Optical biopsy of fixed human skin with backward-collected optical harmonics signals

Harmonics-based optical microscopy has been widely applied in biomedical researches due to its noninvasiveness to the studied biomaterials. Due to momentum conservation consideration, most previous studies collect harmonics generation signals in a forward geometry, especially for third harmonic generation signals. However, the adopted forward transmission type geometry is not feasible for future clinical diagnosis. In this paper, first virtual biopsy based on backward propagating optical higher harmonics, combining second harmonic and third harmonic, is demonstrated in the fixed human skin specimens. In our study, third harmonic generation can provide morphologic information including the distribution of basal cells while second harmonic generation can provide distribution of collagen fibers in dermis. Therefore, this technique is ideal for future noninvasive in vivo skin disease examination without dye. © 2005 Optical Society of America OCIS codes: (180.6900) Three-dimensional microscopy; (180.5810) Scanning microscopy; (190.4160) Multiharmonic generation References and links 1. http://www.skincancerinfo.com/sectionb/biopsy.html 2. M. Rajadhyaksha, M. Grossman, D. Esterowitz, R. H. Webb, and R. R. Anderson, “In vivo confocal scanning laser microscopy of human skin: melanin provides strong contrast,” J. Invest. Dermatol. 104, 946-952 (1995). #8523 $15.00 USD Received 22 August 2005; revised 24 September 2005; accepted 24 September 2005 (C) 2005 OSA 3 October 2005 / Vol. 13, No. 20 / OPTICS EXPRESS 8231 3. M. Rajadhyaksha, S. Gonzalez, J. M. Zavislan, R. R. Aderson, R. H. Webb RH, and R. R. Anderson, “In vivo confocal scanning laser microscopy of human skin II: advances in instrumentation and comparison with histology,” J. Invest. Dermatol. 113, 293-303 (1999). 4. B. R. Masters, and P. T. C. So, “Confocal microscopy and multi-photon excitation microscopy of human skin in vivo,” Opt. Express 8, 2-10 (2001). http://www.opticsexpress.org/abstract.cfm?URI=OPEX-8-1-2 5. P. J. Caspers, G. W. Lucassen, and G. J. Puppels, “Combined in vivo confocal Raman spectroscopy and confocal microscopy of human skin,” Biophy. J. 85, 572-580 (2003). 6. L. D. Swindle, S. G. Thomas, M. Freeman, and P. M. Delaney, “View of normal human skin in vivo as observed using fluorescent fiber-optic confocal microscopic imaging,” J Invest. Dermatol. 121, 706-712 (2003). 7. G. J. Tearney, M. E. Brezinski, B. E. Bouma, S. A. Boppart, C. Pitris, J. F. Southern, and J. G. Fujimoto, “In vivo endoscopic optical biopsy with optical coherent tomography,” Science 276, 2037-2039 (1997). 8. Y. Yasuno, S. Makita, Y. Sutoh, M. Itoh, and T. Yatagai, “Birefringence imaging of human skin by polarization-sensitive spectral interferometric optical coherent tomography,” Opt. Lett. 27, 1803-1805 (2002). 9. B. Povazay, K. Bizheva, A. Unterhuber, B. Hermann, H. Sattmann, A. F. Fercher, W. Drexler, A. Apolonski, W. J. Wadsworth, J. C. Knight, P. S. J. Russell, M. Vetterlein, and E. Scherzer, “Submicron axial resolution optical coherence tomography,” Opt. Lett. 27, 1800-1802 (2002). 10. W. Denk, J. H. Strickler, and W. W. Webb, “Two-photon laser scanning fluorescence microscopy,” Science 249, 73-76 (1990). 11. E. H. K. Stelzer, S. Hell, S. Lindek, R. Stricker, R. Pick, C. Storz, G. Ritter, and N. Salmon, “Nonlinear absorption extends confocal fluorescence microscopy into the ultra-violet regime and confines the illumination volume,” Opt. Commun. 104, 223-228 (1994). 12. B. R. Masters, P. T. C. So, and E. Gratton, “Multiphoton excitation fluorescence microscopy and spectroscopy of in vivo human skin” Biophys. J. 72, 2405-2412 (1997) 13. P. T. C. So, H. Kim, and I. E. Kochevar, “Two photon deep tissue ex vivo imaging of mouse dermal and subcutaneous structures,” Opt. Express 3, 339-350 (1998). http://www.opticsexpress.org/abstract.cfm?URI=OPEX-3-9-339 14. U. K. Tirlapur, K. König, C. Peuckert, R. Krieg, and J. J. Halbhuber, “Femtosecond near infrared laser pulses elicit generation of reactive oxygen species in mammalian cells leading to apoptosis-like death,” Exp. Cell Research 263, 88-97 (2001) 15. K. König K, P. T. C. So, W. W. Mantulin, and E. Gratton, “Cellular response to near-infrared femtosecond laser pulses in two photon microscope,” Opt. Lett. 22, 135-136 (1997). 16. I-H. Chen, S.-W. Chu, C.-K. Sun, P.-C. Cheng, and B.-L. Lin, “Wavelength dependent cell damages in biological multi-photon confocal microscopy: A micro-spectroscopic comparison between femtosecond Ti:sapphire and Cr:forsterite laser sources,” Opt. Quantum. Electron. 34, 1251-1266 (2002). 17. Y. Guo, P. P. Ho, H. Savage, D. Harris, P. Sacks, S. Schantz, F. Liu, N. Zhadin, and R. R. Alfano, “Second-harmonic tomography of tissues,” Opt. Lett. 22, 1323-1325 (1997). 18. G. Peleg, A. Lewis, M. Linial, and L. M. Loew ”Nonlinear optical measurement of membrane potential around single molecules at selected celluar sites,” Proc. Natl. Acad. Sci. 96, 6700-6704 (1999). 19. S.-W. Chu, I.-H. Chen, T.-M. Liu, P.-C. Cheng, C.-K.Sun, and B.-L. Lin, “Multimodal nonlinear spectral microscopy based on femtosecond Cr:foristerite laser,” Opt. Lett. 26, 1909-1911 (2001). 20. E. Brown, T. Mckee, E. Ditomaso, E. Pluen, B. Seed, Y. Boucher, and R. K. Jain, “Dynamic imaging of collagen and its modulation in tumors in vivo using second-harmonic generation,” Nat. Med. 9, 796-800 (2003) 21. S.-W. Chu, S.-Y. Chen, T.-H. Tsai, T.-M. Liu, C.-Y. Lin, H.-J. Tsai and C.-K. Sun, “In vivo developmental biology study using invasive multi-harmonic generation microscopy,” Opt. Express 11, 3093-3099 (2003). http://www.opticsexpress.org/abstract.cfm?URI=OPEX-11-23-3093 22. C.-K. Sun, S.-W. Chu, S.-Y. Chen, T.-H. Tsai, T.-M. Liu, C.-Y. Lin and H.-J. Tsai, “Harmonics generation microscopy for developmental biology,” J. Struct. Biol. 147, 19-30 (2004). 23. C.-K. Sun, C.-C. Chen, S.-W. Chu, T.-H. Tsai, Y.-C. Chen and B.-L. Lin, “Multi-harmonic-generation biopsy of skin,” Opt. Lett. 28, 2488-2490 (2003). 24. Y. Barad, H. Eizenberg, M. Horowitz, and Y. Silberberg, “Nonlinear scanning laser microscopy by third harmonic generation,” Appl. Phys. Lett. 70, 922-924 (1997). 25. P. J. Campagnola, M. D. Wei, A. Lewis, L. M. Loew, “High resolution non-linear optical microscopy of living cells by second harmonic generation,” Biophys. J. 77, 3341-3349 (1999). 26. A. T. Yeh, N. Nassif, A. Zoumi, and B. Tromberg, “Selective corneal imaging using combined second-harmonic generation and two-photon excited fluorescence,” Opt. Lett. 27, 2082-2084 (2002). 27. A. Zoumi, A. T. Yeh, and B. Tromberg, “Imaging cells and extracellular matrix in vivo by using second harmonic generation and two photon excited fluorescence,” Proc. Natl. Acad. Sci. 99, 11014-11019 (2002). 28. R. M. Williams, W. R. Zipfel, and W. W. Webb, “Interpreting second-harmonic generation images of collagen I fibrils,” Biophys. J. 88, 1377-1386 (2005). 29. M. Han, G. Giese, and J. F. Bille, “Second harmonic generation imaging of collagen fibrils in cornea and sclera,” Opt. Express 13, 5791-5797 (2005) http://www.opticsexpress.org/abstract.cfm?URI=OPEX-13-15-5791 #8523 $15.00 USD Received 22 August 2005; revised 24 September 2005; accepted 24 September 2005 (C) 2005 OSA 3 October 2005 / Vol. 13, No. 20 / OPTICS EXPRESS 8232 30. J. X. Cheng, A. Volkmer, L. D. Book, and X. S. Xie, “An epi-detected coherent anti-stokes Raman scattering (E-CARS) microscope with high spectral resolution and high sensitivity,” J. Phys. Chem. B 105, 1277-1280 (2001). 31. A. Volkmer, J. X. Cheng, and X. S. Xie, “Vibrational imaging with high sensitivity via epidetected coherent anti-Stokes Raman scattering microscopy,” Phys. Rev. Lett. 87, 023901 (2001). 32. J. X. Cheng, A. Volkmer, and X. S. Xie, “Theoretical and experimental characterization of coherent anti-Stokes Raman scattering microscopy,” J. Opt. Soc. Am. B 19, 1363-1375 (2002). 33. A. Volkmer, “Vibrational imaging and microspectroscopies based on coherent anti-Stoke Raman scattering microscopy,” J. Phys. D: Appl. Phys. 38, R59-R81 (2005). 34. J. X. Cheng, and X. S. Xie, “Green's function formulation for third-harmonic generation microscopy,” J. Opt. Soc. Am. B 19, 1604-1610 (2002). 35. C.-K Sun, S.-W Chu, S.-P. Tai, M. -C. Chan, I C. Hsiao, C. -H. Lin, Y. -C. Chen, and B. -L. Lin, “Origin of backward second harmonic generation emission in a biological sample examined by laser scanning microscopes”, in Technical Digest of Conference on Laser and Electro-optics/Quantum Electronics and Laser Science Conference (CLEO/QELS 2005), Baltimore, MD, USA, paper CWP7 (2005). 36. T.-H. Tsai, S.-P. Tai, W.-J. Lee, H.-Y. Huang, Y.-H. Liao and C.-K. Sun, “Signal degradation study in fixed human skin using confocal microscopy and higher-harmonic optical microscopy,” submitted to Biophys. J.. 37. S.-Y. Chen, S.-P. Tai, T.–H. Tsai and C.–K. Sun, ”Direct backward-emitted third-harmonic generation and its application to clinical microscopy”, in Technical Digest of Conference on Laser and Electro-optics/Quantum Electronics and Laser Science Conference (CLEO/QELS 2005), Baltimore, MD, USA, paper QMI3 (2005). 38. D. Yelin and Y. Silberberg, “Laser scanning third-harmonic-generation microscopy in biology,” Opt. Express 5, 169-175 (1999). http://www.opticsexpress.org/abstract.cfm?URI=OPEX-5-8-169 39. J. A. Squier, M. Muller, G. J. Brakenhoff, and K. R. Wilson, “Third harmonic generation microscopy,” Opt. Express 3, 315-324 (1998). http://www.opticsexpress.org/abstract.cfm?URI=OPEX-3-9-315 40. P. J. Campagnola, A. C. Millard, M. Terasaki, P. E. Hoppe, C. J. Malone, and W. A. Mohler, “Three-dimensional high-resolution second-harmonic generation imaging of endogenous structural proteins in biological tissues,” Biophys. J. 81, 493-508 (2002). 41. S.-W. Chu, I.-S. Chen, T.-M. Liu, C.-K. Sun, S.-P. Lee, B.-L. Lin, P.-C. Cheng, M.-X. Kuo, D.-J. Lin, and H.-L. Liu, “Nonlinear bio-photonic crystal effects revealed with multi-modal nonlinear microscopy,” J. Microsc. 208, Part 3, 190-200 (2002).


Introduction
In skin disease diagnosis, biopsy is a powerful tool.However, traditional physical biopsy requires the removal, fixation, and staining of tissues, cells, or fluids from the lesions of the patient [1].This histological procedure is not only time-consuming but also invasive and painful for patients.In addition, physical biopsy procedure per se may potentially put patients in the risk of the spreading of tumor cells.Moreover, unless tedious serial sectionings are performed, there is no guarantee for a missing diagnosis due to a local invasion not present in the given histologically-sectioned sample examined by the pathologist.Therefore, a non-invasive in vivo optical virtual biopsy, which can provide non-invasive, highly penetrative, three-dimensional (3D) imaging with sub-micron spatial resolution, is highly desired.
In previously studies, reflectance and fluorescence confocal laser scanning microscopies were applied to virtual biopsy of human skin with ultraviolet, visible, near infrared (NIR) and 1064nm light [2][3][4][5][6].Confocal laser scanning microscopy provides a significant improvement in both axial and lateral resolutions over conventional optical coherent tomography [7][8][9] and epifluorescence microscopy because it eliminates out-of-focus fluorescence with a spatial filter in the form of a confocal aperture.Confocal laser scanning systems, especially those with a short wavelength laser sources, are known for out of focus absorption, which would cause not only out-of-focus photo-bleaching but also cell damages [10,11].Utilizing two-photon induced fluorescence under NIR excitation, two-photon fluorescence microscopy (TPFM) was demonstrated in 1990 [10].With a NIR wavelength and a signal intensity that is in quadratic dependence on the laser intensity, TPFM showed high axial/depth discrimination even without a confocal pin-hole.Recently, optical biopsy based on laser scanning TPFM was successfully demonstrated by So's group [4,12,13].Two-photon fluorescence biopsy of skin based on 780 nm femtosecond light was shown to provide high resolution imaging from the skin surface through the epidermal-dermal junction.By comparing optical biopsies based on the confocal microscopy and TPFM of human skins [4], they found that TPFM can reduce photobleaching of the fluorophores, improve background discrimination, and minimize the photodamage to living cell specimens.Most importantly, they indicated that TPFM permits additional structures to be identified including the collagen fibers in the dermis.
However for future clinical applications without surgery, current 700-850 nm based TPFM technology still presents several limitations including limited penetration depth, in-focus cell damages, multi-photon phototoxicity due to high optical intensity in the 800 nm wavelength region, and toxicity due to required exogenous fluorescence markers.Several mechanisms causing nonlinear photon damages with 740-800 nm femtosecond excitations have recently been identified including oxidative photodamages caused by two-photon excitation of endogenous and exogenous fluorophores [14] or multi-photon-induced plasma generation [15,16].Previous researches also indicated that using shorter wavelength for more efficient two-photon excitation also produces greater photo-damages [14,15].These reports on the multi-photon-absorption-induced photo-damages by Ti:sapphire lasers in TPFM indicated that to reduce possible tissue damage while performing in vivo optical biopsy of human skin with an high intensity light source is very important.These photodamage phenomena also limit the maximum optical intensity applicable in a TPFM based biopsy system.Therefore, the signal intensity and the penetration depth are limited.Previous studies suggested that with an 80MHz Ti:sapphire laser and a NA~1.25 objective, 50% reduction in cloning efficiency could be observed in the Chinese hamster ovary cells when the average power was >2mW and the cells 100% stopped to clone for an average power >6mW [15].
In contrast to laser-induced fluorescence, higher harmonic generations (HHG), including second harmonic generation (SHG) and third harmonic generation (THG), are known to leave no energy deposition to its interacted matters due to the energy conservation characteristic and the emitted HHG photon energy is the same as the total annihilated excitation photon energy.With no energy release to its interacted specimens, this energy-conservation characteristic provides the optical "noninvasiveness" nature desirable for microscopy application, especially for in vivo clinical imaging [17][18][19][20].Due to its nonlinear nature, the generated SHG intensity depends on the square of the incident light intensity, while the generated THG intensity depends on the third power of the incident light intensity.Similar to multi-photon induced fluorescence processes, these nonlinear dependencies allow localized excitation and provide intrinsic sectioning capability.It is thus important to utilize these endogenous HHG signals for clinical imaging purpose to replace unnecessary usage of invasive and toxic fluorophores in common TPFM.In previous HHG microscopy studies of in vivo vertebrate embryos based on a femtosecond Cr:forsterite laser [21,22], complex developmental processes throughout the 1.5-mm-thick zebrafish embryos from initial cell proliferation, gastrulation, to the tissue formation could all be continuously visualized in vivo without any treatment on the live specimens.No optical damage could be found even after long-term (12-h) continuous observation with 100-mW average incident power onto one embryo, corresponding to a total energy exposure over 1000 J, confirming the noninvasive nature of the higher harmonic generation processes.Besides, excellent 3D resolution (sub-micron) and high penetration capability (1.5mm into the surface of the live specimen) were successfully demonstrated [22].It is thus highly desired to apply the HHG microscopy to in vivo virtual biopsy of human skins for non-invasive and highly-penetrative 3D imaging with sub-micron spatial resolutions.In a previous HHG study of the excised mouse skins, the results suggested that HHG microscopy could provide sub-micron resolution deep-tissue biopsy images in animal skins without using fluorescence and exogenous markers [23].
Unfortunately, the HHG microscope generally utilizes a transmission detection scheme, especially for THG microscopy.However for clinical optical biopsy applications, the backward-collection (or epi-collection) geometry is preferred due to optically thick human bodies.Although most of previous harmonics optical microscopes were based on a forward signal-collection geometry due to the momentum-conservation characteristic of higher harmonic generations [21,22,24,25], it does not mean that the HHG microscope based an epi-collection geometry is unfeasible, especially for the THG modality that is with a third order coherent process.While epi-SHG microscopy has already been widely adopted for imaging [20,23,[26][27][28][29], epi-detected coherent anti-Stoke Raman scattering (CARS) microscopy, which is another coherent third-order nonlinear microscopy just like THG microscopy, was also demonstrated to be able to suppress non-resonant background signals [30,31].Previous studies in CARS microscopy, including theoretically simulations and experimental studies, showed that the backward-propagating CARS signals can be generated in a very thin (sub-μm) layer compared with the pump laser wavelength under tightly focused excitation fields [30][31][32][33].They also indicated that this theoretical concept should be generally applicable to any nonlinear coherent microscopy including THG microscopy [34].
In this paper, the first harmonics-generation optical biopsy (HOB) system, based on backward-collected SHG and THG signals, is demonstrated in excised and fixed human skin.In this study, we confirm the fact that abundant THG and SHG signals can be collected with the backward collection geometry.Just like forward propagating SHG and THG, epi-THG is found to provide morphologic information in both epidermis and dermis layers including the size, shape, and distribution of basal cells with a sub-micron spatial resolution while epi-SHG can provide distribution information of collagen fibers in dermis.By comparing virtual biopsy images taken with backward and forward propagating HHG, we found that the epi-SHG can provide more information than the forward SHG and the collected signal intensity of the epi-THG could be higher than that of forward collected THG at a depth of 30μm below skin surface for a 1mm-thick specimen.Our results suggest that HOB based on backward collection geometry should be ideal for clinical in vivo skin diagnosis in the future.

Material and methods
The fixed integument specimen was removed from the back of a seventy-year-old female patient and preserved in formalin at 4℃.The experimental protocols were approved the National Taiwan University Hospital Institutional Review Board (NTUH-IRB).The thickness of the specimen under observation was about 1 mm. Figure 1 shows our HOB system setup.The study of the harmonics optical biopsy of fixed human skin was performed using a home-built femtosecond Cr:forsterite laser centered at 1230 nm with a 130 fs pulsewidth and a 110 MHz repetition rate.The spectral full width half maximum of the laser output was 15nm.The infrared laser beam was first shaped by a telescope and then directed into a modified beam scanning system (Olympus Fluoview300) and a microscope (Olympus BX-51).An IR water-immersion objective (Olympus LUMplanFL/IR 60X/NA 0.9/Working distance 2mm) was used to focus laser beam into the observed skin specimen.The scanning rate of Fluoview300 was 1000 lines/s corresponding to ~two frames per second for a 512×512 resolution.The observed specimen was mounted on a stage with an axial resolution of 25nm.The generated backward (or epi-) SHG and THG signals were collected using the same objective.A chromatic beam splitter, DM1 (Chroma technology 865dcxru), was used to direct the backward SHG and THG signals into a home-built photon detection system.In this photon detection system, the backward SHG and THG were seperated by another chromatic beam splitter, DM2 (Chroma technology 490DRXR) and detected by two separate PMTs (Harmamatsu R4220P for THG and Harmamatsu R943-02 for SHG) with 410nm (THG) and 615nm (SHG) narrow band interference filters (Chroma technology D410/10X and D615/10X) in front.With a system limitation for up to only 3 channels, we kept one channel for forward propagation SHG or THG imaging for comparison.The transmitted forward SHG and THG were collected by a NA 1.4 achromatic oil immersion condenser, and then directed into the PMT (Harmamatsu R928P) with 410nm (for THG) or 615 nm (for SHG) narrow band interference filters (Chroma technology D410/10X or D615/10X) in front.To filter out the fundamental laser wavelength, two color filters, CF1 and CF2 (CVI SKG5), were inserted to the systems as shown in Fig. 1.Theoretically, the lateral resolution of SHG microscopy is 500nm [35,36].For THG microscopy, the lateral resolution is 410nm [36,37].Practical spatial resolution in the human skin has been measured in a previous study and was found to be depth dependent [36].

HOB of fixed human skin based on backward propagating signal collection geometry
To demonstrate the feasibility for future clinical skin diagnosis, Fig. 2 shows an example of the vertically sectioned backward-(B-) SHG (Fig. 2(a)), B-THG ((Fig.2(b)), and the combined (Fig. 2(c)) images taken from the fixed human epidermis to dermis based on the backward signal-collection geometry.From the vertically sectioned HOB images, the general histological structures in both epidermis and dermis layers can be identified through the THG modality due to its sensitivity to local optical inhomogeneities [18,38,39].For example, the coverglass-formalin interface and the morphology of stratum corneum, stratum granulosum, stratum spinosum, and the basal layer can all be easily picked up through the B-THG modality (for example, see Fig. 2(b)).SHG, on the other hand, mainly reflects the distribution of connective tissues in the dermis layer [16,40,41] (Fig. 2(a)) due to strong SHG generation in collagen fibers.Through the B-SHG image (for example, see Fig. 2(a)), other structures like the dermal capillary surrounded by connective tissues can also be identified.In Fig. 2, we can obtain the structural information in dermis with a depth of 360μm below skin surface, which demonstrate the excellent penetration capability of the backward collection HOB.By enhancing the imaging contrast, we are able to observe the dermis structure with a depth more than 400μm (not shown).The penetration depth of HOB is comparable with previous studies by using reflection confocal microscopy based on a 1064nm excitation [3].Currently, the limitation of penetration depth is found to be dominated by the distortion of fundamental laser beam profile [36].While the laser beam propagates in the skin, the laser beam intensity and phase profile are both distorted by non-uniformly distributed microstructures in the skin, especially due to collagen fibrils in dermis, and the spatial resolution and the efficiency of HHG are both been reduced.Besides, the generation and collection efficiencies of the harmonics signals decrease with depth due to the greater attenuation losses of both the laser intensity and HHG signals.To further improve the image quality of the biopsy images deep inside the bio-tissues, some imaging processing (for example, deconvolution) on the displayed row data (as shown in this article) should be developed.Cooling PMT to reduce thermal noises and inserting pre-amplifiers in the detection system to improve HOB system sensitivity can also improve the penetration depth capability.From epidermis to dermis of the studied fixed human skin, we can specify coverglass-formalin interface (d), stratum corneum (e), stratum granulosum & spinosum (f) and the basal layer (g) from the B-THG image.In the B-SHG image, the dermal papilla (h) and dermal capillaries (i) are clearly resolved.Image size: 240μm×360μm Figure 3 shows examples of the horizontally sectioned HOB images taken from the fixed human skin sample at different depths.In this figure, micro-structures through epidermis including stratum corneum (Fig. 3(a)), stratum spinosum (Fig. 3(b)), and the basal layer (Fig. 3(c)) can be all clearly resolved through B-THG.In Fig. 3(c), not only basal cells but also sub-cellular information inside basal cells is picked up by the B-THG biopsy with a sub-micron resolution.Figure 3(d) shows horizontally sectioned HOB image taken in the dermal papilla area.While SHG reveal the rich collagen fibril structures in the dermal layer, other histological structures like dermal papilla, Meissner's corpuscle and capillary networks can also be clearly identified.From Fig. 3(d), we can also observe basal cells (through B-THG) surrounding the collagen fibrils.It can be found that the B-SHG signals are concentrated in the dermis layer and can provide information of the structure and distribution of collagen fibrils that may assist clinical diagnosis of connective tissue diseases such as scleroderma, tuberous sclerosis, dermal mucinosis, and solar keratosis, etc.On the other hand, B-THG can provide structure contour and histological information in both epidermis and dermis.Fig. 3. Horizontally sectioned HOB images of fixed human skin at depths of (a) 29μm, (b) 54μm, (c) 90μm, and (d) 156μm from the specimen surface.B-SHG and B-THG signals are denoted by green and blue colors respectively.Using B-HOB, we can identify the morphology of (a) stratum corneum, (b) stratum spinosum, and (c) the basal layer in epidermis where distinctive morphological and optical properties displayed during keratinocyte maturation process could be appreciated.Besides, the capillary networks (e) and the Meissner's corpuscle (f) located within the dermal papilla could be clearly observed by HOB.These observations indicate that B-HOB could be an ideal platform of noninvasive skin diagnostic tool without staining.Image size: 120μm×120μm.Figure 4 is a movie of a sequential set of horizontally sectioned HOB images.The movie is composed of 100 horizontally sectioned images and the depth difference between two adjacent images is 1.5-μm.Different profiles between two adjacent images can be found in Fig. 4, indicating excellent axial resolution of the demonstrated B-HOB with a 0.9NA objective.Unlike traditional physical biopsy, the demonstrated new modality of B-HOB enables direct acquisition of high resolution 3-D structural images of human skin and provides an ideal new platform for future clinical noninvasive optical virtual biopsy without using fluorescence and confocal pinhole.

Comparison of HOB based on backward and forward propagating HHG
The difference between virtual biopsy images based on backward and forward collection geometries is also an interesting and important issue.Therefore, we have also compared the images taken in the human skin by HOB based on forward and backward propagating signal collection geometries.

B-THG vs. F-THG
Figure 5(a-d) and Fig. 5(e-h) show the B-THG and forward-(F-) THG images of horizontal sections at depths of 30μm, 45μm, 72 μm, and 87 μm below the skin surface.By comparing the images of B-THG and F-THG (Fig. 5(i-l)), their patterns co-localize quite well and thus we can conclude that their biophysical origins are the same.We also analyzed the intensity ratio of the backward and forward propagating THG signals in the fixed human skin specimen.The B-THG/F-THG ratio can be found to be dependent on the thickness of the skin specimen.For in vivo clinical diagnosis, the collected signal intensity of the backward propagating harmonics generation should be much stronger than the forward one due to the optically thick human body.For the study shown in Fig. 5, the thickness of the fixed skin specimen is 1mm and we found that the B-THG/F-THG ratio decreased with the imaging depth.Surprisingly, the B-THG/F-THG ratio was 1.3~4 at a depth of <30μm beneath the skin surface, indicating that the collected signal intensity of the backward propagating THG is higher than that of the forward propagating THG.Similar to the epi-CARS, the strong B-THG signals could be attributed to direct backward THG emission [37] rather than the back-scattering of the forward THG.The ultrathin THG interaction layers, such as the sub-cellular organelles and cell membranes, could release the phase matching condition and could be the reason why such a strong B-THG signal could be seen.Besides, the collagen fibrils in the dermis are known to be highly scattering [2][3][4] and could decrease the detected THG intensity in the forward path.Although the B-THG/F-THG ratio decreased to a value of only 0.1-0.3 at a depth more than 60μm, we still could obtain clear B-THG images.

B-SHG vs. F-SHG
Unlike THG concentrates in the epidermis, SHG concentrates in the dermis at a depth more than 100μm and mainly reflects the distributions of collagen fibrils in the dermis.In our study, we compared the horizontally sectioned B-SHG (Fig. 6(a-d)) and F-SHG (Fig. 6(e-h)) images, including image patterns and signal intensities at a depth from 100μm to 150μm below skin surface in the dermis region.Comparing images simultaneously taken with B-SHG and F-SHG (Fig. 6 (i-l)), difference between these two modalities can be found and a B-SHG/F-SHG ratio ranging from 0.04 to 0.8 can be observed at different positions.Previous studies have revealed that B-SHG/F-SHG ratio in a collagen fibril is highly dependent on its diameter [28,35].The B-SHG/F-SHG ratio can achieve ~1 with the diameter of collagen fibrils less than λ/10 (~62nm in our case) [28].In a recent study, Han et al. used B-SHG and F-SHG microscopies to study collagen fibrils in cornea and sclera [29].They indicated that F-SHG is dominated in the cornea while B-SHG and F-SHG are comparable in the sclera.They mentioned that the collagen fibrils in the cornea form parallel-arranged collagen fibril bundles and thus the B-SHG suffers strong coherent destructive interference, while the collagen fibrils in sclera are randomly arranged.In our study, we observe that B-SHG (for example Fig. 6(a-d)) can pick up more details on the collagen fibril distribution than F-SHG (for example Fig. 6(e-h)).Based on previous findings [28,29,35], our observed difference between B-SHG and F-SHG images can also be attributed to the coherence effect.With a distribution of collagen fibrils with different diameters, the F-SHG image will be dominated by the thicker or parallel-arranged collagen fibrils due to the coherent effect while the image dynamic range is limited.On the other hand, due to the destructive interference effect, the B-SHG signal intensity from thicker or parallel-arranged collagen fibrils is not necessarily stronger than the B-SHG signal intensity from thinner or randomly arranged collagen fibrils.With a limited dynamic range, the B-THG can thus reveal more details in the collagen fibril distribution than the F-SHG.

Conclusion
First harmonics optical biopsy based on a backward signal-collection geometry is demonstrated in fixed hyuman skin.This novel optical biopsy technique does not require invasive staining and the harmonic generation processes leave no energy deposition to the examined tissue.Without complex physical biopsy procedures, it also can provide high spatial resolution images of human skin in all three dimensions with a penetration depth well into the dermis level.By improving detection system sensitivity and by applying deconvolution techniques in the future, virtual biopsy images with even higher quality and even better penetration capability could be achieved.Comparing images taken by HOB based on forward and backward signal-collection geometries, we can confirm that the biophysical origins of B-THG and F-THG images in the epidermis of human skin are the same due to the optically thin interaction thickness.On the other hand, with collagen fibrils of different diameters and arrangements, B-SHG and F-SHG images in the dermis of human skin look different due to coherent and destructive interference effects.With a limited dynamic range, B-SHG can provide more details in the distribution of collagen fibrils in the dermis than F-SHG.Based on intrinsic contrast mechanisms without fluorescence, our study confirms that backward-HOB can resolve detailed histological structures of human skin from epidermis into dermis.Therefore, HOB based on backward propagating signal collection geometry will be ideal for noninvasive in vivo skin disease examination without dye.

Fig. 2 .
Fig. 2. Vertically sectioned (a) B-SHG, (b) B-THG, and (c) combined HOB images of fixed human skin.B-SHG and B-THG signals are denoted by green and blue colors respectively.From epidermis to dermis of the studied fixed human skin, we can specify coverglass-formalin interface (d), stratum corneum (e), stratum granulosum & spinosum (f) and the basal layer (g) from the B-THG image.In the B-SHG image, the dermal papilla (h) and dermal capillaries (i) are clearly resolved.Image size: 240μm×360μm

Fig. 4 .
Fig. 4. (1.62 MB) Movie of stack of horizontal sections in the fixed human skin (13.5 MB version).Using B-HOB, a series of horizontal sections from epidermis to dermis are demonstrated.The movie is composed of 100 horizontal images.The optical depth difference between adjacent images is 1.5 μm.Image size: 120μm×120μm

Fig. 5 .
Fig. 5. Horizontally sectioned (a-d) B-THG and (e-f) F-THG images taken at different depths inside the epidermis of the fixed human skin specimen.(i-l) are the combined images for comparison.B-THG and F-THG signals are denoted by blue and red colors respectively.Image size: 80μm×80μm

Fig. 6 .
Fig. 6.Horizontally sectioned (a-d) B-SHG and (e-f) F-SHG images taken at different depths inside the dermis of the fixed human skin specimen.(i-l) are the combined images for comparison.B-SHG and F-SHG signals are denoted by green and red colors respectively.Image size: 80μm×80μm