Open Access
Add to BibSonomyAbstract
Coronary stents are the most commonly used treatment in the United States to repair narrowed or weakened arteries. The ability to visualize the stent during the stenting procedure and post-surgery is crucial to correctly place the stent with respect to the vessel stenosis, and to identify its position within the vessel wall. Current imaging modalities suffer from low contrast, resolution and/or unfavorable artifacts that can inhibit correct visualization of the stent in the artery. We demonstrated the effectiveness of a combined intravascular photoacoustic and intravascular ultrasound imaging method for high resolution and sufficient contrast imaging of commercial stents with respect to the vessel wall.
©2009 Optical Society of America
1. Introduction
Coronary stents are currently the most widely used coronary intervention in the United States. While the procedure is more than 95% successful [1], stents have brought along several unique issues including restenosis, hyperplasia and stent drift. The ability to visualize stents both during the stenting procedure and during post-surgery follow-up is important in order to correctly assess the stent with respect to the plaques and vessel, and also identify its apposition within the vessel wall. Immediately following a stenting procedure, it is important to determine the relation of the stent struts to the vessel wall. Ideally, the stent is deployed in contact with the lumen wall; however, malapposition can occur resulting in the stent detaching itself from the wall. This detachment can cause turbulent eddies to form in the vessel which can lead to thrombosis in the area of the stent. It is important when monitoring the stent to determine how much restenosis has formed around the stent struts. The distance that the stent struts are embedded into the vessel wall must be determined to assess stent viability. Currently, the most common method for assessing stent position is x-ray coronary angiography/fluoroscopy [2]. However, this procedure is problematic due to its use of ionizing radiation and possible complications in using iodinated contrast agents. Furthermore, x-ray fluoroscopy can only depict a two-dimensional projection which can lead to an underestimation of lumen diameter and the stent apposition within the lumen.
Magnetic resonance imaging has been used to image stents due to its avoidance of radiation exposure and iodine contrast agents; however, the metallic composition of stents can cause susceptibility artifacts which can obscure the stent lumen and make it very difficult to visualize the relation between the stent and the vessel wall [1, 3]. Long scan times and low resolution also remain a major limitation. Multi-slice computed tomographic angiography (MSCTA) has been shown to image much faster than MRI; however, its low resolution and artifacts in metallic stents make assessing the surrounding vessel difficult [4].
Both intravascular ultrasound (IVUS) and optical coherence tomography (OCT) have reached widespread usage in catheterization labs. IVUS can detect signal reflections from the stent struts, but has insufficient contrast to determine the struts’ position against the vessel wall [5]. Ultrasound contrast of stents is affected by the background tissue environment, which is also acoustically scattering. In addition, extraneous beams of ultrasound generated by the ultrasound transmit pulse and then scattered by the metallic stents will obscure the edges of the stent borders [6]. These blurred edges are image artifacts that can reduce the spatial registration of the imaged stent. OCT directly competes with these disadvantages with a resolution of 10-20 µm but has severe depth limitations, allowing only a penetration depth of about 2 mm [7, 8]. The presence of blood flowing through the vessel limits this depth even further, requiring clinicians to flush the vessel during the imaging procedure [5, 9]. Furthermore, the tissue behind the stent strut becomes hidden due to scattering shadows in OCT, which prevents complete diagnosis of the stent’s relation to the vessel lumen [5].
To counteract these disadvantages, we introduce a combined intravascular photoacoustic and intravascular ultrasound (IVPA/IVUS) based method for imaging clinical off-the-shelf stents with sufficient contrast, resolution, and depth penetration to visualize the stent and surrounding tissue [10]. In photoacoustic imaging the tissue is irradiated by a short laser pulse to produce a small thermal expansion and acoustic response. The light absorbed in a specific local region is converted into heat, and then converted into pressure due to the thermoelastic expansion of the tissue. An ultrasound transducer detects the produced pressure, which is linearly related to the optical absorption coefficient and the localized laser fluence. Therefore, photoacoustic imaging – a technology for remote assessment of optical absorption at sufficient depth, has all the prerequisites to become a diagnostic tool capable of providing morphological and functional information of the blood vessel [11]. Moreover, spectroscopic IVPA/IVUS imaging has shown its potential to differentiate the tissue composition in atherosclerotic plaques [12]. Thus, besides imaging stents, the same technology may be used to assess the type of tissue near the stent.
In this paper, we demonstrate the ability of IVUS and IVPA to image deployed coronary stents. Experiments were performed using commercial stents embedded into vessel-mimicking phantoms. The off-the-shelf stents were visualized quantitatively allowing for accurate measurements of stent position within the vessel. The accuracy of these measurements suggests that photoacoustic imaging of various types of metal implants embedded into tissue is possible.
2. Materials/methods
To test the feasibility of IVUS/IVPA to image stents, imaging studies were performed using 33 mm length, BX Velocity stents (Cordis) [13] deployed in two vessel phantoms mimicking both acoustic and optical scattering properties of tissue. Both phantoms were about 25 mm long with 10 mm outer diameter made of 8% polyvinyl alcohol (PVA). As prepared, PVA is an optically scattering material. For acoustic contrast, 0.1% by weight silica particles of 5 μm size were added to the phantom material to act as ultrasound scatterers.
The first phantom simulated an atherosclerotic vessel with a 3 mm inner diameter. The BX Velocity stent was embedded 1 mm into the inner lumen of the vessel, completely encased within the lumen wall (Fig. 1 ). Eight millimeters of the stent protruded from one end of the phantom.
Fig. 1 Photographs of the first vessel mimicking phantom with a stent embedded in the lumen wall. A 5.0 mm inner diameter stainless steel stent was molded 1 mm deep within the vessel wall from the lumen. An 8 mm long region of the stent was left bare outside of the phantom vessel, which was constructed from 8% cross-linked polyvinyl alcohol (PVA).
The second vessel phantom consisted of three different regions where the stent, relative to the inner lumen, was embedded, deployed (adjacent to the vessel wall), and malapposed (detached from the vessel wall) (Fig. 2 ). In this phantom, the stent was embedded approximately 1.0 mm within the vessel wall, directly adjacent to the wall, and approximately 1.0 mm malapposed from the vessel wall in the lumen in the different regions, respectively. The malapposed region was formed by separating the stent from the PVA with a plastic mold. Although the stent was separated from the PVA vessel, the stent itself was covered with a thin film of PVA due to the molding process during the phantom preparation.
Fig. 2 Cut-away diagram of the second vessel phantom consisting of three regions of varying distances between the stent and the vessel wall to model embedded (within the vessel wall), deployed (adjacent to the vessel wall), and malapposed (separate from the vessel wall) stents.
The laboratory prototype of the IVUS/IVPA imaging setup is shown in Fig. 3 . Briefly, a tunable pulsed Nd:YAG pumped optical parametric oscillator laser system (Vibrant B, Opotek, Inc.) was used at 800 nm wavelength to optically illuminate the vessel phantom (Fig. 3a). The photoacoustic signal was detected by a 40 MHz IVUS imaging catheter (2.5 French, Atlantis SR Plus, Boston Scientific, Inc) which was placed within the lumen of the PVA phantom. The laser beam and the IVUS sensing element were first aligned prior to the experiment to ensure detectibility of the photoacoustic response from the phantom. In addition to collecting photoacoustic signals, ultrasound pulse-echo signals were also collected using the IVUS imaging catheter connected to an ultrasound pulser/receiver. The cross-sectional IVUS and IVPA images were obtained by rotating the vessel and acquiring 256 photoacoustic/ultrasound A-lines. By longitudinally incrementing the phantom 248 µm each frame, a series of 100 cross-sectional images was collected to reconstruct a three dimensional image of the phantom.
Fig. 3 (a) Overall IVUS/IVPA imaging system. (b) Ex-vivo prototype photoacoustic imaging setup: the vessel was placed in a water tank and externally illuminated using an optical fiber. The ultrasound imaging catheter is placed inside the lumen and the vessel was rotated incrementally as IVUS/IVPA A-lines were collected. (c) Diagram of the integrated IVUS/IVPA imaging catheter where the IVUS probe and fiber-optical light delivery system are combined for in-vivo intravascular imaging.
In this ex-vivo prototype imaging setup, external light delivery is unrealistic for in-vivo imaging (Fig. 3(b)). During an in-vivo IVPA imaging procedure using the integrated IVUS/IVPA imaging catheter [14], light would be fed into the vessel using a fiber optical delivery system (Fig. 3(c)) where optical absorption and scattering of blood will diminish the incident laser fluence on the vessel lumen and stent. Therefore, to maximize light penetration at a given fluence and obtain sufficient image contrast, we used light at 800 nm wavelength. At this range, blood is more optically scattering than it is optically absorbing [15]. To mimic this scattering media, milk was substituted for the 30 cm−1 scattering coefficient of oxygenated blood at 800 nm. The tip of the optical fiber was placed 5 mm away from the surface of the vessel to simulate the in-vivo IVUS/IVPA imaging where light would travel through blood (scattering environment) before reaching the vessel wall. Therefore, it is conjectured that photoacoustic images of the phantom submersed into milk will be affected by light scattering similar to in-vivo imaging.
3. Results
The cross-sectional ultrasound image (40 dB display dynamic range) and photoacoustic image (15 dB range) of the first vessel phantom with the stent are shown in Fig. 4 . The photoacoustic image shows very high contrast between the stent and the background. Indeed, photoacoustic signal from the stent struts is very strong due to the high optical absorption of metal compared to the vessel which has little to no photoacoustic response. By comparison, the ultrasound image visualizes the complete cross-section of the vessel including the structure and thickness variation of the vessel wall. Since the ultrasound and photoacoustic images are already spatially co-registered, combining the two images shows complementary information. The location of the stent (IVPA image) is given in relation to the vessel (IVUS image). Quantitatively, the radial distance from individual struts to the lumen wall, measured from the IVUS/IVPA image, varied between 0.7 to 1.2 mm. These measurements are in agreement with the design of the phantom – the phantom mold was milled to separate the lumen wall from the stent by approximately 1.0 mm.
Fig. 4 (a) Ultrasound (IVUS) and (b) photoacoustic (IVPA) cross-sectional images of the stent deployed within the vessel phantom. (c) Overlay of the two images together shows the position of the stent struts with respect to the thickness of the vessel wall.
To visualize the entire vessel wall and the stent, a three-dimensional (3D) image of the phantom was produced (Fig. 5 ). To visualize the structure of the stent in the context of the structure of the phantom, the transparency of the ultrasound image was modified such that the photoacoustic signal can be seen, showing the structure of the stent.
Fig. 5 Three-dimensional (3D) reconstruction of the vessel and stent. The 3D image was created by acquiring a stack of cross-sectional images and combining them together. The ultrasound and photoacoustic signals can be displayed with different transparency in the reconstructed image to show the position and shape of the stent within the vessel.
The intravascular ultrasound, photoacoustic and combined IVUS/IVPA images of the second vessel phantom are shown in Fig. 6 . By imaging at various locations along the length of the vessel, the distance between the stent and the lumen wall was assessed. In the region of the embedded stent (Fig. 6(a)), the distance of the stent struts to the lumen wall was measured to range from 0.7 to 1.0 mm. In the region where the stent was merely adjacent to the vessel wall (Fig. 6(b)), the image visualized the correct position of the stent struts relative to the vessel wall. Finally, in the malapposed region (Fig. 6(c)) the stent struts were measured to be 0.8 to 1.1 mm away from the wall. The position of the stent struts varied due to preparation of the PVA phantom mold. However, the diameter of the stent was measured to be a constant 5.0 mm throughout the entire stent, in agreement with the specification set out by the manufacturer. Due to the molding process used in fabricating the malapposed section (Fig. 6(c)), a thin layer of PVA was formed on the surface of the stent. The presence of this PVA film explains the additional inner ring where the stent is located. Nevertheless, IVPA imaging was unaffected. The images in Fig. 6(c) demonstrate that far more complex geometries and positioning of stents within the vessel can be imaged using IVUS/IVPA imaging.
Fig. 6 Intravascular ultrasound and photoacoustic and combined IVUS/IVPA images from the three different regions in the stented vessel. (a) Stent embedded within the vessel. (b) Stent adjacent to lumen wall. (c) Stent detached from lumen wall.
Similar to the first phantom, a set of 80 cross-sectional images were combined to produce a 3D image of the vessel and stent (Fig. 7 ). By displaying the cut-away of the 3D reconstruction, the shape and position of the stent within the vessel is easily assessed (Fig. 7(d)). Again, the photoacoustic image was able to visualize the stent (Fig. 7(b)) and to correctly measure the inner diameter of the stent to be exactly 5.0 mm, the reported diameter for the stent. Furthermore, the stent could be visualized in the context of the vessel geometry (Fig. 7(c)) to identify the regions of embedded, deployed, and malapposed stent (Fig. 7(d)).
Fig. 7 3D-reconstruction of the tri-sectional phantom. Individual cross sections can show the position of the stent within the vessel. (a) Ultrasound 3D reconstruction of the phantom showing the structure of the vessel. (b) Photoacoustic reconstruction of the stent structure which can be used to assess the condition of the stent. (c) Photoacoustic image of the stent overlaid with the ultrasound image of the vessel can show the position of both. (d) Cut-away image of the reconstruction, allowing for accurate assessment of the stent within the vessel.
The experiments were also performed to identify the influence of luminal blood on the quality of photoacoustic imaging. The comparison between the blood-simulating environment and the water environment showed virtually no difference in either the ultrasound or photoacoustic images (Fig. 8 ). As expected, photoacoustic signal intensity showed a small reduction in the peak signal intensity from the stent due to the optical scattering properties of the milk. Nevertheless, image quality was not significantly reduced – the stent within the vessel wall was clearly visualized in both images.
Fig. 8 IVUS/IVPA imaging of vessel-mimicking phantom in transparent (i.e. non-scattering) and scattering medium. Ultrasound images displayed at 55 dB. Photoacoustic images in water and scattering medium displayed at 15 dB and 10 dB, respectively.
4. Discussion
The combined IVUS and IVPA imaging system was able to image several millimeters deep which is necessary for the intravascular imaging of coronary arteries. The high-frequency 40 MHz IVUS catheter allowed for an axial resolution of several tens of micrometers in both IVUS and IVPA images [10]. The resolution of IVUS/IVPA imaging is slightly worse compared to OCT imaging, but the larger imaging depth allowed for complete viewing of the 10 mm diameter vessel. Furthermore, the full structure of the phantom is visible, unobscured behind the stent struts, thus allowing the position of the stent within the vessel wall to be visible, whether or not the stent was hidden from view.
Overall, the diameter of the stent was correctly measured throughout the stent according to the specification set out by the manufacturer. An in-vivo environment can cause implanted stents to be crushed or deformed which can lead to problems such as thrombosis or stent drift. Photoacoustic imaging allows for accurate assessment of not only the size of the stent but also the stent shape in relation to the implanted environment.
The stent struts were always fully visible in the photoacoustic image, but at best only partially visible in the ultrasound images. In these regions visible in ultrasound, the phantom background is hypoechoic and therefore the stent appears in the ultrasound image with sufficient contrast against this background. In the phantoms used in this experiment, silica particles were added to provide acoustic scattering. Regions that did not contain uniformly mixed silica particles appeared hypoechoic under ultrasound. However, under conditions where the local background regions were highly acoustically scattering, the stent was difficult to visualize under ultrasound alone. Even in the malapposed region of the phantom (Fig. 6c) where traces of the stent struts can be seen, the presence of the PVA film reduced the ability to locate the struts using ultrasound alone. In the photoacoustic image, conversely, the contrast is determined by optical absorption. Since metal struts of the stent are strong light absorbers, the photoacoustic image revealed the precise location of the metal struts.
Since IVPA imaging is based on the same hardware as IVUS, it is understandable that photoacoustic imaging of stents may be susceptible to similar artifacts seen in IVUS. These include artifacts due to off-centered positioning of the transducer within the lumen and acoustic reverberation artifacts within the metal stent implant. IVPA can also bring its own issues, including light delivery through tissue or through other optically attenuating materials.
The optical attenuation of light in blood should not detrimentally affect imaging of stents in-vivo. In our experiment, the optic fiber was placed at a greater distance from the vessel surface than it would be from the inner boundary of the vessel wall in in-vivo imaging. In other words, in in-vivo experiments the light would travel a smaller distance and be attenuated less. Therefore, the resulting images in-vivo should be able to assess stent viability. Furthermore, laser pulse energies in our experiments, measured at the fiber output, were approximately 1.75 mJ. Due to the high optical absorption properties of metals, the energy used to obtain a photoacoustic response can be decreased. For IVPA imaging in real-time, necessary in the clinical environment, lower energies may be required to increase the laser pulse repetition rates. For these applications, a tunable OPO laser system may not be a suitable choice. However, pulsed Nd:YAG lasers operating at 1064 nm could be considered as these lasers can have pulse repetition rates from 1 to 10 kHz. While the pulse energy of these lasers is only upwards of 100 µJ, fluences similar to this experiment could be obtained by decreasing the fiber size or focusing the laser beam to a smaller diameter. It should be noted that smaller fiber sizes will undoubtedly be required in a clinical IVUS/IVPA system in order to meet lumen clearance diameters for the combined imaging catheter containing the IVUS transducer and IVPA optical fiber.
The use of IVUS and IVPA for imaging stents is a natural progression since IVUS/IVPA has been shown to detect and differentiate fatty plaques in atherosclerotic coronary arteries. Coronary stents are commonly used to treat blood vessels that have narrowed due to the presence of vulnerable plaques. Recent studies have shown that stent positioning can drift over time, leading to the need to detect stent shape and location with respect to the site of atherosclerosis while also determining the progression of plaque vulnerability. This makes the combined IVUS/IVPA imaging a natural and feasible method in the diagnosis and treatment of atherosclerosis. This initial study shows that IVUS/IVPA is a promising modality to image stents in vivo.
Acknowledgments
This work was partially supported by the National Institutes of Health under grants HL 096981, EB 004963 and HL 084076. The authors would like to acknowledge Mr. James Amirian and Dr. Richard Smalling of the University of Texas Health Science Center in Houston for providing the stents.
References
1. D. Maintz, R. M. Botnar, R. Fischbach, W. Heindel, W. J. Manning, and M. Stuber, “Coronary magnetic resonance angiography for assessment of the stent lumen: a phantom study,” J. Cardiovasc. Magn. Reson. 4(3), 359–367 (2002). [CrossRef] [PubMed]
2. D. R. Elgort, C. M. Hillenbrand, S. Zhang, E. Y. Wong, S. Rafie, J. S. Lewin, and J. L. Duerk, “Image-guided and -monitored renal artery stenting using only MRI,” J. Magn. Reson. Imaging 23(5), 619–627 (2006). [CrossRef] [PubMed]
3. J. Hug, E. Nagel, A. Bornstedt, B. Schnackenburg, H. Oswald, and E. Fleck, “Coronary arterial stents: safety and artifacts during MR imaging,” Radiology 216(3), 781–787 (2000). [PubMed]
4. D. Maintz, K. U. Juergens, T. Wichter, M. Grude, W. Heindel, and R. Fischbach, “Imaging of coronary artery stents using multislice computed tomography: in vitro evaluation,” Eur. Radiol. 13(4), 830–835 (2003). [PubMed]
5. Y. Kawase, K. Hoshino, R. Yoneyama, J. McGregor, R. J. Hajjar, I. K. Jang, and M. Hayase, “In vivo volumetric analysis of coronary stent using optical coherence tomography with a novel balloon occlusion-flushing catheter: a comparison with intravascular ultrasound,” Ultrasound Med. Biol. 31(10), 1343–1349 (2005). [CrossRef] [PubMed]
6. G. S. Mintz, S. E. Nissen, W. D. Anderson, S. R. Bailey, R. Erbel, P. J. Fitzgerald, F. J. Pinto, K. Rosenfield, R. J. Siegel, E. M. Tuzcu, and P. G. Yock, “American College of Cardiology Clinical Expert Consensus Document on Standards for Acquisition, Measurement and Reporting of Intravascular Ultrasound Studies (IVUS). A report of the American College of Cardiology Task Force on Clinical Expert Consensus Documents,” J. Am. Coll. Cardiol. 37(5), 1478–1492 (2001). [CrossRef] [PubMed]
7. P. Barlis, K. Dimopoulos, J. Tanigawa, E. Dzielicka, G. Ferrante, F. Del Furia, and C. Di Mario, “Quantitative analysis of intracoronary optical coherence tomography measurements of stent strut apposition and tissue coverage,” Int. J. Cardiol (2009).
8. T. L. Slottow, R. Pakala, T. Okabe, D. Hellinga, R. J. Lovec, F. O. Tio, A. B. Bui, and R. Waksman, “Optical coherence tomography and intravascular ultrasound imaging of bioabsorbable magnesium stent degradation in porcine coronary arteries,” Cardiovasc. Revasc. Med. 9(4), 248–254 (2008). [CrossRef] [PubMed]
9. I. K. Jang, B. E. Bouma, D. H. Kang, S. J. Park, S. W. Park, K. B. Seung, K. B. Choi, M. Shishkov, K. Schlendorf, E. Pomerantsev, S. L. Houser, H. T. Aretz, and G. J. Tearney, “Visualization of coronary atherosclerotic plaques in patients using optical coherence tomography: comparison with intravascular ultrasound,” J. Am. Coll. Cardiol. 39(4), 604–609 (2002). [CrossRef] [PubMed]
10. S. Sethuraman, S. R. Aglyamov, J. H. Amirian, R. W. Smalling, and S. Y. Emelianov, “Intravascular photoacoustic imaging using an IVUS imaging catheter,” IEEE Trans. Ultrason. Ferroelectr. Freq. Control 54(5), 978–986 (2007). [CrossRef] [PubMed]
11. S. Sethuraman, J. H. Amirian, S. H. Litovsky, R. W. Smalling, and S. Y. Emelianov, “Ex vivo Characterization of Atherosclerosis using Intravascular Photoacoustic Imaging,” Opt. Express 15(25), 16657–16666 (2007). [CrossRef] [PubMed]
12. S. Sethuraman, J. H. Amirian, S. H. Litovsky, R. W. Smalling, and S. Y. Emelianov, “Spectroscopic intravascular photoacoustic imaging to differentiate atherosclerotic plaques,” Opt. Express 16(5), 3362–3367 (2008). [CrossRef] [PubMed]
13. J. Butany, K. Carmichael, S. W. Leong, and M. J. Collins, “Coronary artery stents: identification and evaluation,” J. Clin. Pathol. 58(8), 795–804 (2005). [CrossRef] [PubMed]
14. B. Wang, A. B. Karpiouk, and S. Y. Emelianov, “Design of catheter for combined intravascular photoacoustic and ultrasound imaging”.Proceedings of the 2008 IEEE Ultrasonics Symposium 1150–1153 (2008).
15. D. J. Faber, M. C. Aalders, E. G. Mik, B. A. Hooper, M. J. van Gemert, and T. G. van Leeuwen, “Oxygen saturation-dependent absorption and scattering of blood,” Phys. Rev. Lett. 93(2), 028102 (2004). [CrossRef] [PubMed]
