A non-invasive method of imaging laser irradiated blood vessels has been achieved using Color Doppler Optical Coherence Tomography (CDOCT). This method may increase understanding of the mechanisms behind treatment of vascular disorders. The CDOCT system used a 1280 nm center wavelength superluminescent diode. A 585 nm, 360 μs pulsed dye laser was used to irradiate hamster dorsal skin flap window preparations. Irradiation sites were imaged with CDOCT prior to, immediately after, and 24 hours after laser irradiation. The processed CDOCT signal provided an estimate of the blood flow velocity. An increase in the blood vessel backscattered signal was observed as blood or vessel walls were coagulated. A decrease in damaged blood vessel reflectivity occurred after twenty four hours.
©1998 Optical Society of America
Pulsed dye lasers have been used to treat a variety of vascular disorders, including port wine stains and telangiectasias. The desired effect of laser treatment is destruction of the abnormal blood vessels and sparing of the surrounding tissue. Histological analysis from biopsies has historically been used to investigate the effects of laser irradiation of blood vessels. For pulsed dye lasers (PDLs) in the 577–590 nm wavelength and hundreds of microseconds pulse duration range, commonly seen effects are agglutination and extravasation of erythrocytes, and destruction of vessel endothelium and perivascular collagen [1,2]. A drawback of histology is that it is necessarily a one-time snapshot. Thus it is impossible to follow the response of a single vessel over a period of hours or days. Another limitation of histology is the mechanical artifacts introduced during tissue processing
We have used color Doppler optical coherence tomography (CDOCT) to image laser irradiated blood vessels for 24 hours. CDOCT is an augmentation of conventional optical coherence tomography (OCT), a recently developed technique which permits non-invasive imaging of biological tissues, including skin, with micron scale resolution and high dynamic range (>100 dB) [3–5]. CDOCT adds high resolution tomographic velocity mapping simultaneous with conventional OCT imaging by detecting the Doppler shift of light backscattered from moving objects [6,7]. Since a source of movement in skin is flowing erythrocytes, CDOCT is an ideal method for estimating blood flow velocity. The advantages of CDOCT are that it is non-invasive, can be repeated as often as desired, and can provide depth-resolved imaging and quantitative blood flow velocity data.
2. Materials and Methods
2.1 The dorsal skin flap window preparation
Cutaneous blood vessels were provided by hamster dorsal skin flap window preparations. The window preparation exposed 10–300 μm subdermal vessels while maintaining a full thickness of normal skin. The window preparation including surgical technique has been described previously . Briefly, a double thickness of dorsal skin was lifted and sutured to an aluminum holding fixture. A 1 cm circle of skin was cut from one thickness, exposing subdermal vessels and connective tissue of the opposite side. A glass window protected the exposed skin. Animals were anesthetized during surgery, imaging and irradiation with Ketamine and Xylazine (3:4 ratio) 0.1ml/ 100g. All procedures were performed in accordance with University of Texas at Austin Institutional Animal Care and Use Committee approved protocols.
2.2 Color Doppler Optical Coherence Tomography system
The CDOCT system (Figure 1) was similar to one described previously . The system incorporated a 1280 nm center wavelength superluminescent diode in a Michelson interferometer. Heterodyne techniques were used to measure the light reflection from a small volume of tissue. Changing the length of the interferometer reference arm allowed a-scans (depth scans) to be made, and the beam was scanned laterally across the tissue to build up a two-dimensional image. Two images were created from the coherently demodulated signal: the first (magnitude) by detecting the magnitude, and the second (Doppler) by performing a short-time Fourier transform on the signal. The latter operation yielded the signal strength at Doppler shift frequencies, then mean flow velocity was estimated from the centroid of the Doppler shift spectrogram and the angle between the sample probe optical axis and the blood vessel.
The system had a lateral resolution of 20 μm and an axial resolution of 20 μm (magnitude images) or 45 μm (Doppler images). Each image, consisting of 200 a-scans with 512 pixels each, was acquired in approximately 20 seconds. Lateral and depth (adjusted for an average tissue index of refraction of 1.4) dimensions of the images were 0.5 and 0.7 mm, respectively. Images were obtained from the window side of the preparation.
2.3 Laser irradiation
The laser used in this study was a pulsed dye laser (PDL) operating at 585 nm, 360 μsec pulse duration, and 5 mm spot size (Candela SPTL-1, Wayland, MA). Selected blood vessels of window preparations were irradiated with pulse energies above and below the threshold for blood vessel coagulation. Irradiation was performed from the window side of the preparation, so there was very little competition for laser light from the overlaying connective tissue. Photographs and CDOCT images were taken of the blood vessels before and immediately after laser irradiation, and at 1 hour and 24 hours after irradiation. In this paper, an example series of photographs and CDOCT images is shown for two irradiations of a single hamster window preparation arteriole/venule pair.
3. Results and Discussion
3.1 General observations with CDOCT imaging
The CDOCT system successfully mapped window preparation structure and blood flow velocity. A photograph of a window preparation, with an approximately 100 μm diameter venule (V) and 70 μm diameter arteriole (A), is shown in Figure 2a. The white circle represents the extent of the laser beam (5 mm diameter), and the white line shows the location of CDOCT imaging. CDOCT magnitude and Doppler images are given in Figure 2b and 2c, respectively.
Blood vessels were clearly visible in the Doppler CDOCT image of Figure 2c, and flow velocity was estimated. These vessels were not as obvious in the magnitude image of Figure 2b. This trend was evident in all the CDOCT images, and occurred because at 1280 nm the optical properties of blood are similar to surrounding tissues. The blood was slightly more absorbing so often a shadow was cast below vessels. Blood vessels were clearly seen as dark regions in the magnitude images under two conditions: 1) the blood vessel was filled with a relatively nonscattering fluid (which occurred in some damaged vessels), and 2) the light backscattered from the blood was Doppler-shifted to a frequency rejected by the CDOCT system bandpass filter (which occurred in some fast-flowing vessels). Features visible in all magnitude images are connective tissue, muscle layer (thin horizontal dark and light band), and fat.
3.2 Example Irradiations with the pulsed dye laser
The hamster window preparation shown in Figure 2 was irradiated by the PDL. When a pulse of radiant exposure 4.5 J/cm2 was applied to the window preparation, embolized coagula formed in both the venule and arteriole. These coagula were quickly flushed by moving blood through the vessels and out of the field of view of the window preparation. Subsequent photographs (Figure 3a) and CDOCT images (Figure 3b,c) of the window model confirmed that this irradiation had no appreciable effect on either the diameter or flow velocity of the arteriole and venule.
The window was observed for 1 hour to assure that the laser pulse caused no changes in blood vessel morphology or blood flow. Then, another irradiation was performed. A photograph of the window preparation immediately following a 6.0 J/cm2 pulse is presented in Figure 4a. Embolized coagula again formed in the arteriole and venule, however, these coagula were of sufficient size that the arterial coagulum became lodged at a narrower diameter branch point and stopped the flow of blood (small arrow). The venous coagulum passed from the field of view of the window preparation.
The large arrow in Figure 4a shows the location of a hemorrhage in a smaller diameter blood vessel (not imaged with CDOCT). Many of these smaller vessels appeared to be completely coagulated by the 6.0 J/cm2 pulse. Vessel coagulation and hemorrhage may be causes of purpura seen clinically during vascular treatments with the pulsed dye laser.
CDOCT images were taken at the same location as Figure 3b,c and within 10 minutes of the radiant exposure of 6.0 J/cm2. The magnitude image in Figure 4b revealed a more highly scattering region at the location of the venule (arrow), indicating that a small fixed coagulum was attached to its superficial wall. As can be seen in the Doppler CDOCT image in Figure 4c, the blood flow stopped in the venule as well as the arteriole. The window photograph gave no evidence of vessel damage sufficient to stop blood flow although the venule appeared slightly constricted in some locations.
Unlike embolized coagula in arterioles, venous embolized coagula do not frequently lodge and stop blood flow. Venous blood vessel diameters normally become larger downstream of the blood flow. However, it was possible that the venule constricted outside the field of view of the window preparation and that the coagulum lodged at that point. A combination of narrowed diameter from fixed coagulum and vessel constriction, and possible lodging of an embolized coagulum, may have been the cause of blood flow stoppage in the venule. These venous effects were only temporary, as CDOCT revealed that venule blood flow was restored to normal flow velocity within 24 hours after irradiation.
One hour after irradiation, the photographs and CDOCT images (not shown) of blood vessels were similar in appearance to those taken immediately after irradiation. The blood flow was still stopped in the arteriole and venule and the arterial coagulum was still present. Hemorrhage from several small blood vessels became more evident.
A photograph of the hamster window preparation 24 hours after irradiation is presented in Figure 5a. The arterial coagulum is still present and the ruptured small vessels have created a large region of hemorrhage.
CDOCT images were taken at the same location as Figure 4b,c 24 hours after the radiant exposure of 6.0 J/cm2. The small fixed venous coagulum no longer appeared in the magnitude CDOCT image of Figure 5b. The arteriole (A) can be seen in the magnitude image as a low-reflectance region. This is because the erythrocytes have begun to settle out of the stagnant blood, causing the vessel to appear more clear and pinkish. The Doppler CDOCT image of Figure 5c confirms that there was no flow in the arteriole but that the venous flow has returned to normal.
4. Summary and Conclusion
A summary of the response of blood vessels to PDL irradiation is given in Table 1. Common PDL laser responses seen in this study and in over 100 other irradiations include embolized coagulum, vessel blood and vessel coagulation, diameter alterations, and hemorrhage. The latter two responses are not normally considered in analytical models but were frequent results of laser irradiation in this study.
A previous study  with 585nm, 450 μs PDL irradiation of 45–113 μm diameter chick chorioallantoic membrane (CAM) blood vessels described effects of vasoconstriction/dilation, temporary occlusion, permanent occlusion, capillary extravasation, and hemorrhage. All these effects were seen in this study, although dilation was uncommon and laser-induced capillary extravasation was difficult to distinguish from normal surgical and healing response. The authors of the CAM study found a 50 % probability of any response and a severe response (permanent damage) at 3–5 and 5–6 J/cm2, respectively, in general agreement with our findings. They also noted that arterioles were more easily damaged than venules. Lodging of embolized coagula, as seen in this study, is a likely factor.
This study showed that CDOCT can be used to image blood vessels before and after laser irradiation, and that these images correlate with photographs of the window preparation. CDOCT is an excellent method of estimating blood flow velocity and is sensitive to changes in blood vessel diameter caused by laser-induced constriction or dilation. Magnitude CDOCT images showed coagulated material as bright regions and static, fluid-filled vessels as dark regions which cast no shadow. The hamster window preparation remains usable for over a week. Therefore, this procedure may be useful for describing the longer term process of vessel healing and vascular system response.
The authors thank Candela Corporation for the loan of the SPTL-1 laser. Funding for this research was provided in part by grants from the Office of Naval Research Free Electron Laser Biomedical Science Program (N00014-91-J-1564) and the Albert and Clemmie Caster Foundation (JKB, AJB), and the National Science Foundation (BES-9624617) (JAI).
References and links
1. J. G. Morelli, O.T. Tan, J. Garden, R. Margolis, Y. Seki, J. Boll, J. M. Carney, R. R. Anderson, H. Furumoto, and J. A. Parrish, “Tunable dye laser (577 nm) treatment of port wine stains,” Lasers Surg. Med. 6, 94–99 (1986). [CrossRef] [PubMed]
2. O. T. Tan, P. Morrison, and A. K. Kurban, “585 nm for the treatment of port-wine stains,” Plastic Reconstruct. Surg. 86, 1112–1117 (1990). [CrossRef]
3. D. Huang, E. A. Swanson, C. P. Lin, J. S. Schuman, W. G. Stinson, W. Chang, M. R. Hee, T. Flotte, K. Gregory, C. A. Puliafito, and J. G. Fujimoto, “Optical coherence tomography,” Science 254, 1178–1181 (1991). [CrossRef] [PubMed]
5. J. K. Barton, T. E. Milner, T. J. Pfefer, S. J. Nelson, and A. J. Welch, “Optical low coherence reflectometry to enhance Monte Carlo modeling of skin,” J. Biomed. Opt. 2, 226–234 (1997). [CrossRef] [PubMed]
7. J. A. Izatt, M. D. Kulkarni, S. Yazdanfar, J. K. Barton, and A. J. Welch, “In vivo bidirectional color Doppler flow imaging of picoliter blood volumes using optical coherence tomography,” Opt. Lett. 22, 1439–1441 (1997). [CrossRef]
8. Z. F. Gourgouliatos, A. J. Welch, and K. R. Diller, “Microscopic instrumentation and analysis of laser-tissue interaction in a skin flap model,” J. Biomech. Eng. 113, 301–307 (1991). [CrossRef] [PubMed]
9. S. Kimel, L. O. Svaasand, M. Hammer-Wilson, M. J. Schell, T. E. Milner, J. S. Nelson, and M. W. Berns, “Differential vascular response to laser photothermolysis,” J. Invest. Derm. 103, 693–700 (1994). [CrossRef] [PubMed]