Principles and Device Developments

Ultrasound-Based Approaches
IVUS systems produce tomographic images by performing a series of pulse/echo sequences, or vectors, in which an acoustic pulse is emitted and the subsequent reflections from the tissue are detected. Each vector is acquired by directing the ultrasound beam from the catheter in a slightly different direction from the previous vector by mechanical or electrical means. A gray-scale IVUS image is made with all the vectors (commonly 256 vectors), with each vector acquired at a different angle of rotation.

Several clinically relevant properties of the ultrasound image, such as the resolution, depth of penetration and attenuation of the acoustic pulse by tissue, are dependent on the geometric and frequency properties of the transducer. A crystal transducer emits a signal that spans a range of frequencies. The higher the center frequency, the better the radial resolution (Figure 1) but the lower the depth of penetration. Conventional IVUS catheters used in the coronary arteries have center frequencies that range from 20 MHz to 40 MHz, thus providing theoretical lower limits of resolution (calculated as half the wavelength) of 39 and 19 µm, respectively. In practice, the radial resolution is at least 2 to 5 times poorer, as determined by factors such as the length of the emitted pulse and the position of the imaged structures relative to the transducer.

View larger version (23K): [in a new window] Figure 1. Higher-frequency IVUS. IVUS images of a coronary artery at (A) 30 MHz, (B) 40 MHz, and (C) 50 MHz. The higher center-frequency images have better radial resolutions (theoretically around 70 µm at 50 MHz).

For peripheral and intracardiac echocardiographic (ICE) applications, larger imaging catheters with lower center frequencies (8 to 20 MHz) are produced in both mechanical and electrical configurations. In addition, a phased-array intracardiac echocardiography catheter is available that provides a sector image with color/spectral Doppler and multiple frequency imaging (5.0 to 10 MHz) capabilities.

Optical Approaches
Angioscopy

Intracoronary angioscopy is an endoscopic technology that allows direct visualization of the surface color and superficial morphology of atherosclerotic plaque, thrombus, neointima, or stent struts. The light source emits a high-intensity white light to illuminate the target object through the fiber optic catheter. The imaging catheter contains a flexible fiber optic bundle of several thousand pixels; the latest-generation catheter, which incorporates 6000 fibers, is 0.75 mm in outer diameter with a microlens that provides a 70° field of view and a focused depth that ranges from 1 to 5 mm. Although conventional delivery systems were equipped with a distal balloon to create a blood-free field for optical imaging, an alternative system uses a smaller catheter to continuously flush an optically clear liquid in front of the tip of the angioscope for transient blood displacement. To circumvent the subjectivity of color interpretation, quantitative colorimetric methods have been proposed. One research group is also developing a side-view imaging catheter to overcome several inherent limitations of the conventional forward-view catheter configuration (Figure 2). 

View larger version (18K): [in a new window] Figure 2. In vivo stent images in a coronary artery obtained by a prototype side-view angioscopy system. This system enables straight observation of vessel surface structures, whereas perspective representation by conventional forward-view angioscopy often visualizes only the proximal aspect of protruding objects. Images courtesy of Dr Jun-ichi Kotani.

Optical Coherence Tomography
Optical coherence tomography (OCT) generates real-time tomographic images from backscattered reflections of infrared light. The greatest advantage of this optical technology is its resolution that is significantly higher (10 to 15-µm axial resolution and 20 to 25-µm lateral and out-of-plane resolution) than ultrasound-based approaches. This improvement comes at the expense of poorer penetration through blood and tissue (1 to 3 mm). The imaging catheter includes a fiber optic core with a microlens and a prism at the distal tip to generate a focused scanning beam perpendicular to the catheter axis. The current OCT catheter has a stationary outer sheath so that the inner imaging core can rotate safely to provide cross-sectional images. This design also facilitates an automated pullback of the imaging core. At present, 1 company (LightLab Imaging Inc, Westford, Mass) has commercialized intravascular OCT technology by providing dedicated imaging wires and occlusion balloon catheters.

In standard OCT systems (time-domain OCT), the optical engine includes a superluminescent diode as a source of low-coherent infrared light, typically with a wavelength around 1300 nm. The high propagation speed of light requires OCT to use interferometric techniques to determine the depth of the reflector.

To reduce ischemia during blood-free optical imaging, several groups have been developing rapid-scan OCT systems, referred to as Fourier-domain OCT or optical frequency domain imaging.³ This technique measures optical echo time delay by use of a light source whose light output can be rapidly swept over a range of wavelengths (eg, 1260 to 1360 nm). Fourier transform techniques enable conversion of the frequency domain (or wavelength-dependent) data to a time-domain representation. Although conventional time-domain systems have a frame rate of 4 to 20 frames per second, frequency-domain OCT achieves acquisition at 80 to 110 frames per second, which allows for comprehensive scanning of long arterial segments during a short balloon occlusion or even 1 bolus liquid flush without occlusion. The first clinical study with this technology is being initiated to investigate vulnerable plaque hypotheses in a prospective multicenter fashion.

Intravascular Spectroscopy
Spectroscopy determines the chemical composition of plaque substances based on analysis of spectra induced by interaction of light with the tissue materials. To date, several forms of photonic spectroscopy have been adapted for tissue characterization, including diffuse reflectance near-infrared (NIR), Raman, and fluorescence spectroscopy. When tissues are exposed to a light beam that contains a broad spectrum of wavelengths, wavelengths absorbed by the illuminated molecules will be missing from the spectrum of the original light after it has traversed the tissue. Diffuse reflectance NIR spectroscopy analyzes the amount of this absorbance as a function of wavelengths within the NIR window (700 to 2500 nm; Figure 3). In contrast, Raman spectroscopy uses a light beam of a single wavelength and monitors shifts in wavelength as some of the incident photons interact with the molecules in the tissue and gain or lose energy. Raman spectroscopy measures this inelastic scattering (shift in wavelength) because it contains unique information on the substance with which the photons interacted. Under a certain condition, the photons can excite molecules to a higher energy level, the decay from which releases the energy difference in the form of light. Fluorescence spectroscopy utilizes photoluminescence or luminescent emission to identify the properties of the tissue being illuminated.

View larger version (21K): [in a new window] Figure 3. NIR absorbance spectra from 4 chemical components. Spectral analysis uses the overall shape of the spectra over a range of wavelengths. In this example, the regions around 1200 nm separate the cholesterols from the collagens, whereas the regions around 1500 nm provide more discrimination among the cholesterols. Reproduced from Caplan et al. Journal of the American College of Cardiology. 2006;47:92–96, with permission from the publisher. Copyright © 2006, Elsevier.

For diffuse reflectance NIR spectroscopy, a coronary catheter system is under clinical evaluation for commercial use (InfraReDx, Inc, Cambridge, Mass). The 3.2F rapid-exchange NIR catheter contains fiber optic bundles for delivery and collection of light within a protective outer sheath. The catheter directs the light to the vessel wall with a mirror located at the tip, and spectra acquisition can be done at 5 ms through flowing blood.

Because of the low intensity of Raman light scattered from the vessel wall, the development of intravascular Raman spectroscopy systems is technically more challenging. One research group has developed a fiber optic probe that consists of a central fiber for laser delivery surrounded by 15 fibers for sample light collection.⁴ Both fibers have a dielectric filter to block the Raman signal generated by the fiber material. The system uses an 830-nm diode laser as a light source, and the data acquisition time has been reported as 1 second in an in vivo study. To circumvent blood interference, another group developed a real-time Raman system with prototype catheters equipped with an eccentric balloon or a basket configuration, which enables gentle contact of the probe against the vessel wall. A prototype intravascular NIR fluorescence probe (0.018-in single fiber) is also being tested with animal models.

Magnetic Resonance Imaging–Based Approaches
Magnetic Resonance Imaging (MRI) evaluates tissue with an electromagnetic radiofrequency pulse application within a strong static magnetic field. The main advantage of MRI compared with other imaging modalities is its ability to achieve strong contrast between soft tissue components (Figure 4). However, this plaque characterization is currently limited to relatively large or superficial arteries owing to the significant falloff of signal-to-noise ratio in deeper regions imaged when an external coil is used. Cardiac motion is also a significant factor that limits the resolution of coronary artery imaging with MRI. The placement of an MRI probe within the artery is one possible solution, because it would allow a high signal-to-noise ratio at the level of the arterial wall. The addition of an MRI coil to a catheter device is useful not only for imaging but also for tracking catheter position in real-time MRI interventions.

View larger version (90K): [in a new window] Figure 4. Comparison of histology (left), IVUS (middle), and MRI (right) images of human coronary specimens. Top, Atheromatous (arrows) plaque with a fibrous cap and luminal thrombus. Bottom, Calcified plaque. MRI discriminates the tissue types better than IVUS. Deep calcium behind the superficial calcium is also visualized by MRI. T2W indicates T2 weighted; PDW, proton density weighted; and ca, calcium.

The most common approach to intravascular MRI (IV-MRI) is the combination of an intravascular receiver coil and an external MRI scanner. Using balloon-injured hyperlipidemic rabbits, one group reported the feasibility of a nonobstructive IV-MRI coil for in vivo imaging of atherosclerosis.⁵ The intravascular probe was based on a single-loop receiver coil design and included a lumen for a 0.014-in guidewire. More recently, a significantly miniaturized detector coil (a 0.030-in wire-based device) has been validated in ex vivo and in vivo human iliac arteries.⁶ This device is made from nitinol tubing with mechanical properties similar to standard guidewires. A theoretical safety concern is the heat generation that can occur under certain circumstances during imaging; however, a recent rabbit study using a guidewire-based IV-MRI showed no abnormal changes of coagulation factors, clinical manifestations of blood coagulation disorders, or histopathological thermal damage in target vessels.⁷

Another interesting approach is to incorporate both magnets and coils within the catheter, which would permit standalone imaging with no external scanner. The current system (TopSpin Medical, Lod, Israel) is specifically designed for tissue characterization, providing a color-coded tissue component map, rather than anatomic imaging. The first-generation probe used in pilot clinical evaluation is 1.73 mm in diameter and provides depth-of-view imaging in the radial plane of 250 µm, lateral resolution of a 60° sector, and slice thickness of 2 mm (Figure 5). To eliminate motion artifacts, the probe needs to be stabilized against the arterial wall by gentle inflation of a partially occlusive side balloon. A 4.5F second-generation catheter with dual sensors at 180° angular displacement is currently being developed for faster data acquisition with no manual rotation of the catheter.

View larger version (41K): [in a new window] Figure 5. A self-contained IV-MRI catheter (left) and a sample image (right). Lipid within vessel wall is displayed in yellow. Photographs courtesy of TopSpin Medical, Inc.

Intravascular Thermography
The concept of plaque temperature measurement as a marker of local inflammation process was originally proposed on the basis of observations from human carotid endarterectomy specimens.⁸ In an ex vivo study with a needle thermistor,⁸ intimal surface temperature correlated positively with cell density (mostly macrophages) and inversely with the distance of the cell clusters from the luminal surface.

The local vessel wall temperature can be assessed by 2 approaches: thermocouple- and infrared-based measurements. The first approach incorporates a single or multiple thermal sensors at the catheter tip and requires contact of the thermal sensors with the vessel wall. The second approach uses an infrared fiber optic catheter to provide thermal imaging without direct vessel wall contact. In either platform, one technical consideration is the influence of blood flow on measured temperature. Given a large volume of coronary blood flow, this "cooling effect" can significantly impact the acuity of this modality to detect a small thermal heterogeneity, especially in the presence of turbulent flow.^9,10

	Current Clinical Applications and Limitations

Of all the technologies, IVUS is the most mature and widely used intravascular imaging technique. To date, IVUS has provided significant insights into biologically mediated processes of the vasculature, such as the extent of plaque burden, vascular remodeling, and restenosis in patients with coronary artery disease. This technology has also been established as a significant clinical tool in the evaluation and guidance of interventional techniques, including balloon angioplasty, atherectomy, conventional stenting, brachytherapy, and most recently, drug-eluting stents (DES).

Over the past year, the use of IVUS has grown considerably, driven by recent public concerns about the safety of DES. In the era of bare-metal stents, a number of studies demonstrated that IVUS guidance could improve the outcomes of stenting by identifying several morphological or morphometric risk factors of stent thrombosis or restenosis. Importantly, recent IVUS studies have suggested that the majority of these risk factors, such as stent underexpansion and residual reference disease, continue to be significant determinants of DES failure.^11–14 On the other hand, late-acquired incomplete strut apposition has been reported more frequently in DES than in bare-metal stents.^15,16 Although a recent clinical study indirectly suggested a possible link of this IVUS finding with late DES thrombosis,¹⁷ its exact role in the pathogenesis of this rare clinical event remains unknown.

Intracoronary angioscopy is another established diagnostic modality that clinical investigators worldwide have been using to gain additional insights into the pathophysiology of coronary lesions. Because of the high sensitivity of angioscopy to detect intraluminal thrombus in vivo, as well as its unique ability to evaluate the surface color of plaque, this imaging technology has found an important clinical niche, particularly in the field of vulnerable plaque and acute coronary syndromes. More recently, several investigators have successfully used this technology for macroscopic evaluation of neointimal coverage and local thrombus formation over DES struts in clinical settings.^18–20

Both imaging modalities, however, have several technical limitations. Limited spatial resolutions do not allow direct visualization of microstructure, including endothelialization of stent surface, a thin fibrous cap over a lipid core, or microvasculature within the vessel wall. Accurate 3-dimensional (3D) image orientation, precise discrimination of tissue components, and functional assessment of vessel wall in terms of biomechanical properties or plaque activity (atherogenic or inflammatory process) would also facilitate better understanding of vascular pathology and the optimization of interventional strategies.