IVUS-based imaging modalities for tissue characterization: similarities and differences
- First Online:
- Cite this article as:
- Garcìa-Garcìa, H.M., Gogas, B.D., Serruys, P.W. et al. Int J Cardiovasc Imaging (2011) 27: 215. doi:10.1007/s10554-010-9789-7
- 972 Downloads
Gray-scale intravascular ultrasound (IVUS) is the modality that has been established as the golden standard for in vivo imaging of the vessel wall of the coronary arteries. The use of IVUS in clinical practice is an important diagnostic tool used for quantitative assessment of coronary artery disease. This has made IVUS the de-facto invasive imaging method to evaluate new interventional therapies such as new stent designs and for atherosclerosis progression-regression studies. However, the gray-scale representation of the coronary vessel wall and plaque morphology in combination with the limited resolution of the current IVUS catheters makes it difficult, if not impossible, to identify qualitatively (e.g. visually) the plaque morphology similar as that of histopathology, the golden standard to characterize and quantify coronary plaque tissue components. Meanwhile, this limitation has been partially overcome by new innovative IVUS-based post-processing methods such as: virtual histology IVUS (VH-IVUS, Volcano Therapeutics, Rancho Cordova, CA, USA), iMAP-IVUS (Bostoc Scientific, Santa Clara, CA, USA), Integrated Backscatter IVUS (IB-IVUS) and Automated Differential Echogenicity (ADE).
KeywordsIntravascular ultrasound Radiofrequency data analysis Atherosclerosis Tissue characterization
Atherogenesis is the leading cause of cardiovascular mortality and morbidity in the developed world. The imaging of coronary atherosclerosis and more in particular the high risk atheromatous plaque has made an explosive progress the last decade with the advent of innovative techniques that focus on the high-resolution visualization of the coronary vascular wall. Gray-scale intravascular ultrasound (IVUS) is the modality that has been established as the golden standard for in vivo imaging of the vessel wall of the coronary arteries [1, 2]. The use of IVUS is an important diagnostic tool used for quantitative assessment of coronary artery disease. This has made IVUS the de-facto invasive imaging method of choice to evaluate new interventional therapies such as new stent designs [3, 4, 5] and for atherosclerosis progression-regression studies [6, 7, 8, 9, 10, 11, 12, 13, 14]. However, the gray-scale representation of the coronary vessel wall and plaque morphology in combination with the limited resolution of the current IVUS catheters makes it difficult, if not impossible, to identify qualitatively (e.g. visually) the plaque morphology similar as that of histopathology, the golden standard to characterize and quantify coronary plaque tissue components. Meanwhile, this limitation has been partially overcome by new innovative IVUS-based post-processing methods such as: virtual histology IVUS [15, 16] (VH-IVUS, Volcano Therapeutics, Rancho Cordova, CA, USA), iMAP-IVUS  (Bostoc Scientific, Santa Clara, CA, USA), Integrated Backscatter IVUS  (IB-IVUS) and Automated Differential Echogenicity  (ADE).
Intravascular ultrasound (IVUS) principles
The size of current IVUS catheters are ranging from 2.6 to 3.5 French (0.87–1.17 mm) and are inserted into the coronary arteries through 6-French guiding catheters. The principle of IVUS imaging is based on the oscillatory movement (expansion and contraction) of a piezoelectric transducer (crystal) in order to produce sound waves when electrically excited. There are two major different transducer designs : (1) the mechanical single element rotating device and (2) the electronic phased array. The mechanical rotating element device uses a single piezoelectric transducer that rotates with 1,800 rotations per minute, while the electronic phased array device uses multiple stationary placed piezoelectric transducers which are sequentially activated. The generated sound waves by the transducers propagates through the different tissues and is reflected according to the acoustic properties of the tissue it travels through .
Gray-scale IVUS based atheromatous plaque classification is limited due to its low spatial resolution and for the usual IVUS transducers (20 and 40 MHz) for which the axial resolution is 200 μm and the lateral 200–250 μm. Based on their visual appeareance, not necessarily histological composition, atheromas have been classified in four categories by gray-scale IVUS: (1) soft plaque (lesion echogenicity less than the surrounding adventitia), (2) fibrous plaque (intermediate echogenicity between soft (echolucent) atheromas and highly echogenic calcified plaques), (3) calcified plaque (echogenicity higher than the adventitia with acoustic shadowing), and (4) mixed plaques (no single acoustical subtype represents >80% of the plaques) .
IVUS based imaging modalities for tissue characterization
To overcome the limitations of qualitative visual interpretation of the IVUS images and to describe the coronary plaque morphology, several post-processing methods for computer-assisted quantification have been developed during the recent years. There are two basic different approaches: (1) Signal-based analysis (the so called raw radiofrequency analysis or RF-analysis) and (2) Image-based analysis. The different methods will be described below:
Tissue characterization using virtual histology IVUS (VH-IVUS)
Validation studies of IVUS and IVUS based imaging modalities
IVUS for coronary atheromatous lesions compared to histology. Atheromatous plaque was classified as echodense, echolucent, heterogeneous or calcified by each observer and by one observer on separate occasions
Overall inter- and intra-observer reproducibility for plaque-type (Kappa 0.87[0.80–0.94] and 0.89[0. 85–0.93 respectively]) and focal calcification (0.78[0.74–0.82] and 0.88[0.84–0.92]) was high
Agreement for overall plaque type between intravascular ultrasound and histology occurred in 89% of sites (Kappa 0.73[0.69–0. 77]). Specificity ≥90%
IVUS, high frequency transducer(40 MHz) for plaque composition compared to histomorphology
Lipid pools were observed by histology in 30 sections (25%). IVUS revealed the presence of lipid pools in 19 of these sections (16%; sensitivity 65%]. Specificity ≥95%
Lipid/necrotic areas were defined by IVUS as large echolucent intraplaque areas surrounded by tissue with higher echodensity
Coronary plaque classification with intravascular ultrasound radiofrequency data analysis
Autoregressive classification schemes performed better than those from classic Fourier spectra with accuracies of 90.4% for fibrous, 92.8% for fibrolipidic, 90.9% for calcified, and 89.5% for calcified-necrotic regions in the training data set and 79.7, 81.2, 92.8, and 85.5% in the test data, respectively
Accuracy of in vivo coronary plaque morphology assessment: a validation study of in vivo virtual histology compared with in vitro histopathology
Predictive accuracy from all patients data: 87.1% for fibrous, 87.1% for fibro-fatty, 88.3% for necrotic core, and 96.5% for dense calcium regions, respectively
Sensitivities: NC:67.3%, FT:86%, FF:79.3%, DC:50%. Specificities: NC:92.9%, FT:90.5%,FF:100%, DC:99%
Automated coronary plaque characterisation with intravascular ultrasound backscatter: ex vivo validation
The overall predictive accuracies were 93.5% for FT, 94.1% for FF, 95.8% for NC, and 96.7% for DC
Sensitivities: NC:91.7%, FT:95.7%, FF:72.3%, DC:86.5%. Specificities: NC:96.6%, FT:90.9%, FF:97.9%, DC:98.9%
In vivo plaque characterization using intravascular ultrasound-virtual histology in a porcine model of complex coronary lesions
Compared with histology, IVUS-VH correctly identified the presence of fibrous, fibro-fatty, and necrotic tissue in 58.33, 38.33, and 38.33% of lesions, respectively
Sensitivities: fibrous 76.1%, fibro-fatty 46%, and necrotic core 41.1%
Van Herk 
Validation of in vivo plaque characterisation by virtual histology in a rabbit model of atherosclerosis
VH-IVUS had a high sensitivity, specificity and positive predictive value for the detection of non-calcified thin cap fibroatheroma (88, 96, 87%, respectively) and calcified thin cap fibroatheroma (95, 99, 93%, respectively). These values were respectively 82, 94, 85% for non-calcified fibroatheroma and 78, 98, 84% for calcified fibroatheroma. The lowest values were obtained for pathological intimal thickening (74, 92, 70%, respectively). For all plaque types, VH-IVUS had a kappa-value of 0.79
Unreliable assessment of necrotic core by VHTM IVUS in porcine coronary artery disease
No correlations were found between the size of the necrotic core determined by VH IVUS and histology. VH IVUS displayed necrotic cores in lesions lacking cores by histology
In vivo quantitative tissue characterization of human coronary arterial plaques by use of integrated backscatter intravascular ultrasound and comparison with angioscopic findings
r:0,954 for each category, DC, FF, FT, NC
Diagnostic accuracy of optical coherence tomography and integrated backscatter intravascular ultrasound images for tissue characterization of human coronary plaques
Sensitivities: DC:100% FT:94% Lipid pool:84%
Specificities: DC:99% FT:84% Lipid pool:97%
Development of integrated backscatter intravascular ultrasound for tissue characterization of coronary plaques
IB classified fibrous, lipid-rich and fibrocalcific plaque components with a high accuracy of 93, 90 and 96%, respectively
Characterisation of atherosclerotic plaque by spectral similarity of radiofrequency intravascular ultrasound signals
Ex vivo validation demonstrated accuracies at the highest level of confidence as: 97, 98, 95, and 98% for necrotic, lipidic, fibrotic and calcified regions respectively
Three-dimensional and quantitative analysis of atherosclerotic plaque composition by automated differential echogenicity
Areas of hypoechogenicity correlated with the presence of smooth muscle cells. Areas of hyperechogenicity correlated with presence of collagen, and areas of hyperechogenicity with acoustic shadowing correlated with calcium
Tissue characterization using i-MAP-IVUS
Similarities and differences of IVUS and IVUS-based imaging modalities
Type of device
Mecanical and electrical
Mechanical and electrical
Fibrous: light green
Necrotic core: red
Necrotic core: pink
Necrotic core: blue
Fibrofatty: light green
Backscatter radiofrequency signal analysis
Fast Fourier trannformation
Fast Fourier transformation
Tissue characterization by integrated backscatter IVUS analysis (IB-IVUS)
IB-IVUS analysis is an alternative approach, as compared to the 2 previous ones, using the RF-signals of the IVUS catheters to characterize coronary plaque tissue components. IB-IVUS analyses the RF-signals generated by the 40 MHz mechanically rotating IVUS catheters by applying a fast Fourier transformation of the frequency components of the backscattered signals calculating the intensity of the signal measured in decibels (dB). Different tissue components reflect the RF-signals at different power levels, which, according to the developers, could be used to differentiate various tissue components. Analogue to the other methods it also applies a colour coded overlay onto the gray-scale IVUS images. It comprises the following different tissue components: (1) Calcification, (2) Fibrous tissue, and (3) Lipidic [18, 22, 23].
Tissue characterization using, automated differential echogenicity (ADE)
As above described, IVUS has become over the past 20 years an important clinical intracoronary imaging tool. It not only facilitates clinical practice but it is also a reference method of which its quantitative parameters are often used as endpoints in first-in-man studies and larger clinical trials. However, the limited resolution, as compared to histology, and gray-scale (256 shades of gray are used when the image is optimal) representation, for which we humans have only limited capabilities with our eyes to make a distinction between them (on average we can only discriminate 8–12 different gray-levels) combined with the large amount of images acquired during a pullback examination (in an IVUS study of 4 cm there are 2,400 individual cross-sections acquired), requires another representation of the information we are looking for. Of course quantification is mandatory for research purposes and thus automated or semi-automated computer-assisted methods are necessary. In addition, humans can interpret colour-information much better than gray-scale information alone and therefore the plaque compositional tools as described in this overview are valuable additions to the use of standard IVUS gray-scale imaging alone. Until recently, there was only one tool (e.g. VH IVUS) commercially and thus widespreadly available. This has changed, now more methods are available which could potentially lead to different outcomes as they are using different mathematical methods and software algorithms. Certainly, in the near future the different methods will possibly be applied side-by-side in multi-center studies and it is yet unknown if this could result in possible different study outcomes, this needs to be explored in future research.
The name giving of virtual histology IVUS (VH-IVUS) could possibly lead to expectations that the results derived by this method are one-to-one comparable to histology, which is unfortunately not the case. The development of the RF-signal based methods have been performed empirically, in brief: explanted vessels are imaged by IVUS, the pathologist performs histology and the histology results are cross-correlated to the IVUS images. The derived and correlated RF-signal patterns have been used to build the signal pattern databases which are later used to identify the plaque tissue components in vivo. The crux in this development process is the cross-correlation between histology and IVUS, which due to the large differences in image resolution is difficult (by example the lateral resolution of IVUS is 200–250 μm while that for histology is 5 μm). Although histology is considered the golden standard it is also not free from possible artifacts, by example during the fixation and staining process. Interpretation of the images by different pathologists could also result in interobserver related-biases. It is therefore not realistic to expect similar results from the IVUS derived tissue compositional tools as from quantitative histology. However, if appropriately used, and with justified expectations, these quantitative plaque tissue compositional tools, with all of their limitations, can be of great additional value to investigate plaque compositional changes which cannot be performed with any other imaging method. Great care must thus be taken when trying to prove a hypothesis using these methods which could be out of range of the capabilities of the methods and the basic method, e.g. IVUS itself. An example of expectations which could lead to disappointing results when the RF-based methods are applied is within metallic stented segments. The RF-based methods do not have the signal profile of metallic stent struts in their tissue signal profile databases and this will thus result in an artificial identification (in some research papers this is identified as misuse). This could later lead to unpredictable results and thus care must be taken applying these methods for research for which they are not designed.
However, encouraging results have been achieved applying the tissue composition methods to detect and to quantify as a surrogate the absorption of the recently introduced bioabsorbable BVS stent (BVS, Abbott Vascular, Santa Clara,CA, USA). This information could help to explore into more detail the overall performance of this new stent platform.
A major limitation of the RF-based methods, as compared to the image-based echogenicity method, is that they do not take into account acoustic shadowed areas. They divide every individual RF-beam into small regions of interest and compare the found signal profiles to their databases of known tissue profiles without taking into account what signal is found in front or behind that window (e.g. the larger picture is missing). This results in detection of tissue behind calcified areas, even if on the gray-scale IVUS image there is only shadow visible, thus the RF-signal will most likely contain noise (thick layers of calcium will reflect all acoustic energy back to the transducer causing the typical bright white appearance on gray-scale IVUS images) (Fig. 2). This could lead to two potential biases: (1) an observer related bias as the outer vessel boundaries are shadowed and thus the outer vessel contour must be interpolated by the observer and (2) a software related bias as the signal processing tools are not taking into account these shadowed areas and will assign the pixels in these regions to any of the tissue components within their database based on the “noise” of the ultrasound waves related to that area. In contrast, the image-based method of echogenicity at first examines every cross-sectional IVUS image if it contains acoustic shadowed areas and classifies tissue within these areas as unknown preventing possible software related deviations.
Although IVUS is proven to be safe and is a well-established method widely available, there are other promising imaging methods, such as optical coherence tomography, but for the time being they are not capable identifying quantitatively tissue components comparable to the described methods in this paper.
In-depth post-hoc analysis of IVUS data can be applied to quantify coronary plaque composition and possible changes of this composition over time which can be aplied to evaluate new therapeutic treatment methods. However, there is still a large scientific debate how these different analysis methods exactly relate to each other and to that of the golden standard of histopathology. If these methods are also capable of identifying possible vulnerable segments is still under investigation.
This article is distributed under the terms of the Creative Commons Attribution Noncommercial License which permits any noncommercial use, distribution, and reproduction in any medium, provided the original author(s) and source are credited.
Conflict of interest