The potential of intravascular photoacoustic (IVPA) imaging to detect atherosclerosis was previously demonstrated using a 532 nm nanosecond pulsed laser and an intravascular ultrasound (IVUS) imaging catheter. However, to differentiate vulnerable plaques, the composition of plaques needs to be imaged. Therefore, we introduce a multi-wavelength photoacoustic imaging method to distinguish various types of plaques. Multi-spectral IVPA imaging of ex vivo samples of normal and atherosclerotic rabbit aorta was performed at several wavelengths within 680–900 nm range. The spectral variation of photoacoustic response was extracted and a spectroscopic analysis was performed. The results of our preliminary study suggest that the spectroscopic intravascular photoacoustic imaging technique can be used to differentiate fibrous and lipid components of the atherosclerotic plaques.
© 2008 Optical Society of America
The vulnerability of an atherosclerotic plaque to rupture causing acute coronary syndromes is primarily dependent on the plaque composition. The major contribution to a vulnerable plaque is extracellular lipid accumulation. However, the inflammatory response to atherosclerosis is often responsible for the development of several other types of vulnerable plaques with varied composition [1, 2]. The complex biochemical pathways following hypercholesterolemic conditions could potentially determine the fate of an atherosclerotic lesion (calcification, collagen synthesis or degradation, blood thrombus, endothelial degradation). For example, the internalization of modified lipoproteins by macrophage cells and subsequent release of matrix metalloproteinase (MMP) lead to a degradation of collagen and other connective tissue matrix proteins . In addition, the choice of appropriate therapy also requires the knowledge of plaque composition. The current challenges in the studies of plaque regression are to identify, and preferably quantify, changes in response to treatment. Therefore, there is a need for an imaging technique in interventional laboratories to detect and differentiate atherosclerosis.
Biomedical optoacoustic (also referred as photoacoustic) imaging is a technique relying on the contrast generated by the optical absorption property of the tissues [3, 4]. The imaging technique involves the detection of photoacoustic signals generated in response to absorption of pulsed laser illumination. In our earlier work [5–7], we demonstrated photoacoustic imaging for intravascular imaging applications. Such a technique, in conjunction with intravascular ultrasound has immense potential in interventional applications, specifically, vulnerable plaque detection [6–8]. The basis for the differentiation of a heterogeneous lesion using photoacoustic imaging is the difference in the optical absorption coefficients of common plaque components (lipid, water, blood, collagen) in the aortic tissue .
In this work, we investigate the potential of multi-wavelength IVPA imaging to differentiate atherosclerosis. Specifically, a rabbit model of atherosclerosis was utilized for this ex vivo study. The laser excitation wavelength range of 680 nm – 900 nm was chosen since the optical absorption coefficient of fat shows a characteristic increase in this range [10, 11]. In addition, this wavelength range is appropriate for in vivo implementation of IVPA imaging where luminal blood absorption is minimized. The wavelength-dependent variation of the photoacoustic signal magnitude could possibly encode the spectral information specific to a constituent of the plaque. A simple, derivative-based spectroscopic analysis of the variation of the time-resolved IVPA signals was employed to test the multi-wavelength photoacoustic imaging approach. The spectroscopic IVPA images of the atherosclerotic rabbit aorta are presented and compared with images obtained from the normal aortic tissue. The preliminary results indicate the potential of IVPA imaging to detect the lipid and fibrous components of the vulnerable plaque.
2. Materials and Methods
The photoacoustic imaging studies were performed on a section of a plaque-rich aorta obtained from a rabbit subjected to a 0.15 % cholesterol diet for 10 months. The low cholesterol dietary regimen was utilized to induce fibro-cellular lesions comprised of inflammatory macrophage cells, lipids and deposits of thick collagen type I. The excised tissue was stored in saline for no more than 4 hours prior to the imaging experiments. The control sample for the imaging studies was obtained from a rabbit subjected to a normal diet for the same period of time.
The laboratory prototype of the intravascular photoacoustic (IVPA) imaging system designed for the ex vivo studies was described in detail previously [5–7]. Briefly, a tunable (λ=680 - 900 nm) pulsed Nd:YAG pumped optical parametric oscillator laser source (Vibrant B, Opotek, Inc.) provided the optical illumination for photoacoustic imaging. The collimated laser beam was directed to the tissue sample from the outside using an optical system containing a prism and lens. The optically induced acoustic response generated through a temporal thermal confinement of laser energy was detected using a coronary IVUS imaging catheter (40 MHz, 2.5 French, Atlantis™ SR Plus, Boston Scientific, Inc). Initially, the optical beam and photoacoustic sensing element were somewhat aligned to obtain the maximum photoacoustic signal from the aorta. Subsequently, the intravascular ultrasound (IVUS) and photoacoustic (IVPA) scans were performed by incrementally rotating the sample. A complete rotation of the sample containing 250 photoacoustic beams was obtained to image an entire cross-section of the aorta. Each photoacoustic A-line signal (or beam) was digitized at a sampling rate of 200 MHz. The IVPA scanning was repeated for multiple laser wavelengths (680, 700, 720, 740, 760, 780, 800, 850 and 900 nm). The location of the imaging plane was noted and the imaged cross-section was sliced for histological examination. The cross-sections were stained using Hematoxylin and Eosin (structural information of plaque deposition), RAM11 (macrophage cells) and Picrosirius red (collagen). The histology images obtained from these staining methods were examined under a microscope and digitized for correlation with the IVPA images.
For IVPA imaging, the photoacoustic response in the polar system of coordinates (angle and depth) was corrected for the depth dependent attenuation of optical energy using an exponential time gain compensation curve. Further, the signals were normalized to compensate for the wavelength dependence of the laser source energy. After applying a digital band pass filter (25 MHz–45 MHz), the signals were scan converted to the Cartesian system of coordinates and visualized using the same display dynamic range of 45 dB. To analyze the spectral variation in the absorption of the various plaque and tissue constituents, a simple method utilizing linear approximation of wavelength dependence of the optical absorption was employed. The first derivative was computed using the finite difference approach
where s is the magnitude of the photoacoustic response, λi is the wavelength at which the derivative is computed and λj is the reference wavelength. The finite approximation was used to obtain a coarse estimation of the first derivative between two discrete wavelengths.
For detailed analysis of the multi-wavelength photoacoustic response in the regions of interest (ROI) demarcated by histology, the relative energy (integral or area under the curve) of the original photoacoustic response was calculated. A 1D trapezoidal numerical integration was evaluated for the envelope of the IVPA response in each of the 20 beams (28.8° azimuthal) covering a depth of 30 samples (225 µm radial). These values of the relative energy were estimated for multiple wavelengths within 680 – 900 nm range.
Finally, the IVPA imaging and spectroscopic analysis were repeated for the control aorta that contained no obvious plaques and compared with the images obtained from the atherosclerotic aorta.
The IVPA images obtained at multiple wavelengths between 680 nm and 900 nm are presented in Fig. 1. The cross-sectional images of the aorta indicate a significant thickening of the wall as a result of the plaque deposition. These images are shown using the same display dynamic range and a 6.75 mm diameter field of view. A characteristic feature in the IVPA images is the change in the amplitude of the photoacoustic signal at different wavelengths. For example, the region in the image at 2 o’clock shows minimum photoacoustic response at 680 nm with progressive increase in the signal magnitude towards 900 nm. The histology of the cross section of the aorta is also presented in Fig. 1. The H&E, RAM11 and Picrosirius red stained images show significant plaque formation with intimal thickening leading to 20% stenosis, presence of macrophage foam cells in a lipid pool and fibrous deposits of collagen (type I and III). There is a good correlation between histology and IVPA images.
The IVPA images of the control aorta covering a field of view of 6 mm are presented in Fig. 2. In comparison to the images of the atherosclerotic aorta, the normal aorta shows a thin wall with nearly homogeneous photoacoustic response. Further, the difference in the magnitude of the photoacoustic signal at various wavelengths of laser excitation is minimal. The histology images (H&E, RAM11 and Picrosirius red) confirm the absence of plaques and the presence of thin fibrous collagen type III.
Figure 3 presents the spectroscopic IVPA images of the atherosclerotic and normal aorta computed using the finite differences approach at λi=900 nm and λj=680 nm. The map of the plaque-rich aorta is fairly heterogeneous with positive and negative values of the first derivative. For example, three regions with distinctly different slopes are indicated by the arrows in Fig. 3(a). The region 1 contains positive slope due to an increase in the absorption of light with increase in the optical wavelength. The region 3 with negative slope indicates a decrease in optical absorption coefficient while the region 2 with negative slope and values close to zero suggests no significant change in optical absorption. Compared to the atherosclerotic vessel, the spectroscopic image of the control aorta in Fig. 3(b) is nearly homogeneous with regions largely containing slope values close to zero.
Furthermore, the energy of the photoacoustic response in several representative regions of atherosclerotic vessel and normal aorta were analyzed (Fig. 4). Clearly, there are significant spatial and spectral variations in the energy of the photoacoustic signal within the plaque-rich vessel. The spatially-averaged relative photoacoustic energy in the regions 1 and 3 in Fig. 4(a) correspond to areas with positive and negative slope respectively. Further, the energy of the photoacoustic signal in region 2 in Fig. 4(a) show minimal variation and is characteristically similar to the values obtained in the regions 4 and 5 in Fig. 4(b).
4. Discussion and Conclusions
Intravascular photoacoustic images of atherosclerotic aorta (Fig. 1) indicate the presence of plaques and are clearly different from the IVPA images of normal aorta (Fig. 2) although the composition of the plaque is not easily discernable from the photoacoustic images. However, the spectroscopic IVPA images (Fig. 3) and the variation of the energy of the photoacoustic response (Fig. 4) identified the constituents of the plaque and tissue as evident from the histological slices of the tissue.
Notable in the spectroscopic photoacoustic images is the presence of regions with different slopes indicating an increase or decrease in the absorption with wavelength. The optical absorption coefficient of lipids increases towards 900 nm [10, 11]. Therefore, to target the detection of lipids which is the major constituent of vulnerable plaques, we specifically chose to evaluate the spectroscopic behavior of the photoacoustic signal near 900 nm where the differences in slopes are largest. Our observation of positive slope values in the spectroscopic IVPA images and the presence of lipids in the histology suggest that the intravascular photoacoustic spectroscopy could possibly detect fat deposits in the plaque. Further, the regions of negative slope in the spectroscopic IVPA images correlate with the thickened fibrous deposits of collagen type I. This structural change in the collagen is an important indicator that determines the stability of a plaque. The rest of the plaque and the normal aorta contain thin fibrillar structures of collagen type III and the spectroscopic IVPA images in these regions show a negligible change in the slope (Fig. 4). The role of collagen in the photoacoustic signal generation in the 680 nm – 900 nm wavelength range is not yet clear as the fibrous proteins are highly scattering with a peak absorption in the ultraviolet range of the spectrum . However, there is a possibility of a heterogeneous mixture of water, collagen and elastin producing photoacoustic signals of a relatively low magnitude and negligible change with wavelength. The thickening of the collagen and an absence of smooth muscle cells in the plaque during inflammation could possibly vary the optical properties of the lesion. The wavelength dependence of the relative energy obtained by the integral of the photoacoustic signal corroborates the observation in the first derivative images.
In this study, the photoacoustic response of the lipid and fibrous collagenous components of the vulnerable atherosclerotic lesion were investigated. The multi-wavelength IVPA imaging technique could be useful in monitoring the response of the lesion to lipid lowering therapy as well as the staging of atherosclerosis. Furthermore, IVPA imaging utilizing the variation in the optical absorption of hemoglobin may play a major role in the detection of thrombus.
In conclusion, we obtained multi-wavelength intravascular photoacoustic images of the excised sample of an aortic tissue with plaques. The wavelength dependent change in the optical absorption resulted in the variation in photoacoustic response from the diseased atherosclerotic aorta. The difference in the slopes of this variation of photoacoustic response generated the contrast to differentiate and identify the lipid and collagen in the plaque. These initial investigations using spectroscopic IVPA imaging of plaques suggests the possibility to simultaneously image and differentiate plaques.
This work was partially supported by the American Heart Association under grant 0655033Y and National Institutes of Health under grants EB004963 and HL084076. The authors would like to acknowledge the technical support from Boston Scientific, Inc., and Mrs. Srivalleesha Mallidi and Mrs. Patty Richards for help with the animal experiments.
References and links
3. A. A. Oraevsky and A. A. Karabutov, “Optoacoustic Tomography,” in Biomedical Photonics Handbook, T. Vo-Dinh, ed. (CRC Press, 2003), pp. 34–31–34–34.
4. P. C. Beard and T. N. Mills, “Characterization of post mortem arterial tissue using time-resolved photoacoustic spectroscopy at 436, 461 and 532 nm,” Phys Med Biol 42, 177–198 (1997). [CrossRef] [PubMed]
5. S. Sethuraman, S. R. Aglyamov, J. H. Amirian, R. W. Smalling, and S. Y. Emelianov, “Development of a combined intravascular ultrasound and photoacoustic imaging system,” Proceedings of the 2007 Photonics West Symposium, Photons Plus Ultrasound: Imaging and Sensing Conference 6086, F1–F10 (2006).
6. 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, 978–986 (2007). [CrossRef] [PubMed]
7. S. Sethuraman, J. H. Amirian, S. H. Litovski, R. W. Smalling, and S. Y. Emelianov, “Ex vivo characterization of atherosclerosis using intravascular photoacoustic imaging,” Opt. Express 15, 16657–16666 (2007). [CrossRef] [PubMed]
8. S. Sethuraman, S. Mallidi, S. R. Aglyamov, J. H. Amirian, S. Litovsky, R. W. Smalling, and S. Y. Emelianov, “Intravascular photoacoustic imaging of atherosclerotic plaques: ex vivo study using a rabbit model of atherosclerosis,” Proceedings of the 2007 Photonics West Symposium, Photons Plus Ultrasound: Imaging and Sensing Conference 6437, 6437291–6437299 (2007).
9. S. A. Prahl, “Optical Properties Spectra compiled by Scott Prahl,” (2001), http://omlc.ogi.edu/spectra/.
10. B. Tromberg, N. Shah, R. Lanning, A. Cerussi, J. Espinoza, T. Pham, L. Svaasand, and J. Butler, “Non-invasive in vivo characterization of breast tumors using photon migration spectroscopy,” Neoplasia 2, 26–40 (2000). [CrossRef] [PubMed]
11. R. L. P. van Veen, H. J. C. M. Sterenborg, A. Pifferi, A. Torricelli, and R. Cubeddu, “Determination of visible near-IR absorption coefficients of mammalian fat using time- and spatially resolved diffuse reflectance and transmission spectroscopy,” J. Biomedical Opt. 10, 0540041–0540046 (2005). [CrossRef]