Search tips
Search criteria 


Logo of nihpaAbout Author manuscriptsSubmit a manuscriptHHS Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
Opt Express. Author manuscript; available in PMC 2010 July 7.
Published in final edited form as:
PMCID: PMC2898895

Ultrahigh sensitive optical microangiography for in vivo imaging of microcirculations within human skin tissue beds


In this paper, we demonstrate for the first time that the detailed cutaneous blood flow at capillary level within dermis of human skin can be imaged by optical micro-angiography (OMAG) technique. A novel scanning protocol, i.e. fast B scan mode is used to achieve the capillary flow imaging. We employ a 1310nm system to scan the skin tissue at an imaging rate of 300 frames per second, which requires only ~5 sec to complete one 3D imaging of capillary blood flow within skin. The technique is sensitive enough to image the very slow blood flows at ~4 μm/sec. The promising results show a great potential of OMAG's role in the diagnosis, treatment and management of human skin diseases.

1. Introduction

Better assessment of cutaneous microcirculations may provide important diagnostic information for pathological conditions in dermatology, for example skin cancer, port wine stain treatment, diabetes and plastic surgery. The clinical and technical tools that can noninvasively image three dimensional (3D) micro-blood vessel networks within human skin in vivo are in demand. Ideally, the tools must be able to resolve the capillary blood flows within dermis, which are normally very slow (in the range of ~100 – 900 μm/s at the resting condition [1], and even slower at diseased states). In addition, such tools must be able to provide depth information with an imaging resolution at a scale of capillary blood vessel (~10 μm). To achieve this goal, several optical imaging modalities have been developed. Very popular methods are the scanning laser Doppler imaging and/or dynamic laser speckle imaging [2]. These methods are based on the Doppler effect that is induced by the moving blood cells in the micro-vessels, and more importantly, they are non-invasive. With these approaches, high flow sensitivity (~10s μm/s) is typically achieved. However, the spatial resolution is low that makes them difficult, if not impossible, to provide detailed visualization of the cutaneous micro-blood vessel networks. In addition, they do not provide depth-resolved imaging capability. Photoacoustic microscopy [3] is another promising imaging technique that provides the volumetric imaging of microcirculations. This technique relies on the transient optical energy deposition within blood (i.e., due to light absorption) and subsequent detection of acoustic emission from the blood volume to achieve blood vessel isolation for imaging. Though it has high enough penetration depth (>1 mm), the relatively low spatial resolution (~50 μm) makes it difficult to resolve the capillary blood vessel networks, which requires an imaging resolution at ~10 μm.

Optical coherence tomography [4,5], especially after its advent of Fourier domain OCT (FDOCT), is a very promising and non-invasive tool that is capable of providing high speed and high sensitive 3D imaging of biological tissues. To isolate the patent blood vessels from the tissue microstructures, numerous efforts have been paid over the past decade. An important effort is the development of phase resolved optical Doppler tomography (PRODT) [6]. This method evaluates the phase difference between adjacent A-scan OCT signals within one B-scan, which is consequently converted into the blood flow velocity. Though PRODT has been widely used, its sensitivity to blood flow is low that makes it difficult to visualize 3D microcirculations, particularly within the human skin, where the blood flow within the capillary vessels is in an order of 0.1 – 0.9 mm/s [1]. To improve the sensitivity of phase resolved OCT method, a significant effort has been made by Vakoc et al [7] who used the phase variance between adjacent B-scans to provide the blood flow imaging. Because the time interval between the adjacent B scan is relatively long (in the order of microseconds), their method is sensitive to slow flows within the capillary vessels. Although this latter approach was demonstrated to be able to provide very impressive images of cortical cerebral vasculature networks in rat, the relatively long imaging time (~25 min) restricts its application for in vivo imaging of human tissues, for example the skin, where the involuntary subject movement is un-avoidable. Besides PRODT method, some other important methods were also proposed, such as resonant Doppler flow imaging [8], joint spectral and time domain imaging [9-11], speckle variance imaging [12], phase variance contrast imaging [13] as well as the single-pass flow imaging [14,15]. However so far, none of these methods has been demonstrated for in vivo imaging of detailed microcirculations within human skin.

Originated from full range complex FDOCT [17,18], optical micro-angiography (OMAG) is a recently developed imaging modality [16]. OMAG has been successfully demonstrated for imaging cerebral blood flow in mice and rat [19, 20] and ocular blood flow [21]. Recent developments of OMAG family also added new techniques like single-pass flow imaging [14,15], and joint spectral and time domain imaging [9-11]. These previous OMAG methods have demonstrated flow sensitivity within the reach for imaging microcirculations within skin tissue beds, for example 160μm/s in [20], and 400 μm/s in [14]. However, they are still yet to be applied for imaging blood flows within microcirculation tissue beds in human dermis. The major reason for this failure may be that 1) the sensitivity to the blood flow is still low, and 2) the imaging acquisition time required for imaging the capillary flows is long for in vivo imaging (see below). However, the cutaneous blood flow in the capillaries is in the range from 100 to 900μm/s, and this value may be slower under the pathological conditions. To image the cutaneous blood flows, the sensitivity of OMAG must be improved, and its imaging acquisition time needs to be reduced.

In this report, we propose a novel extension for OMAG imaging method. In this novel approach, as opposed to the previous OMAG method, the A-line density for one B-scan is decreased, while the B-scan density is increased. The OMAG algorithm is then applied on C-scan direction (elevational direction), rather than B-scan direction (lateral direction) as in the conventional approach. We used a time interval between adjacent B scans at 3.3 ms to achieve flow sensitivity close to 4μm/s. We demonstrate that the detailed 3D micro vasculature network can be obtained by use of this novel OMAG approach.

2. Methods

The system setup used to achieve ultra-high sensitive OMAG (UHS-OMAG) is similar to that used in our previous work [20]. Here we briefly describe its main parameters. The system used a superluminescent diode as the light source, which has the central wavelength of 1310 nm and bandwidth of 65 nm that provided a ~12 μm axial resolution in the air. In the sample arm, we used a 50 mm focal length objective lens to achieve ~16 μm lateral resolution. The output light from the interferometer was routed to a home-built spectrometer, which had a designed spectral resolution of ~0.141 nm that provided a detectable depth range of ~3 mm on each side of the zero delay line. The line rate of the camera was 47,000 per second. With this imaging speed, the signal to noise ratio was measured at ~85 dB with a light power on the sample at ~3 mW.

To achieve ultrahigh sensitive imaging to the flow, we applied a novel scanning protocol in this system. Firstly, for each B scan (i.e. x-direction scan), we acquired 128 A-lines with a spacing of ~15 μm between adjacent lines, thus covering a size of ~2 mm on the tissue. The imaging rate was 300 frames per second (fps). Note that with 47 kHz line scan rate, the theoretical imaging rate should be 367 fps. The reduced imaging rate at 300 fps was due to the data transfer limitations during the handshake between the camera and the computer. Secondly, in y-direction (i.e. C scan direction), we captured 1500 B-scans over 2.0 mm on the tissue, which gave a ~1.3 μm spacing between adjacent B scans, indicating the oversampling factor of ~12 in the C scan direction. The whole 3D data volume was captured within 5 s.

The essential principle of UHS-OMAG is the same as the traditional one [20], except that the OMAG algorithm is applied on slow axis (C scan direction) rather than fast axis (B scan direction). As analyzed in [20], the interference signal of one B-scan captured by the CCD camera can be expressed as the following equation:


where k is the wavenumber; t is the timing when a A-line was captured. ER is the light reflected from the reference mirror; S(k) is the spectral density of the light source used; n is the refractive index of tissue; z is the depth coordinate; a(z, t) is the amplitude of the back scattered light; v is the velocity of moving blood cells in a blood vessel, which is located at depth z1. Because the light backscattered from the sample is quite weak compared to the light reflected from the reference mirror, we do not consider, in Eq. (1), the self cross-correlation between the light backscattered from different positions within the sample. We also do not consider the DC signals because they do not contribute to useful OMAG signals. The conventional OMAG used high pass filtering in the fast scanning axis, i.e. B scan direction, to isolate the optical scattering signals between the static and moving scatters. Thus, the detectable flow velocity is determined by the time spacing, ΔtB, between the adjacent A scans, i.e., v = λ/2nΔtB. If flow velocity in a capillary is v ≤ 100 μm/s, then it would require ΔtB ≥ ~4.7 ms for the system to have a chance to sample the blood cells flowing in the capillary. This time spacing translates into a scanning speed of ~213 A scans per second. Therefore, the total imaging time to acquire a 3D capillary flow image of a tissue volume would be prohibitively long, not ideal for in vivo imaging of capillary blood flows.

In order to image the slow blood flow within capillary vessels while keeping the imaging time at the same order as the conventional approach, we propose to perform the OMAG algorithm along the C scan direction. In this case, Eq. (1) can still be used to represent the spectral interferogram signal captured by the system, except that the time variable, t, is now corresponding to the B-scan numbers in one C-scan. With this modification, the requirement of the oversampling in the B scan direction as in the conventional OMAG system is relaxed, making it possible to have a much faster B scan imaging rate, provided that the line scan camera in the spectrometer is limited or fixed. The detectable flow velocity is determined by the time spacing, ΔtC, between adjacent B scans. In our system setup, the imaging rate is 300 fps, i.e., ΔtC ~ 3.3 ms. Therefore, considering that the C scan direction is densely sampled at an oversampling factor of 12, the detectable flow velocity would be ~ 141 μm/s while the imaging speed is still kept at 47,000 A scans per second. This detectable flow velocity would be sufficient to image the blood flow in capillaries.

In the data processing, the proposed approach first takes a differential operation on the captured B scan spectral interferograms along the C scan direction, i.e.,


where i represents the index of the B scans in the C scan direction. The differential operation is equivalent to the high pass filtering, which suppresses the optical scattering signals from the static elements within scanned tissue volume. Then, we apply fast Fourier transform (FFT) upon every wavenumber k (t is now constant) of Eq. (2) to obtain the depth resolved OMAG flow image with ultrahigh sensitivity to the flow.

The minimum detectable blood flow is determined by the system phase noise floor, which can be expressed by the intensity signal to noise ratio, S, of the OMAG/OCT system by σ Δ [var phi]2=1/S [22]. Thus, with the system signal to noise ratio at 85 dB, the minimum detectable flow velocity would be ~4.0 μm/s. However, bear in mind that if a blood cell moves at 4μm/s, the current OMAG system would not provide a continuous trajectory for this blood cell in the 3D OMAG flow image, i.e., the trajectory would be seen as a broken line.

Because the system is very sensitive to the movement, the bulk motion of the sample will seriously degrade the final image result if we directly apply the OMAG algorithm. To solve this problem, we applied the phase compensation method described in our previous work [20] onto the raw interference signal before applying the OMAG algorithm.

3. Experimental results

To test performance of the UHS-OMAG to image the blood flow, we imaged skin located at the backside of hand of a male volunteer. For comparison, we also obtained the traditional OMAG and PRODT cross-sectional flow images, for which we captured 2000 A scans over 2 mm at an imaging speed of 31,000 A scans per second in order to fulfill the oversampling requirement for these previous methods. The results are shown in Fig. 1 where the images in the top row were from the ultrahigh sensitive system while those at the bottom were from the conventional OMAG and PRODT, respectively. Figures 1(A) and 1(D) are the FDOCT structural images obtained from the captured interferograms. Although similar, they are not exactly the same due to the small subject movement when switching the system among different approaches. However, it would be sufficient to provide fair comparison of performance among ultrahigh sensitive OMAG, conventional OMAG and PRODT methods on their ability to extract slow flow information. Figures 1(B), (C), (E) and (F) are from UHS-OMAG, PRODT (based on the phase difference between adjacent B-scans), conventional OMAG and conventional PRODT (based on phase difference between adjacent A scans in one B scan) methods, respectively. In Figs. 1(C) and (F), we calculated the phase differences only when the structural signal is 15dB above the noise floor. It is apparent that the ultrahigh sensitive OMAG approach outperforms all the other methods. Fig. 1(B) shows blood flows within the papillary dermis (white arrows) where only capillary blood vessels are present, as well as the blood flows within reticular dermis (red arrows) where both the capillary and larger blood vessels are present. By calculating the phase differences between adjacent B scans, the blood flow velocities within the capillaries are within the reach of the UHS-OMAG [see the white arrows in Fig. 1(C)].

Fig. 1
In vivo cross sectional imaging of human skin: (A) B-scan structural image, and corresponding (B) UHS-OMAG and (C) PRODT flow images. (D) B-scan structural image, and corresponding conventional (E) OMAG and (F) PRODT flow images.

Because the conventional OMAG requires oversampling at the fast scanning direction (i.e., B scan), it is not sensitive to the slow blood flow within the capillaries, which are normally below 100μm/s. However, the conventional PRODT approach totally failed in imaging any of blood vessels [see Fig. 1(F)]. It should be noted that there is seen global noise ‘flow’ background in the UHS-OMAG flow image, e.g., in Fig. 2(B), which might be caused by some ‘non-moving’ scatters, such as global motion etc. In this case, the moving-scatter-sensitive optical Doppler OCT technique proposed in [23,24] may be used to further enhance the UHS-OMAG flow imaging quality.

Fig. 2
Cross sectional imaging of a flow phantom in which the intralipid scattering fluid in the capillary is not flowing. (A) B-scan structural image, and corresponding (B) ultrahigh sensitive OMAG flow image, indicating the Brownian motion of particles. White ...

In order to check whether the flow sensitivity of UHS-OMAG approaches the system phase-noise floor (in this case ~4 μm/s as stated in the last section), we used instead a highly scattering flow phantom as the imaging target. The phantom was made of the gelatin mixed with ~1% milk to simulate the background optical heterogeneity of the tissue. In making this background tissue, precaution was taken so that the mixed gel was well solidified to minimize the possible Brownian motion of particles in the background. A capillary tube with an inner diameter of ~400 μm was submerged in this background tissue and ~2% TiO2 particle solution was flowing in it that was controlled by a precision syringe pump. Although such setup can control precisely the flow velocity in the capillary tube, the flow speed as low as ~4 μm/s is difficult, if not impossible, to provide. Considering if the flow is stopped, the Brownian motion of particles is unavoidable in the capillary tube. With our experimental condition, the motion speed of particles due to Brownian motion would be randomly distributed within a range of several tens of microns per second. Due to these reasons, we decided to test whether UHS-OMAG is able to measure the Brownian motion of particles. In the experiments, the capillary tube was made almost perpendicular to the incident sample beam to avoid free fall of the scattering particles within the tube. The imaging results are shown in Fig.2, where Fig. 2(A) is the OMAG/OCT microstructural image of the flow phantom, and Fig. 2(B) is the corresponding UHS-OMAG flow image. From this result, it is clear that UHS-OMAG is able to image the particle movements due to Brownian motion while almost no signals are detected in the background region.

To examine in more detail, we used the phase-resolved technique [6, 20] applied to the adjacent B-scans of the UHS-OMAG flow images to provide the velocity image of the flow phantom above. The result is shown in Fig. 3(A), where it can be seen that the velocity values in the background region are low while those within the capillary tube are contrasted out primarily due to the Brownian motion of the particles. Fig. 3(B) shows a plot of the calculated velocities across the center of the capillary tube at the position marked as the blue line in Fig. 3(A), where the dashed box indicates the position of the capillary lumen. The velocity values of particle movements ranged from approximately -50 to 100 μm/s at this cross-line position. The standard deviation of the values outside the dashed-box region was evaluated to be ~4.5 μm/s, close to the theoretical value of ~ 4 μm/s. From this experiment, we concluded that the proposed UHS-OMAG is sensitive to the flow as low as ~4 μm/s for the system setup used in this study. We also performed the conventional PRODT imaging of the same phantom. In doing so, we set the system imaging rate at 31,000 A scans per second. And the A-line density across the B-scan of ~2.5 mm was set at 4000, indicating the spacing between adjacent A scans was 0.625 μm. The corresponding results are given in Fig. 3(C) and 3(D), respectively, where it is clear that PRODT was totally failed to image the Brownian motion of the particles under the current experimental setup. Note also that the standard deviation of velocity values shown in Fig. 3(D) was ~180 μm/s, thus it is not surprising that PRODT is not able to achieve satisfactory imaging performance.

Fig. 3
Assessment of UHS-OMAG sensitivity to the flow as compared to PRODT. (A) velocity image obtained from the flow phantom assessed by UHS-OMAG, (B) plot of the velocity data across the capillary tube at the position shown as the blue line in (A). (C) and ...

After evaluation of the UHS-OMAG imaging performance, we next show its capability to image the capillary blood flows within dermis in 3D. Figure 4(A) shows a schematic drawing of the blood vessel system of the human skin, in which an interconnected network of vessels is characterized by regular structures on all levels. The human skin is composed of the cutis and the subcutis (hypodermis [HD]). The cutis is further divided into the epidermis (EP) and the dermis (DR). The dermis and subcutis are pervaded with a complex system of blood vessels, while the epidermis is free of vessels. A superficial network comprises the interface between papillary (PD) and reticular dermis (RD), while a lower network is located on the border between dermis and subcutis. Vertical vessels connect both networks and thus make it complete. In the diagram, arteries are shown in red and veins in blue. To show whether the ultrahigh sensitive OMAG is able to image the blood flow within the patent blood vessels as described above, we acquired 3D blood flow images over the palm of a healthy volunteer, as shown in Fig. 4(B), where the black box indicates the scanning area (~2×2 mm2). When performing imaging, the surface of the volunteer's palm was controlled with a roughly 85 degree angle towards the incident sample beam, so that there are Doppler-induced signals, from the horizontally oriented reticular vascular network, received by the system.

Fig. 4
(A) Schematic diagram of the blood vessel system of the human skin; (B) photograph of palm showing the scanning area (2×2 mm2); (C) 3D rendered OMAG image of blood vessels together with the 3D structures; (D) enface movie showing the imaged blood ...

The 3D OMAG imaging result of blood vessel networks is given in Fig. 4(C), shown together with the 3D micro-structural image. To see the OMAG flow signals in more detail, we provide a movie in Fig. 4(D) to show the enface view flying through from the top to the bottom, where the blood flows within blood vessel systems within the skin are clearly delineated. Because the UHS-OMAG sensitivity is as low as ~4μm/s, even the dynamics of sweat glands are imaged. In Fig. 5, we provide the projection views at the different land-mark depths. Figure 5(A) gives the projection view at the depths from 400 to 450μm, which depth corresponds to the papillary dermis where the capillary vessels are dense (e.g., pointed by the arrows). At the depths from 450 to 650μm, the vessels are almost vertical that connects vessel networks between papillary dermis and reticular dermis, seen as the bright spots in Fig. 5(B). The blood vessel network in reticular dermis is given in Fig. 5(C) where the vessel diameter is seen to be smaller than those situated in the hypodermis in Fig. 5(D). These observations from ultrahigh sensitive OMAG are almost identical to that described in the standard reference [25], for example given in Fig. 4(A), demonstrating the power of the ultrahigh sensitive OMAG in the investigations of pathological conditions in dermatology.

Fig. 5
UHS-OMAG provides detailed projection views of microcirculation network at different depths of skin obtained from: (A) 400 – 450μm (closely representing papillary dermis), (B) 450 – 650μm, (C) 650-780μm (closely ...

4. Conclusion

We have demonstrated an ultrahigh sensitive OMAG system to image the volumetric microcirculation within the human skin. It was achieved by applying the OMAG algorithm along the slow scan axis (i.e., the C scan direction), as opposed to the fast axis (i.e., the B scan direction) in the conventional method. Comparing with the conventional OMAG flow image, the new method delivered much better performance to extract slow flow information. We have shown that detailed 3D microvascular images obtained from the human skin by the proposed OMAG is comparable to that described in the standard textbook. Therefore, we expect that the ultrahigh sensitive OMAG may have great value in future clinical investigations of pathological conditions in human skin.

Supplementary Material

Video file


This work was supported in part by research grants from the National Heart, Lung, and Blood Institute (R01 HL093140), National Institute of Biomedical Imaging and Bioengineering (R01 EB009682), and the American Heart Association (0855733G). The content is solely the responsibility of the authors and does not necessarily represent the official views of grant giving bodies.

References and links

1. Stucker M, Baier V, Reuther T, Hoffmann K, Kellam K, Altmeyer P. Capillary blood cell velocity in human skin capillaries located perpendicularly to the skin surface: measured by a new laser Doppler anemometer. Microvascular Research. 1996;52:188–192. [PubMed]
2. Briers JD. Laser Doppler, speckle and related techniques for blood perfusion mapping and imaging. Physiol Meas. 2001;22:R35–R66. [PubMed]
3. Zhang HF, Maslov K, Stoica G, Wang LV. Functional photoacoustic microscopy for high-resolutionand noninvasive in vivo imaging. Nature Biotechnol. 2006;24:848–851. [PubMed]
4. Fercher AF, Drexler W, Hitzenberger CK, Lasser T. Optical coherence tomography – principles and applications. Rep Prog Phys. 2003;66:239–303.
5. Tomlins PH, Wang RK. Theory, development and applications of optical coherence tomography. J Phys D Appl Phys. 2005;38:2519–2535.
6. Zhao YH, Chen ZP, Ding ZH, Ren H, Nelson JS. Real-time phase-resolved functional optical coherence tomography by use of optical Hilbert transformation. Opt Lett. 2002;27:98–100. [PubMed]
7. Vakoc BJ, Lanning RM, Tyrrell JA, Padera TP, Bartlett LA, Stylianopoulos T, Munn LL, Tearney GJ, Fukumura D, Jain RK, Bouma BE. Three-dimensional microscopy of the tumor microenvironment in vivo using optical frequency domain imaging. Nat Med. 2009;15:1219–1223. [PMC free article] [PubMed]
8. Bachmann AH, Villiger ML, Blatter C, Lasser T, Leitgeb RA. Resonant Doppler flow imaging and optical vivisection of retinal blood vessels. Opt Express. 2007;15:408–422. [PubMed]
9. Szkulmowski M, Szkulmowska A, Bajraszewski T, Kowalczyk A, Wojtkowski M. Flow velocity estimation using joint spectral and time domain optical coherence tomography. Opt Express. 2008;16:6008. [PubMed]
10. Szkulmowska A, Szkulmowski M, Szlag D, Kowalczyk A, Wojtkowski M. Three-dimensional quantitative imaging of retinal and choroidal blood flow velocity using joint spectral and time domain optical coherence tomography. Opt express. 2009;17:10584–10598. [PubMed]
11. Grulkowski I, Gorczynska I, Szkulmowski M, Szlag D, Szkulmowska A, Leitgeb RA, Kowalczyk A, Wojtkowski M. Scanning protocols dedicated to smart velocity ranging in Spectral OCT. Opt express. 2009;17:23736–23754. [PubMed]
12. Mariampillai A, Standish BA, Moriyama EH, Khurana M, Munce NR, Leung MKK, Jiang J, Cable A, Wilson BC, Vitkin IA, Yang VXD. Speckle variance detection of microvasculature using swept-source optical coherence tomography. Opt Lett. 2008;33:1530–1532. [PubMed]
13. Fingler J, Schwartz D, Yang C, Fraser SE. Mobility and transverse flow visualization using phase variance contrast with spectral domain optical coherence tomography. Opt Express. 2007;15:12636–12653. [PubMed]
14. Tao Y, Davis AM, Izatt JA. Single-pas columetric biodirectional blood flow imaging spectral domain optical coherence tomography using a modified Hilbert transform. Opt Express. 2008;16:12350–12361. [PubMed]
15. Tao Y, Kennedy KM, Izatt JA. Velocity-resolved 3D retinal microvessel imaging using single-pass flow imaging spectral domain optical coherence tomography. Opt Express. 2009;17:4177–4188. [PMC free article] [PubMed]
16. Wang RK, Jacques SL, Ma Z, Hurst S, Hanson SR, Gruber A. Three Dimensional Optical Angiography. Opt Express. 2007;15:4083–4097. [PubMed]
17. Wang RK. In vivo full range complex Fourier domain optical coherence tomography. Appl Phys Lett. 2007;90:054103.
18. Wang RK. Fourier domain optical coherence tomography achieves full range complex imaging in vivo by introducing a carrier frequency during scanning. Phys Med Biol. 2007;52:5897–5907. [PubMed]
19. Wang RK, Hurst S. Mapping of cerebrovascular blood perfusion in mice with skin and cranium intact by Optical Micro-AngioGraphy at 1300nm wavelength. Opt Express. 2007;15:11402–11412. [PubMed]
20. Wang RK, An L. Doppler optical micro-angiography for volumetric imaging of vascular perfusion in vivo. Opt Express. 2009;17:8926–8940. [PMC free article] [PubMed]
21. An L, Wang RK. In vivo volumetric imaging of vascular perfusion within human retina and choroids with optical micro-angiography. Opt Express. 2008;16:11438–11452. [PubMed]
22. Vakoc BJ, Yun SH, de Boer JF, Tearney GJ, Bouma BE. Phase-resolved optical frequency domain imaging. Opt Express. 2005;13:5483–5492. [PMC free article] [PubMed]
23. Ren HW, Sun T, MacDonald DJ, Cobb MJ, Li XD. Real-time in vivo blood-flow imaging by movingscatterer-sensitive spectral-domain optical Doppler tomography. Opt Lett. 2006;31:927–929. [PubMed]
24. Ren H, Li X. Clutter rejection filters for optical Doppler tomography. Opt Express. 2006;14:6103–6112. [PubMed]
25. Braverman IM. The Cutaneous Microcirculation. J Investig Dermatol Symp Proc. 2000;5:3–9. [PubMed]