FULL ARTICLE

In vivo blood flow imaging of inflammatory humanskin induced by tape stripping using opticalmicroangiography

Hequn Wang**; 1, Utku Baran**; 1; 2, and Ruikang K. Wang*; 1

1 Department of Bioengineering, University of Washington, Seattle, WA, USA2 Department of Electrical Engineering, University of Washington, Seattle, WA, USA

Received 29 January 2014, revised 25 February 2014, accepted 27 February 2014Published online 21 March 2014

Key words: optical coherence tomography, optical microangiography, skin microcirculation, capillary blood flow,tape stripping, inflammatory skin disorders

1. Introduction

It is known that many skin diseases are linkedto either insufficient or overactive vasculature [1].

Among them, skin inflammatory disorders, which re-present a large number of clinical cases in dermatol-ogy, are common examples. Representative inflam-matory skin diseases associated with prominent

# 2014 by WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

Journal of

BIOPHOTONICS

Vasculature response is a hallmark for most inflamma-tory skin disorders. Tape stripping on human skin causesa minor inflammation which leads to changes in mi-crovasculature. In this study, optical microangiography(OMAG), noninvasive volumetric microvasculature invivo imaging method, has been used to track the vascu-lar responses after tape stripping. Vessel density hasbeen quantified and used to correlate with the degree ofskin irritation. The proved capability of OMAG techni-que in visualizing the microvasculature network underinflamed skin condition can play an important role inclinical trials of treatment and diagnosis of inflammatoryskin disorders.

* Corresponding author: e-mail: wangrk@uw.edu** These authors have equal contribution to this article.

UHS-OMAG images of the skin at (a) before tape strip-ping, (b) 1 min, (c) 6 min, and (d) 15 min after stripping.

Scale bar: 0.3 mm.

J. Biophotonics 1–8 (2014) / DOI 10.1002/jbio.201400012

remodeling in microvasculature include psoriasis [2],rosacea [3], atopic dermatitis [4], UV damage [5],bullous pemphigoid [6], and systemic sclerosis [7].Other than skin inflammatory disorders, some sys-temic diseases, such as inflammatory bowel disease[8], rheumatoid arthritis [9], atherosclerosis [10] andasthma [11], have also shown to be associated withperipheral microvascular modifications. Due to theeasy accessibility of skin, the cutaneous microvascu-lature can provide a unique opportunity for the mea-surement of microcirculation status. Therefore, skinmicrovascular function can be considered as a prog-nostic marker when evaluating certain diseases andtreatment effects.

Topical drug delivery has become an increasinglycommon strategy for local therapy and to reducesystemic adverse effects in the treatment of skin dis-eases. However, the barrier function of stratum cor-neum (SC) results in poor penetration of drugs intothe skin, which often limits the efficacy of topicalformulations. Topical therapeutics through skin islimited in efficiency, usually only 50% of the thera-peutics can reach to the target site [12]. In order todeliver the therapeutic drug concentrations into theviable epidermal layer, improved drug permeationacross SC is required. Tape stripping is a commonlyused method to increase the penetration depth ofdrug delivery [13]. It has also been used to investi-gate the SC physiology as well as the kinetics andpenetration depth of topical drugs [14–17]. Duringthe tape stripping process, the skin is usually appliedwith appropriate adhesive tapes, and the SC is se-quentially removed by a serial number of strippingprocedures [18]. This is considered as a minimally in-vasive technique and is the basis of the FDA’s so-called dermatopharmacokinetic (DPK) approach tothe assessment of topical bioavailability and bioequi-vanlence [19].

On the other hand, tape stripping induces me-chanical disruptions of the epidermal barrier thatlead to skin inflammation [20, 21]. Similar to manyother inflammatory skin disorders, microvascular re-modeling is usually involved [22]. Surprisingly, de-spite the popularity of tape stripping in dermatologi-cal research and practices, as well as its potential ofcausing skin inflammation, little work has been doneon revealing the microvasculature changes and/or tocorrelate the degree of the vascular reactions withthe degree of damage produced. To our knowledge,there is only one prior study that has recorded theskin capillary changes after tape stripping procedure,and the authors have concluded that there is not al-ways a correlation between the macroscopic appear-ance of the stimulated area and the microscopic ana-tomic vascular changes [21]. However, this study wasperformed using capillaroscopy, which can onlyimage the very shallow capillaries and may not beable to provide a complete microvascular network.

Therefore, there is a need to visualize the detailedand complete microvasculature changes after tapestripping process.

A variety of non-invasive techniques have beendeveloped to explore skin microcirculation. Capil-laroscopy, which is essentially a method of lookingat the skin capillaries under a microscope, has beenused to directly visualize the skin surface of nailfoldarea [23, 24]. However, being a traditional light mi-croscopy, capillaroscopy has limited light penetra-tion, especially for dark skin subjects. Therefore, ca-pillaroscopy is restricted to image thin skin areasuch as nail-fold, and it is difficult to visualize thecomplete microvascular network. Other techniquessuch as orthogonal polarization spectral imaging(OPSI), which uses polarized light at l ¼ 550 nmand collects the reflected light orthogonal to theemitted light, can be used to image small blood ves-sels of nail bed or lip [25]. To date, it has been usedin applications such as assessment of the sublingualmicrocirculation in critical care patients as a strongpredictor of outcome in severe sepsis [26]. However,OPSI does not allow easy studies on blood perfusionor microvascular dysfunction. Laser Doppler ima-ging (LDI) and laser Doppler flowmetry (LDF) aremore recently developed imaging techniques, whichutilize the Doppler effect and provide a direct meas-ure of microcirculatory flow [27, 28]. Laser Dopplerbased techniques have been previously used toevaluate burns [29], dermal inflammation [30],wound healing [31], and cutaneous ulceration [32].Although they can provide large-scale relativechanges in skin blood perfusion, they have limitedimage resolution to resolve some of the small capil-laries and cannot provide absolute velocity of bloodflow [27].

Optical coherence tomography (OCT), an inter-ferometric method, is another non-invasive imagingtechnique capable of producing large-scale (milli-meters) cross-sectional morphological views of tis-sue microstructures in vivo with a micron-levelimaging resolution [33, 34]. OCT detects the scat-tered photons and the imaging contrast is originatedfrom the variance of refractive index within hetero-geneous tissue [35, 36]. Due to the relatively highresolution, deep imaging depth (1–3 mm), and thereal-time image acquisition, OCT has gained moreand more attention in the field of dermatology. Sofar, OCT has been successfully used to study non-melanoma basal cell carcinoma [37], actinic kerato-sis [38], inflammatory diseases [39, 40], to quantifyskin changes and monitor therapeutic effects [41,42]. Although of great value, traditional OCT tech-nique is difficult, if not impossible, to provide bloodvessel imaging, therefore, limiting it from studyingblood perfusion status. Fortunately, by analyz-ing both the intensity and the phase informationembedded in the OCT spectral interferograms, a

H. Wang, U. Baran, and R. K. Wang: In vivo blood flow imaging of inflammatory human skin2

Journal of

BIOPHOTONICS

# 2014 by WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.biophotonics-journal.org

new technique named optical microangiography(OMAG), has been developed to provide volu-metric vasculature image and 3D blood perfusionmap in microcirculatory tissue beds in vivo [43]. Toimprove the sensitivity to image capillaries, ulatra-high-sensitive OMAG (UHS-OMAG) has been re-cently proposed [44]. It is a variation of OMAGtechnique, which allows for imaging the completemicrovasculature network down to capillary levelby differentiating signals scattered off static tissuesfrom signals scattered off dynamic components suchas moving red blood cells (RBC) within patent ves-sels. During the last few years, OMAG techniquehas been intensively used to study microvasculatureof a variety of biological tissues in vivo. For exam-ple, UHS-OMAG has been used to study the mousecerebral microvasculature and its responses to sys-temic hypoxia, normoxia and hyperoxia [45], whereOMAG is demonstrated to have capability to visua-lize the acute hypoxia and hyperoxia on microhe-modynamic activities, including the passive and ac-tive modulation of microvascular density and fluxregulation in vivo. UHS-OMAG has also been usedto provide depth-resolved retinal microvasculatureimages within human retina, promising a non-inva-sive tool for the diagnosis of eye diseases with sig-nificant vascular involvement, such as diabetic reti-nopathy and age-related macular degeneration [46].For skin applications, Doppler-OMAG (DOMAG)[47], which uses Doppler principle has been com-bined with UHS-OMAG to image capillary mor-phology in human finger cuticle [48] and to studyskin wound healing process in mouse pinna in vivo[49]. UHS-OMAG system has also been used toinvestigate the vascular abnormalities in psoriasis[50], where the microcirculation within the normalskin sites can be differentiated from that in thepsoriatic skin sites.

Based on these previous results, we believe thatOMAG technique is advantageous for visualizingfunctional microvasculature within skin tissue beds.In this paper, we propose to apply tape stripping onin vivo human skin to produce a simple but repre-sentative skin inflammation condition and utilizeOMAG to track the microvascular changes after thisprocedure. Our aim is to explore the feasibility ofOMAG to detect changes in microvasculature underskin irritation and to introduce microvasculature-based biomarkers in evaluating the severity of skininflammation. We believe that the acquired high-quality, detailed microvascular images of the skinwill help understanding the microvasculature re-sponses to tape stripping. In addition, the capabilityof visualizing the complete microvasculature net-work under inflamed skin condition may signifi-cantly broaden the potential applications of OMAGin skin disease evaluations and in other fields of bio-medical research.

2. System and methods

2.1 OMAG system setup

All the measurements of this study were performedusing a fiber-based spectral-domain OCT system de-veloped in house. The detailed information of thissystem has been previously described in reference[51]. Briefly, the system employed an extendedbroadband superluminescent diode (Thorlabs Inc.) asthe light source, which has the central wavelength of1340 nm and bandwidth of 110 nm. The axial resolu-tion of the system was measured to be �7 mm in theair. To focus the light onto the sample, a 10X micro-scope objective lens with a focal length of 18 mm wasused to achieve a lateral resolution of �7 mm. Theoutput signals from the interferometer was directedto a home-made spectrometer system, which providesspectral resolution of �0.141 nm with a detectabledepth range of �3 mm on each side of the zero delayline. The InGaAs linescan camera (Goodrich Inc.)used in the system had an acquisition rate of 92 kHz.With a 3.5 mW illumination power exposed at thesample surface, the developed system was measuredto have a dynamic range of 105 dB. To visualize thevolumetric microvasculature, a unique UHS-OMAGscanning protocol was applied [52]. Each B-frameconsists of 400 A-lines covering a distance of�3.0 mm. The imaging rate was 180 fps. In the slowaxis (C-scan), a total number of 2000 B-frames with5 repetitions in each location were performed alsocovering a distance of �2.8 mm. ED-based clutter fil-tering algorithm [53] was implemented in MA-TLAB� to suppress the effect of tissue motion andextract microvasculature. For Doppler OMAG meas-urement to quantify the velocity of blood flow, 200B-scans and 5000 A-lines covering 1.5 mm � 1.5 mmrange were performed. The imaging speed was set at1.5 fps, which is optimized for imaging blood flow forthe finger skin. Details of this imaging protocol canbe found in [45, 54]. During each measurement, thevolunteer was sitting on a chair with his elbow sup-ported and his finger was gently fixed in a home-made finger holder. For the DOMAG experiments,the stage was tilted 20� with respect to the optical ta-ble in order to have a detectable axial component ofthe RBC velocity. Images are obtained by Dopplerprocessing of complex signals among A-lines. Phasevariance mask is used to separate Doppler flow sig-nals from noisy phase background [54].

2.2 Tape stripping procedure

A customized hand holder, which was used to sup-port the volunteer’s hand and to reduce the involun-

J. Biophotonics (2014) 3

FULLFULLARTICLEARTICLE

# 2014 by WILEY-VCH Verlag GmbH & Co. KGaA, Weinheimwww.biophotonics-journal.org

tary body movement, was designed and employed.Dorsal skin of the second finger from the right handof a 26-year-old male volunteer was measured. In or-der to remove the surface reflection, a thin layer ofmineral oil was applied the skin surface with a glasscoverslip covered on the top. For each stripping pro-cedure, the tape was applied on the finger skin for5 s and was removed rapidly. This procedure was re-peated for 15 times. Since the experiment conditions(movement from the volunteer, the amount of oil,etc.) are difficult to keep stable, data acquired after15 min is discarded. UHS-OMAG images were ac-quired at every 40 s for the first 10 min after the tapestripping procedure, and then continued to acquiredata at every 2.5 min till 15 min after the stripp-ing procedure. For Doppler OMAG measurement,images were acquired at every 5 min till 15 min afterthe procedure.

3. Results and discussion

With the superior lateral and axial resolutions of thesystem, microvascular network even at capillary lev-el can be visualized. In order to determine the de-gree of stimulation and its recovery characteristics,UHS-OMAG imaging was used to monitor the mi-crovasculature changes after applying tape strippingprocedure. Figure 1 compares the microvasculature

responses at different time points after applying tapestripping. Figure 1(a) shows a representative UHS-OMAG image acquired from the finger skin beforetape stripping. The reserve blood vessels were notactivated with most area shown as black in theOMAG image. At 1 min after the procedure, the re-serve blood vessels were strongly activated with nu-merous newly recruited capillaries as shown in Fig-ure 1(b). This activation effect lasted for �2–3 minand started to decrease �5 min after stripping. At15 min after the procedure, the blood flow seemedto get close to its normal physiology status.

The degree of stimulation should decrease withtime after applying tape stripping. To investigatewhether the imaging device has the capability to dif-ferentiate degrees of stimulation which potentiallyimplies the degree of inflammation, we monitoredthe vessel density changes over time. To calculatethe vessel density for each image, segmentation algo-rithm is applied. Briefly in this method, images arebinarized with an adaptive threshold technique spe-cifically tailored for quantifying OMAG results [55],and vessel density is calculated by dividing the num-ber of ones with the total pixel number. Figure 2shows the changes in vessel density after the tapestripping procedure was performed 15 times on thefinger skin of a 26-year-old male volunteer using anelectrical tape. Stars are measured data points whilethe curve shows the polynomial fitting result. Wecan see that the vessel density decreases rapidlyfrom �0.17 to �0.10 within the first 8 min. And thenthe skin gradually stabilizes and the vessel densityreaches �0.09 at 15 min after the procedure.

To better quantify the changes in blood flow ve-locity after applying tape stripping, Doppler OMAGimages were acquired. Figure 3 shows the compari-

Figure 1 Monitor the skin after applying tape stripping atdifferent time points using UHS-OMAG. (a)–(d) were ac-quired at before tape stripping, 1 min, 6 min, and 15 minafter tape stripping procedure, respectively. Yellow circlein (b)–(d) labels the area where obvious decrease in bloodflow was observed. Scale bar is 0.3 mm.

Figure 2 Quantification of the recovery trend after tapestripping. Tape: Scotch 3M 10 mm width electrical tape;number of stripping: 15.

H. Wang, U. Baran, and R. K. Wang: In vivo blood flow imaging of inflammatory human skin4

Journal of

BIOPHOTONICS

# 2014 by WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.biophotonics-journal.org

son of blood flow at 1 min, 5 min and 13 min afterthe procedure. The velocity range is �0.3 mm/s. Asshown in Figure 3(a) and (b), compared to 1 minafter the procedure, the vessels became less densewith a decrease in both the diameter and the velo-city at 5 min after tape stripping. At 13 min after theprocedure (Figure 3(c)), the vessels appeared to besimilar compared to at 5 min, indicating that theblood vessel response has been stabilized and theskin has gone almost back to normal physiologicalstate. Moreover, in order to quantify the overallblood flow changes in the measured skin area, totalblood flow in the corresponding area was calculatedfor each time point using the method described in[56]. Results indicated that total blood flow was�0.023 mm3/s at 1 min after the tape stripping and itdropped to �0.017 mm3/s and �0.014 mm3/s at 5 minand 13 min after the tape stripping, respectively.

To identify the location of the blood vessels andto better correlate the vessel response with the ves-sel location, UHS-OMAG images were merged withstructural OCT images for both before and aftertape stripping. Figure 4(a) and (b) are UHS-OMAGimages taken at 15 min and 1 min after tape strip-ping. The corresponding OCT structural images ta-ken at the cross section labeled in yellow dash linesin (a) and (b) are shown in Figure 4(c) and (d). Fig-ure 4(e) shows the merged image taken at 15 minafter tape stripping, where we can see that the mainflows come from the blood vessels located at thedermis of the skin. However, at 1 min after the strip-ping procedure where the skin is still under irritationstatus, significantly increased blood flow of vesselslocated at the dermal epidermal junction (DEJ) andpapillary dermis layers was observed as shown inFigure 4(f).

Progression of various connective tissue diseases,diabetes, and RP are reflected in morphological andfunctional changes in microvasculature. The abilityof visualizing structural alterations of capillaries

with RBC velocity mapping within a human fingercuticle might be a critical aid in the treatment ordiagnosis of these diseases. The presented imagingmodality enables imaging large number of capil-laries in one 3D scan which is crucial to decreasesubjectivity and to locate the same set of capillariesof patients at each visit. The velocity range is easilyadjustable by slightly changing the frame rate [54]for various clinical studies. Moreover, independentof patient’s skin characteristics, the cutaneous mi-crocirculation in human finger nail-fold area withdifferent depths can also be investigated which isnot possible with alternative methods. We believethere are certain reserve blood vessels existed inthe tissue, which are usually not functional but canbe activated to supply the needs of affected tissuewhen stimulation happens. Vascular reserve origi-nates from the description of coronary autoregula-tion by Mosher in 1964, and the control of coronaryblood flow by an autoregulatory mechanism is ex-plained in [57]. The authors claimed that blood ves-sel reserves exist in healthy subjects, and they iden-tified that coronary blood flow at rest depends onthe determinants of myocardial oxygen demand.Previous literature also suggests that the main func-tions of skin microcirculation are also guaranteedby its functional reserve [58]. Therefore, we believethat the procedure of tape stripping may activatethese reserve blood vessels. In order to comparethe vessel response within the same area, image re-gistration was performed for images acquired at dif-ferent time points. The post-registered imagesshown in Figure 2 represent blood vessels from ex-actly the same area on the volunteer’s finger skin.From the results, we can see that the overall acti-vated reserve blood vessels decreased with time. Inaddition, the decrease in blood vessel density wasnot uniform within the imaging area. For example,the area labeled by a yellow circle in Figure 1(b)–(d) showed a more obvious decrease compared to

Figure 3 Changes in blood flow velocity caused by tape stripping. (a) Doppler OMAG image of the skin at 1 min;(b) 5 min; and (c) 13 min after applying tape stripping. Stripping was performed using Scotch 3M 10 mm width electricaltape for 25 times. Red represents flow against (arterial) and green represents flow towards (venous) the incident beam.Scale bar is 0.15 mm.

J. Biophotonics (2014) 5

FULLFULLARTICLEARTICLE

# 2014 by WILEY-VCH Verlag GmbH & Co. KGaA, Weinheimwww.biophotonics-journal.org

the rest area. This could be due to the non-uniformvessel stimulation caused by tape stripping.

The image registration steps performed in thepost-processing is crucial to have a reliable quantita-tive comparison between the images of the samearea in different time points. We also tried our bestto keep all the crucial parameters, such as the focusof the probe beam and positioning and orientationof the target to be the same throughout the entiremeasurement period. Since it is not possible to keepthe measurement area and all the above-mentionedparameters to be the same, it is not practical to ob-tain the basal vessel density before the stripping pro-cedure is applied. Therefore, the imaging area in Fig-ure 1(b)–(d) may not be exactly the same areacompared to baseline (Figure 1(a)). Moreover, dueto increased discomfort in time significant body

movement from the volunteer starts to occur after15–20 min. Therefore, our monitoring period hasbeen limited to �15–20 min after tape stripping.After applying the stripping procedure, it takes�1 min to apply oil, put coverslip on, and adjust thefocus. Therefore, the first data point that we mana-ged to acquire is 1 min after the tape stripping. Fromthe curves shown in Figure 2, the skin stabilizes ataround 15 min after applying 15 times tape stripping,and the vessel density well characterizes the degreeof irritation. Other parameters, such as vessel di-ameter, may also be used to evaluate the degree ofstimulation, which may correspond to the degree ofinflammation. Moreover, with the newer develop-ment of ultrafast swept source OCT system, the im-aging time will be significantly reduced, which maypromote OCT system to be applied clinically.

From Figure 3, because of the same limitationsmentioned above, it was not practical to acquire acomparable baseline Doppler OMAG image beforethe tape stripping. Based on the results shown in Fig-ure 1–3, we believe that the skin should have al-ready settled down and gone back to its normal phy-siological condition at 15 min after tape stripping.Therefore, Figure 4(a), (c) and (e) can be consideredas baseline images representing normal skin condi-tion. We can see that the blood flow at finger skinwas not very active as shown in Figure 4(a) and (e).However, a dramatic increase in the number of func-tional capillaries and vessels was observed at 1 minafter the procedure as shown in Figure 4(b) and (f).This is because at 1 min after the procedure, the skinis still under irritation, where blood vessel responsesare still active.

Our results in general show that OMAG imagingtechnique is capable of monitoring microcirculationresponses for skin under inflamed condition. OMAGcan also be used to monitor changes in the degree ofinflammation. Inflammatory skin conditions such aspsoriasis and rosacea are considered to be associatedwith microvascular modifications [2, 3]. Therefore,OMAG has great potential to be applied to investi-gate for these skin orders. Moreover, tape strippingprocedure is heavily involved in drug delivery re-search because the removal of stratum corneum al-lows better penetration of the drugs. From our re-sult, there are clear vessel responses associated withtape stripping procedure which may cause differentpenetration efficiencies when delivering the drug atdifferent time points. Therefore, OMAG may beused to help determine the optimal time point fordrug delivery after tape stripping and eventuallyachieve the best efficacy. In the future, we plan tostudy inflammatory diseases covering different de-grees of inflammation using OMAG to further con-firm the capability of our technique in studying in-flammatory skin disorders and its related diseases.More quantification parameters/indices will also be

Figure 4 OCT structural image merged with OMAG flowimage. Maximal projection view of UHS-OMAG imagestaken at (a) 15 min and (b) 1 min after tape stripping; (c)and (d) are corresponding OCT structural images taken atthe cross section labeled in yellow dash lines in (a) and(b), respectively; (e) and (f) are merged images of theOCT structural image and the OMAG flow image taken at15 min and 1 min after tape stripping. The OMAG flow in-formation is labeled in red, while the OCT structural im-age was displayed in grayscale. The blue dash line in (c)–(f) marks the dermal-epidermal junction (DEJ). Scale baris 0.3 mm.

H. Wang, U. Baran, and R. K. Wang: In vivo blood flow imaging of inflammatory human skin6

Journal of

BIOPHOTONICS

# 2014 by WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.biophotonics-journal.org

derived to better correlate the inflammation/irrita-tion status.

4. Conclusion

In summary, tape stripping is a common procedureused both in clinic and in research. Characterizationof the detailed effects especially on the microcircula-tion changes after tape stripping will help us deter-mine the degree of inflammation caused by the pro-cedure. Our results have shown that OMAG is ableto detect microvasculature changes followed by tapestripping, therefore, indicated that the OMAG tech-nique may potentially be used to detect and differ-entiate between normal and irritated skin. More-over, the presented results have improved theunderstanding on the microcirculation dynamics be-fore and after tape stripping. In general, appropriatetape stripping causes only minimal inflammation tohuman skin. It should be considered as a relativelysafe method to study drug delivery and related to-pics.

Acknowledgments The work was supported in part byNational Institutes of Health grants (R01HL093140, andR01EB009682).

Author biographies Please see Supporting Informationonline.

References

[1] P. Carmeliet. Nat Med 9(6), 653–660 (2003).[2] I. M. Braverman. J Invest Dermatol 59(1), 91–98

(1972).[3] A. H. Gomaa, M. Yaar, M. M. Eyada, and J. Bhawan.

J Cutan Pathol 34(10), 748–753 (2007).[4] R. Agha-Majzoub, R. P. Becker, D. E. Schraufnagel,

and L. S. Chan. Microcirculation 12(6), 455–476(2005).

[5] K. Yano, K. Kajiya, M. Ishiwata, Y. K. Hong, T. Miya-kawa, and M. Detmar. J Invest Dermatol. 122(1),201–208 (2004).

[6] L. F. Brown, T. J. Harrist, K. T. Yeo, M. Stahle-Back-dahl, R. W. Jackman, B. Berse, K. Tognazzi, H. F.Dvorak, and M. Detmar. J Invest Dermatol. 104(5),744–749 (1995).

[7] O. Distler, J. H. Distler, A. Scheid, T. Acker,A. Hirth, J. Rethage, B. A. Michel, R. E. Gay, U. Mul-ler-Ladner, M. Matucci-Cerinic, K. H. Plate, M. Gass-mann, and S. Gay. Circ Res 95(1), 109–116 (2004).

[8] J. H. Chidlow, D. Shukla, M. B. Grisham, and C. G.Kevil. Am J Physiol Gastrointest Liver Physiol293(1), G5–G18 (2007).

[9] E. M. Paleolog. Arthritis Res 4(Suppl 3), S81–S90(2002).

[10] G. K. Hansson. N Engl J Med 352(16), 1685–1695(2005).

[11] D. Ribatti, I. Puxeddu, E. Crivellato, B. Nico, A. Vac-ca, and F. Levi-Schaffer. Clin Exp Allergy 39(12),1815–1821 (2009).

[12] G. Cevc. Expert Opin Inv Drug 6(12), 1887–1937(1997).

[13] J. J. Escobar-Chavez, V. Merino-Sanjuan, M. Lopez-Cervantes, Z. Urban-Morlan, E. Pinon-Segundo,D. Quintanar-Guerrero, and A. Ganem-Quintanar. JPharm Pharm Sci 11(1), 104–130 (2008).

[14] D. Bommannan, R. O. Potts, and R. H. Guy. J InvestDermatol 95(4), 403–408 (1990).

[15] N. Higo, A. Naik, D. B. Bommannan, R. O. Potts, andR. H. Guy. Pharm Res 10(10), 1500–1506 (1993).

[16] K. Tojo and A. R. C. Lee. J Invest Dermatol 92(1),105–108 (1989).

[17] J. J. Escobar-Chavez, D. Quintanar-Guerrero, andA. Ganem-Quintanar. Drug Dev Ind Pharm 31(4–5),447–454 (2005).

[18] H. Pinkus, J Invest Derm 16, 383–386 (1951).[19] V. P. Shah. US Dept of Health and Human Services,

Rockville, 1–19 (1998).[20] S. Holzmann, C. H. Tripp, M. Schmuth, K. Janke,

F. Koch, S. Saeland, P. Stoitzner, and N. Romanil. JInvest Derm 122(5), 1165–1174 (2004).

[21] Y. Rapp, F. S. Glickman, and L. Frank. Arch Derma-tol 88(3), 257–266 (1963).

[22] R. Huggenberger and M. Detmar. J Invest DermatolSymp P 15(1), 24–32 (2011).

[23] A. C. Shore. Brit J Clin Pharmaco 50(6), 501–513(2000).

[24] D. Giuggioli, A. Manfredi, M. Colaci, C. U. Manzini,A. Antonelli, and C. Ferri. Autoimmun Rev 12(11),1058–1063 (2013).

[25] W. Groner, J. W. Winkelman, A. G. Harris, C. Ince,G. J. Bouma, K. Messmer, and R. G. Nadeau. NatMed 5(10), 1209–1212 (1999).

[26] D. De Backer, K. Donadello, Y. Sakr, G. Ospina-Tas-con, D. Salgado, S. Scolletta, and J. L. Vincent. CritCare Med 41(3), 791–799 (2013).

[27] S. Clark, F. Campbell, T. Moore, M. Jayson IV, T. A.King, and A. L. Herrick. Microvasc Res 57(3), 284–291 (1999).

[28] H. Debbabi, P. Bonnin, P. H. Ducluzeau, G. Lefther-iotis, and B. I. Levy. Am J Hypertens 23(5), 541–546.

[29] C. L. Riordan, M. McDonough, J. M. Davidson,R. Corley, C. Perlov, R. Barton, J. Guy, and L. B.Nanney. J Burn Care Rehabil 24(4), 177–186 (2003).

[30] G. F. Clough. J Physiol 516(2), 549–557 (1999).[31] P. Ljung, S. Bornmyr, and H. Svensson. Acta Orthop

66(1), 59–63 (1995).[32] D. J. Newton, F. Khan, J. J. F. Belch, M. R. Mitchell,

and G. P. Leese. J Foot Ankle Surg 41(4), 233–237(2002).

[33] D. Huang, E. Swanson, C. Lin, J. Schuman, W. Stin-son, W. Chang, M. Hee, T. Flotte, K. Gregory, andC. A. Puliafito. Science 254(5035), 1178–1181 (1991).

J. Biophotonics (2014) 7

FULLFULLARTICLEARTICLE

# 2014 by WILEY-VCH Verlag GmbH & Co. KGaA, Weinheimwww.biophotonics-journal.org

[34] P. H. Tomlins and R. K. Wang. J Phys D Appl Phys38(15), 2519–2535 (2005).

[35] J. M. Schmitt. IEEE J Sel Top Quant 5(4), 1206–1215(1999).

[36] R. K. Wang and Z. Ma. Opt Lett 31, 3001–3003(2006).

[37] T. Gambichler, A. Orlikov, R. Vasa, G. Moussa,K. Hoffmann, M. Stucker, P. Altmeyer, and F. G. Be-chara. J Dermatol Sci 45(3), 167–173 (2007).

[38] M. Mogensen, B. M. Nurnberg, J. L. Forman, J. B.Thomsen, L. Thrane, and G. B. E. Jemec. Brit J Der-matol 160(5), 1026–1033.

[39] K. Yano, K. Kajiya, M. Ishiwata, Y. K. Hong, T. Miya-kawa, and M. Detmar. J Invest Dermatol. 122(1),201–208 (2004).

[40] J. Welzel, M. Bruhns, and H. H. Wolff. Arch Derma-tol Res 295(2), 50–55 (2003).

[41] T. Gambichler, S. Boms, M. Stucker, G. Moussa,A. Kreuter, M. Sand, D. Sand, P. Altmeyer, andK. Hoffmann. Arch Dermatol Res 297(5), 218–225(2005).

[42] L. Themstrup, M. Mogensen, and G. B. E. Jemec,22nd World Congress of Dermatology (2011).

[43] R. K. Wang, S. L. Jacques, Z. Ma, S. Hurst, S. R. Han-son, and A. Gruber. Opt Express 15(7), 4083–4097,(2007).

[44] L. An, J. Qin, and R. K. Wang, Opt Express 18,8220–8228 (2010).

[45] Y. Jia, P. Li, and R. K. Wang. J Biomed Opt 16(9),096019 (2011).

[46] L. An, T. T. Shen, and R. K. Wang. J Biomed Opt16(10), 106013 (2011).

[47] R. K. Wang and L. An. Opt Express 17(11), 8926–8940 (2009).

[48] U. Baran, L. Shi, and R. K. Wang. J Biophotonicsdoi: 10.1002/jbio.201300154 (2013).

[49] Y. Jung, S. Dziennis, Z. Zhi, R. Reif, Y. Zheng, andR. K. Wang. PloS one 8(2), e57976 (2013).

[50] J. Qin, J. Jiang, L. An, D. Gareau, and R. K. Wang.Laser Surg Med 43(2), 122–129 (2011).

[51] Y. Jia and R. K. Wang. J Neurosci Meth 194, 108–115(2010).

[52] R. K. Wang, L. An, P. Francis, and D. Wilson. OptLett 35, 1467–1469 (2010).

[53] S. Yousefi, Z. Zhi, and R. K. Wang. IEEE T BiomedEng 58(8), 2316–2323 (2011).

[54] L. Shi, J. Qin, R. Reif, and R. K. Wang. J BiomedOpt 18(10), 106015–106015 (2013).

[55] R. Reif, J. Qin, L. An, Z. Zhi, S. Dziennis, and R. K.Wang. J Biomed Imaging 2012, 509783 (2012).

[56] V. J. Srinivasan, S. Sakadzic, S., I. Gorczynska, S. Ru-vinskaya, W. Wu, J. G. Fujimoto, and D. A. Boas. OptExpress 18(3), 2477–2494 (2010).

[57] P. Mosher, J. Ross, P. A. Mcfate, and R. F. Shaw. CircRes 14(3), 250–259 (1964).

[58] M. Rossi and A. Carpi. Biomed Pharmacother 58(8),427–431 (2004).

H. Wang, U. Baran, and R. K. Wang: In vivo blood flow imaging of inflammatory human skin8

Journal of

BIOPHOTONICS

# 2014 by WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.biophotonics-journal.org