, and 65.1%, respectively. A significant inverse relationship was present between calcium and strain levels (r = -0.20, P = .03). After adjusting for univariate predictors, the contact of NC with the lumen was identified as the only independent predictor of high strain (OR 5.0, 95% CI 1.7-14.1, P = .003).

CONCLUSION

In the present study, IVUS-VH showed an acceptable sensitivity to detect high strain. In turn, the specificity was low. Of interest, a significant inverse relationship was present between calcium and strain levels.

Association of plaque characterization by intravascular ultrasound virtual histology and arterial remodeling

Positive remodeling is more often observed in lesions of patients who have acute coronary syndromes or vulnerable (rupture-prone) plaques. However, there are few data that correlate plaque morphology, composition, and arterial remodeling in vivo. We evaluated coronary plaque characterization of lesions with positive remodeling using intravascular ultrasound (IVUS) radiofrequency data analysis. Seventy-seven nonbifurcation native coronary lesions (in 50 patients) were imaged in vivo using 30-MHz IVUS transducers. Lesions were classified into 4 plaque types, fibrous, fibrofatty, dense calcium, and necrotic core, by using processing of the radiofrequency signal validated in vitro. The remodeling index was calculated as the lesion external elastic membrane area divided by the proximal reference external elastic membrane area. Lesions were divided into 2 groups: positive remodeling (remodeling index>1.0, 26 lesions) and intermediate/negative remodeling (remodeling index<or=1.0, 51 lesions). Total plaque volume and the volume of each plaque type were averaged over the length of the lesion. Reference segment plaque compositions were similar. Mean lesion fibrofatty plaque area was significantly larger in lesions with positive remodeling than in lesions with intermediate/negative remodeling (1.2+/-0.7 vs 0.8+/-0.4 mm2, p=0.001; 26.3+/-6.6% vs 19.8+/-5.7%, p<0.001, of total plaque volume). The same results were obtained at the minimum lumen site and in the subgroup of patients who had acute coronary syndromes. Further, there was a linear relation between remodeling index and fibrofatty plaque area (r=0.26, p=0.02). In conclusion, in vivo IVUS radiofrequency data analysis demonstrates that positive remodeling occurs in lesions with more fibrofatty plaque.

In vivo intravascular ultrasound-derived thin-cap fibroatheroma detection using ultrasound radiofrequency data analysis

OBJECTIVES

The purpose of this study was to assess the prevalence of intravascular ultrasound (IVUS)-derived thin-cap fibroatheroma (IDTCFA) and its relationship with the clinical presentation using spectral analysis of IVUS radiofrequency data (IVUS-Virtual Histology [IVUS-VH]).

BACKGROUND

Thin-cap fibroatheroma lesions are the most prevalent substrate of plaque rupture.

METHODS

In 55 patients, a non-culprit, non-obstructive (<50%) lesion was investigated with IVUS-VH. We classified IDTCFA lesions as focal, necrotic core-rich (> or =10% of the cross-sectional area) plaques being in contact with the lumen; IDTCFA definition required a percent atheroma volume (PAV) > or =40%.

RESULTS

Acute coronary syndrome (ACS) (n = 23) patients presented a significantly higher prevalence of IDTCFA than stable (n = 32) patients (3.0 [interquartile range (IQR) 0.0 to 5.0] vs. 1.0 [IQR 0.0 to 2.8], p = 0.018). No relation was found between patient's characteristics such as gender (p = 0.917), diabetes (p = 0.217), smoking (p = 0.904), hypercholesterolemia (p = 0.663), hypertension (p = 0.251), or family history of coronary heart disease (p = 0.136) and the presence of IDTCFA. A clear clustering pattern was seen along the coronaries, with 35 (35.4%), 31 (31.3%), 19 (19.2%), and 14 (14.1%) IDTCFAs in the first 10 mm, 11 to 20 mm, 21 to 30 mm, and > or =31 mm segments, respectively, p = 0.008. Finally, we compared the severity (mean PAV 56.9 +/- 7.4 vs. 54.8 +/- 6.0, p = 0.343) and the composition (mean percent necrotic core 19.7 +/- 4.1 vs. 18.1 +/- 3.0, p = 0.205) of IDTCFAs between stable and ACS patients, and no significant differences were found.

CONCLUSIONS

In this in vivo study, IVUS-VH identified IDTCFA as a more prevalent finding in ACS than in stable angina patients.

Motion artifact reduction for IVUS-based thermal strain imaging

Thermal strain imaging (TSI) using intravascular ultrasound (IVUS) has the potential to identify lipid pools within rupture-prone arterial plaques and serve as a valuable supplement to current IVUS systems in diagnosing acute coronary syndromes. The major challenge for in vivo application of TSI will be cardiac motion, including bulk motion and tissue deformation. Simulations based on an artery model, including a lipid-filled plaque, demonstrate that effective bulk motion compensation can be achieved within a certain motion range using spatial interpolation. We also propose a practical imaging scheme to minimize mechanical strains caused by tissue deformation based on a linear least squares fitting strategy. This scheme was tested on clinical data by artificially superimposing thermal displacements corresponding to different temperature rises. Results suggest a 1-2 degrees C temperature rise is required to detect lipids in an atherosclerotic plaque in vivo.

Coronary artery remodelling is related to plaque composition

OBJECTIVE

To assess the potential relation between plaque composition and vascular remodelling by using spectral analysis of intravascular ultrasound (IVUS) radiofrequency data.

METHODS AND RESULTS

41 coronary vessels with non-significant (< 50% diameter stenosis by angiography), < or = 20 mm, non-ostial lesions located in non-culprit vessels underwent IVUS interrogation. IVUS radiofrequency data obtained with a 30 MHz catheter, were analysed with IVUS virtual histology software. A remodelling index (RI) was calculated and divided into three groups. Lesions with RI > or = 1.05 were considered to have positive remodelling and lesions with RI < or = 0.95 were considered to have negative remodelling. Lesions with RI > or = 1.05 had a significantly larger lipid core than lesions with RI 0.96-1.04 and RI < or = 0.95 (22.1 (6.3) v 15.1 (7.6) v 6.6 (6.9), p < 0.0001). A positive correlation between lipid core and RI (r = 0.83, p < 0.0001) and an inverse correlation between fibrous tissue and RI (r = -0.45, p = 0.003) were also significant. All of the positively remodelled lesions were thin cap fibroatheroma or fibroatheromatous lesions, whereas negatively remodelled lesions had a more stable phenotype, with 64% having pathological intimal thickening, 29% being fibrocalcific lesions, and only 7% fibroatheromatous lesions (p < 0.0001).

CONCLUSIONS

In this study, in vivo plaque composition and morphology assessed by spectral analysis of IVUS radiofrequency data were related to coronary artery remodelling.