The DIVER Microscope for Imaging in Scattering Media

Review The DIVER Microscope for Imaging in Scattering Media

Alexander Dvornikov 1, Leonel Malacrida 1,2 and Enrico Gratton 1,*

1 Laboratory for Fluorescence Dynamics, Department of Biomedical Engineering, University of California Irvine, Irvine, CA 92697, USA; [email protected] (A.D.); [email protected] (L.M.) 2 Departamento de Fisiopatología, Hospital de Clínicas, Facultad de Medicina, Universidad de la República-Uruguay, Montevideo 11400, Uruguay * Correspondence: [email protected]

Received: 11 May 2019; Accepted: 19 June 2019; Published: 21 June 2019

Abstract: We describe an advanced DIVER (Deep Imaging Via Emission Recovery) detection system for two-photon fluorescence microscopy that allows imaging in multiple scattering media, including biological tissues, up to a depth of a few mm with micron resolution. This detection system is more sensitive to low level light signals than conventional epi-detection used in two-photon fluorescence microscopes. The DIVER detector efficiently collects scattered emission photons from a wide area of turbid samples at almost any entrance angle in a 2π spherical angle. Using an epi-detection scheme only photons coming from a relatively small area of a sample and at narrow acceptance angle can be detected. The transmission geometry of the DIVER imaging system makes it exceptionally suitable for Second and Third Harmonic Generation (SHG, THG) signal detection. It also has in-depth fluorescence lifetime imaging (FLIM) capability. Using special optical filters with sin-cos spectral response, hyperspectral analysis of images acquired in-depth in scattering media can be performed. The system was successfully employed in imaging of various biological tissues. The DIVER detector can be plugged into a standard microscope stage and used as an external detector with upright commercial two-photon microscopes.

Keywords: non-linear microscopy; hyperspectral imaging; FLIM; scattering media; ﬂuorescence

1. Introduction The ability to visualize features in deep layers of biological tissue with high resolution is a very sought-after feature of imaging systems employed in medical diagnostics, as well as in clinical and biological applications. Deep-tissue imaging has many applications, all based on the necessity of exploring cells and molecules in the intact organism. As a general goal, it may serve as a kind of “optical anatomo-pathology” tool that removes the requirement of biopsies and tissue fixation and allows access to structures and functions in the native physiological environment. Both geometrical and optical properties of a sample and characteristics of the microscope system affect the achievable imaging depth. While transparent specimens can be easily imaged with a traditional microscope, biological tissue is an intrinsically turbid medium, which produces strong multiple scattering, and absorption and which exhibits inhomogeneity of the refractive index. These features make traditional light and fluorescence microscopy inefficient at depths more than 100–200 µm from the surface of a sample [1], even with the aid of staining and the addition of fluorescent markers to improve the contrast. In order to access deeper layers of turbid samples, high power short pulsed lasers emitting in near infrared (NIR) are required. This approach allows deeper light penetration into media, because of the reduced absorption and scattering at longer wavelengths. The advent of multi-photon microscopy

Methods Protoc. 2019, 2, 53; doi:10.3390/mps2020053 www.mdpi.com/journal/mps Methods Protoc. 2019, 2, 53 2 of 12 has improved the achievable imaging depth, but the current microscope technology is still limited to an imaging depth of about 1 mm [2–5]. We built a two-photon fluorescence microscope, the DIVER (Deep Imaging Via Emission Recovery) that is capable of imaging in turbid media simulating brain optical properties up to the depth of a few mm. Such an increased penetration depth is the result of a novel photon harvesting system which collects emission photons from a wide area of the sample in contrast to the conventional detection scheme that collects photons only from a relatively small area, defined by the objective Numerical Aperture (NA), and therefore the loss of most emission photons. In our microscope light losses are also reduced by means of refractive index matching throughout the optical path. The system has proven effective on imaging of biological and artificial samples, such as murine colon and small intestine, vasculature in the skin, subcutaneous xenograft tumors in mice and various tissue phantoms [6–9]. Due to its transmission geometry, the DIVER imaging system is also particularly suitable for Second and Third Harmonic Generation (SHG, THG) signal detection. SHG is a coherent process that produces photons at exactly half the wavelength of the excitation light. Second harmonic generation is a useful optical tool for biological and medical imaging and diagnostics [10]. On the other hand, THG is generated at the interphase of a sharp change in the index of refraction [11]. The physical principle for both processes does not involve absorption to generate contrast, like in the case of multi-photon fluorescence microscopy, but arises from the polarization of specific endogenous structures lacking center of symmetry (SHG) or mismatch in the refraction index (THG), thus not requiring staining of the sample. SHG and TGH also work with NIR lasers frequencies, which allows imaging at an increased depth in biological tissue with high resolution. Unlike fluorescence, which produces an isotropic emission, SHG and THG mainly propagate in the direction of the excitation beam [12,13]. Due to its intrinsic forward-directed nature, SHG and THG detection is better accomplished in an imaging system working in transmission geometry. SHG is a very powerful imaging technique, both ex vivo and in vivo, and, for instance, has been successfully used in imaging of diseases of connective tissue [14]. Some types of collagen, the most abundant component of the extra-cellular matrix (ECM), generate a very strong second harmonic signal. Many diseases, such as ovarian [15] and breast cancer [16] are characterized by a remodeling of the ECM. Additionally, myosin, present in large amounts in biological samples, gives a strong SHG signal and the imaging of the arrangement of the fibrils can give insight into some muscle diseases [17]. We were able to obtain SHG images of biological and non-biological specimens in depth using very low excitation light power, thus causing less photo-damage, when compared to traditional microscope geometry [18,19]. The DIVER system is also equipped with a FLIMBox card that allows in depth fluorescence lifetime imaging (FLIM). FLIM is a powerful imaging technique that maps the spatial distribution of the unique lifetimes of intrinsic and extrinsic fluorophores, such as NADH, FAD, and fluorescent proteins [20,21]. FLIM has many applications to explore various biological processes, such as protein-protein interactions and others [22,23]. Traditionally, FLIM analysis was performed in the time domain and has been computationally demanding [24]. The phasor-approach, introduced by Jameson et al. [25] and later on extended to FLIM data analysis in the frequency domain by Digman et al. [26], has greatly simplified and enhanced data processing. The phasor-approach was also used for data analysis in hyperspectral imaging. The spectral phasor approach introduced by Fereidouni et al. [27] shows major advantages over classical multicomponent deconvolution to resolve hyperspectral images. In the spectral phasor approach the spectra are transformed to the real and imaginary components of the Fourier transformation, then the spectral shift and broadening of the full width at the half maximum (FWHM) are related to the phase increase or modulation decrease, respectively [28,29]. Unlike deconvolution, the spectral phasor approach does not require the knowledge of a spectral basis set for the analysis, which are instead needed for a classical fitting approach [29,30]. In particular, the application of spectral phasors to quantify dipolar Methods Protoc. 2019, 2, 53 3 of 12 Methods Protoc. 2019, 2, x FOR PEER REVIEW 3 of 12 fluorescencerelaxations at is the important interphase in simplifying of membranes the usingstudy LAURDAN of membrane fluorescence dynamics and is important structures in in simplifying cells and tissuesthe study [30,31]. of membrane dynamics and structures in cells and tissues [30,31]. RecentlyRecently we havehave shown shown that that using using the the set ofset two of filterstwo filters with thewith sine the and sine cosine and spectral cosine responsespectral responseit is possible it is topossible resolve to spectral resolve components spectral compon in hyperspectralents in hyperspectral images with images single with channel single detectors, channel detectors,such as the such DIVER, as the and DIVER, in strongly and in scattering strongly sca systemsttering [32 systems]. This technique[32]. This technique is particularly is particularly useful for usefulhyperspectral for hyperspectral imaging in imaging biological in tissues, biological where tissues, emitted where photons emitted are comingphotons from are allcoming directions from andall directionshyperspectral and detectors,hyperspectral which detectors, require collimatedwhich requ light,ire collimated are not very light, effi arecient. not very efficient.

2.2. Materials Materials and and Methods Methods TissueTissue phantoms phantoms imitating imitating brain brain optical optical proper propertiesties and and used used in in experiments experiments were were prepared prepared accordingaccording to to [9,33]. [9,33]. TissueTissue samples, samples, such such as asmouse mouse brain, brain, rat lungs, rat lungs, mouse mouse bone, liver, bone, kidney, liver, kidney,sciatic nerve, sciatic adipose nerve, tissueadipose or tissue bacteria or bacteria were generously were generously provided provided by various by various research research laboratories laboratories and and researchers researchers workingworking at at Laboratory Laboratory for for Fluorescence Fluorescence Dynamics (LFD) on other projects. CyanCyan and Green plasticplastic filtersfilters (Amazon, (Amazon, color color filters, filters, cat# cat# B016Q0BA6A) B016Q0BA6A) were were used used as cosineas cosine and andsine sine filters filters in hyperspectral in hyperspectral imaging imaging experiments. experiments. DataData acquisition acquisition and and analysis analysis was was performed performed using using the the Globals Globals for for Images-SimFCS Images-SimFCS software software by by G-SoftG-Soft Inc. Inc. (Champaign, (Champaign, IL-USA). IL-USA). TheThe concepts concepts of DIVERDIVER detectordetector operation operation can can be be found found in in [7 ,[7,8].8]. Originally Originally it wasit was designed designed for for the thepurpose purpose of deep of deep imaging imaging in turbid in turbid media, media, particularly particularly in biological in biological tissues, but tissues, later it but was later successfully it was successfullyused in various used projects in various related projects to diseases related to in diseases organs, suchin organs, as kidney, such as liver kidney, and others,liver and owing others, its owingexceptional its exceptional ability to utilizeability SHG,to utilize THG, SHG, FLIM THG, and recentlyFLIM and for recently Hyperspectral for Hyperspectral imagining [imagining9,19,32,34]. [9,19,32,34].Here we present Here thewe recentpresent advances the recent in advances the DIVER in imaging the DIVER system. imaging system. TheThe custom custom made made DIVER DIVER microscope microscope and and its its key key components components are are schematically schematically shown shown in in Figure Figure 1.1. ItIt utilizes utilizes the the Spectra Spectra Physics Physics InSight InSight DS+ DS+ femtosecondfemtosecond laser laser for for two-photon two-photon excitation, excitation, tunable tunable in in thethe spectral spectral range range of of 680–1300 680–1300 nm. nm. An An Acousto-optical Acousto-optical Modulator Modulator (AOM) (AOM) by by AA AA Opto-Electronics Opto-Electronics (Orsay-France),(Orsay-France), is is used used to to adjust adjust the the excitation excitation beam beam power. power. XY-Scanner, XY-Scanner, Cambridge Cambridge Technology Technology (Bedford,(Bedford, MA, MA, USA) USA) is is coupled coupled to to scanning scanning opti optics,cs, constructed constructed using using four four achromatic achromatic doublets doublets (Edmund(Edmund Optics, Optics, Inc, Inc, Barrington, Barrington, NJ, NJ, USA) USA) in insymmetric symmetric arrangement arrangement (Plössl-type (Plössl-type scan scan lens), lens), to minimizeto minimize aberrations aberrations and and allow allow wide wide field field of ofview view imaging imaging [35]. [35 ].This This combination combination of of 15 15 mm mm clear clear apertureaperture scanning mirrorsmirrors andand 50 50 mm mm diameter diameter lenses lenses allows allows us us to achieveto achieve 2–3 2–3 fold fold wider wider field field of view of view(FOV) (FOV) compared compared with commercialwith commercial microscopes, microscopes, using using the same the same objective objective lenses lenses for imaging. for imaging.

FigureFigure 1. 1. SchematicsSchematics of of the the DIVER DIVER (Deep (Deep Imaging Imaging Via Via Emission Emission Recovery) Recovery) imaging imaging system. system.

TheThe DIVER DIVER detector, detector, with with construction details details shown shown in Figure 22a,A, is placedplaced inin thisthis customcustom imagingimaging system system on on the the three-axis three-axis FTP-2050-XYZLE FTP-2050-XYZLE microscope microscope stage stage (ASI, (ASI, OR- OR- USA) USA) equipped equipped with with linearlinear encoders encoders for for high high precision precision sample sample positioning. positioning. The The imaging imaging system system is is also also equipped equipped with with an an H7422P-40H7422P-40 epiepi-detector-detector (Hamamatsu (Hamamatsu Photonics Photonics K.K., K.K., Japan). Japan). Both Both the the DIVER DIVER and and epiepi-detector-detector are are connectedconnected to to a a 320 320 Fast-FLIMbox Fast-FLIMbox (ISS, (ISS, Inc., Inc., Champa Champaign-Illinois,ign-Illinois, USA) USA) to to allow allow FLIM FLIM measurements measurements and phasor analysis. The DIVER detector can be directly plugged in microscope stages and used as

Methods Protoc. 2019, 2, 53 4 of 12

Methods Protoc. 2019, 2, x FOR PEER REVIEW 4 of 12 and phasor analysis. The DIVER detector can be directly plugged in microscope stages and used as an anexternal external detector detector in commercial in commercial upright upright microscopes, microscopes, such assuch for exampleas for example Olympus Olympus BX 61, as BX shown 61, as in shownFigure 2inB,C. Figure 2B,C.

FigureFigure 2. ((AA)) The The DIVER DIVER detector detector construction. construction. (B) Picture (B) Picture of the commercialof the commercial Olympus BX61Olympus microscope BX61 microscopewith the DIVER with detector.the DIVER (C) detector. Picture of ( theC) Picture DIVER detectorof the DIVER that can detector be plugged that incan the be microscope plugged in stage. the microscope stage. In the DIVER detector, a photomultiplier tube (PMT) with a wide 18 18 mm photocathode area × (R7600P-300,In the DIVER Hamamatsu) detector, is a used. photomultiplier The PMT is attached tube (PMT) to the with sealed a wide chamber 18 × containing 18 mm photocathode filters/shutter areawheel (R7600P-300, (see Figure2 A).Hamamatsu) The chamber is used. is filled The with PMT propylene is attached glycol to to the eliminate sealed airchamber gaps in containing the optical filters/shutterpath from the wheel sample (see to Figure detector. 2A). Two The 25 chamber mm round is filled BG39 with filters propylene (SCHOTT glycol North to eliminate America Inc.,air gapsDuryea, in the PA-USA) optical path are used from as the windows sample to to detec seal thetor. chamberTwo 25 mm and round also to BG39 block filters the excitation (SCHOTT IR North laser Americalight. The Inc, filter Duryea, wheel canPA-USA) accommodate are used up as to windows eight optical to seal filters the of chamber 25 mm diameter and also (one to positionblock the is excitationused as a lightIR laser block) light. that The can filter be easily wheel exchanged can accommodate by users. The up filter to eight wheel optical is rotated filters by theof 25 stepper mm diametermotor to (one the desired position position; is used the as operatinga light block) software that allowscan be automaticeasily exchanged sequential by imageusers. acquisitionThe filter wheelusing diis ffrotatederent filters by andthe stepper laser powers. motor to the desired position; the operating software allows automaticSamples sequential are placed image directly acquisition on the using top window different of thefilters DIVER and laser detector. powers. A drop of water is used to fill theSamples air gap are between placed thedirectly sample on andthe top detector window window of the to DIVER reduce detector. reflections; A drop for the of water same reasonis used a todrop fill the of index air gap matching between microscope the sample oil and is useddetector to fill window the gap to between reduce the reflections; PMT and for the the bottom same windowreason aof drop the chamber.of index matching Matching microscope index of refraction oil is us ined theto fill optical the gap path between from a sample the PMT surface and the to thebottom PMT windowwindow of maximizes the chamber. photon Matching transmission, index whichof refracti otherwiseon in the significantly optical path decreases from a due sample to reflections surface to on theinterfaces. PMT window This approach maximizes is especially photon transmission, important for which efficient otherwise collection significantly of scattered decreases emission photonsdue to reflectionsthat are propagating on interfaces. at various This approach directions is andespecial angles.ly important If there is anfor air efficient gap between collection optical of scattered elements, emissionphotons arrivingphotons atthat critical are anglepropagating will be lostat various due to total directions internal and reflection. angles. If there is an air gap betweenThis optical simple detectorelements, construction photons arriving has proven at tocritical be very angle efficient will for be collection lost due of scatteredto total emissioninternal reflection.photons from a wide area of sample surface and practically at any entrance angle in a 2pi solid angle. ThisThis feature simple significantly detector enhancesconstruction detector has sensitivityproven to whenbe very compared efficient with for ancollection epi-detection of scattered scheme, emissionwhere most photons of emitted from surfacea wide photonsarea of sample are lost, surface especially and whenpractically imaging at any in scattering entrance angle media. in a 2pi solid angle. This feature significantly enhances detector sensitivity when compared with an epi-detection3. Results scheme, where most of emitted surface photons are lost, especially when imaging in scattering media. 3.1. Imaging in Scattering Media: the DIVER Versus Epi-Detector 3. ResultsTocompare imaging performance of the DIVER detector and the epi-detector we imaged fluorescent beads embedded in scattering silicone samples, containing Titanium oxide particles for scattering. 3.1.The Imaging scattering in Scattering properties Media: of these the samplesDIVER Versus were adjustedEpi-Detector to match scattering properties of biological tissues,To compare for instance imaging brain tissueperformance (Figure3 ofcaption). the DIVER detector and the epi-detector we imaged fluorescentThe images beads atembedded different depthsin scattering were acquiredsilicone samples, at the same containing time by Titanium both detectors oxide andparticles then for the scattering.ratio of signal The intensitiesscattering properties was plotted of as these a function samples of were imaging adjusted depth, to as match shown scattering in Figure properties3. At the top of biologicalsurface of tissues, the sample for instance the signal brain intensities tissue (Figure were set 3 caption). to be about the same for both detectors.

Methods Protoc. 2019, 2, x FOR PEER REVIEW 5 of 12 Methods Protoc. 2019, 2, x FOR PEER REVIEW 5 of 12 The images at different depths were acquired at the same time by both detectors and then the The images at different depths were acquired at the same time by both detectors and then the ratio of signal intensities was plotted as a function of imaging depth, as shown in Figure 3. At the top ratio of signal intensities was plotted as a function of imaging depth, as shown in Figure 3. At the top surfaceMethods Protoc.of the2019 sample, 2, 53 the signal intensities were set to be about the same for both detectors. 5 of 12 surface of the sample the signal intensities were set to be about the same for both detectors.

Figure 3. Comparison of the DIVER and the epi-detector efficiency. Images of fluorescent beads in FigureFigure 3.3. ComparisonComparison ofof thethe DIVERDIVER andand thethe epi-detectorepi-detector eefficiency.fficiency. ImagesImages ofof fluorescentfluorescent beadsbeads inin scatteringscattering sample sample (left (left) )and and dependence dependence of of epiepi- to to DIVER DIVER de detectortector signals signals ratio ratio on on imaging imaging depth depth scattering sample (left) and dependence of epi- to DIVER detector signals ratio on imaging depthμ ((rightright).). Samples Samples measured: measured: transparent transparent (no (no TiO TiO2 particlesparticles added); added); scattering scattering (scattering (scattering coefficient coefficient s (right). Samples measured: transparent (no TiO22 particles added); scattering (scattering coefficient μs = 3.5 mm−1); stronger1 scattering (μs = 10 mm−1). 1 µs = 3.5 mm−1 − ); stronger scattering (µs = 10− mm1 − ). = 3.5 mm ); stronger scattering (μs = 10 mm ). WithWith increase increase of of imaging imaging depth depth in in scattering scattering samples samples the the performance performance of of DIVER DIVER detector detector was was With increase of imaging depth in scattering samples the performance of DIVER detector was betterbetter than than the the epi-detector; epi-detector; moreover, moreover, at at certain certain depths, depths, images images can can be be acquired acquired only only with with the the better than the epi-detector; moreover, at certain depths, images can be acquired only with the DIVERDIVER detector. ThisThis imaging imaging depth depth dependence dependence is more is more pronounced pronounced in stronger in stronger scattering scattering samples. DIVER detector. This imaging depth dependence is more pronounced in stronger scattering samples.For transparent For transparent samples the samples DIVER the detector DIVER also detector performs also slightly performs better slightly with imaging better depthwith imaging increase, samples. For transparent samples the DIVER detector also performs slightly better with imaging depthdue to increase, the fact thatdue the to microscopethe fact that objective the microsco focal pointpe objective moves towardfocal point the DIVERmoves detector.toward the This DIVER aspect depth increase, due to the fact that the microscope objective focal point moves toward the DIVER detector.in turn increases This aspect the in eff turnective increases Numerical the Apertureeffective Numerical of the DIVER Aperture detector, of whilethe DIVER for the detector, epi-detector while it detector. This aspect in turn increases the effective Numerical Aperture of the DIVER detector, while forremains the epi-detector the same. it remains the same. for the epi-detector it remains the same. FigureFigure 44 shows shows examples examples of imagesof images acquired acquired with with the DIVER the DIVER detector detector in various in tissuevarious samples. tissue Figure 4 shows examples of images acquired with the DIVER detector in various tissue samples.The achievable The achievable imaging depth imaging depends depth on depends tissue optical on tissue properties, optical however, properties, we foundhowever, that usuallywe found the samples. The achievable imaging depth depends on tissue optical properties, however, we found thatsame usually samples the could same besamples imaged could 2–3 timesbe imaged deeper 2–3 using times the deeper DIVER using system the thanDIVER using system theepi-detector than using that usually the same samples could be imaged 2–3 times deeper using the DIVER system than using theor commercialepi-detector microscopes or commercial (like microscopes Zeiss LSM-710 (lik ore Zeiss Olympus LSM-710 LSM-BX61). or Olympus For example, LSM-BX61). using For the the epi-detector or commercial microscopes (like Zeiss LSM-710 or Olympus LSM-BX61). For example,DIVER detector using the we DIVER could imagedetector the we fixed could mouse image brain the fixed tissue mouse from abrain transgenic tissue from mouse a transgenic expressing example, using the DIVER detector we could image the fixed mouse brain tissue from a transgenic mouseThy-1-YFP-16, expressing which Thy-1-YFP-16, is a motor, which sensory is anda motor, central sensory neuron and marker, central up neuron to 0.9 marker, mm depth, up to Figure 0.9 mm4A, mouse expressing Thy-1-YFP-16, which is a motor, sensory and central neuron marker, up to 0.9 mm depth,while usingFigure the 4A, epi-detector while using the the imaging epi-detector depth the was imaging limited depth to about was 0.3 limited mm. to about 0.3 mm. depth, Figure 4A, while using the epi-detector the imaging depth was limited to about 0.3 mm.

Figure 4. A Figure 4. (A( ) )3D 3D reconstitution reconstitution of of a a brain brain section section of of a a transgenic transgenic mouse mouse expressing expressing Thy-1-YFP-16 Thy-1-YFP-16 as as a a markerFigure of4. neurons.(A) 3D reconstitution (B) 3D intensity of a and brain fluorescence section of lifetime a transgenic imaging mouse (FLIM) expressing images of Thy-1-YFP-16 autofluorescence as a marker of neurons. (B) 3D intensity and fluorescence lifetime imaging (FLIM) images of frommarker liver of tissue neurons. of a mouse (B) 3D under intensity high fat and diet. fluorescence The phasor plotlifetime allows imaging isolation (FLIM) of diff erentimages lipids of autofluorescence from liver tissue of a mouse under high fat diet. The phasor plot allows isolation of (greenautofluorescence and red cursors). from liver (C) tissue 3D intensity of a mouse and FLIMunder imageshigh fat of diet. a GFP The expressing phasor plot mouse allows embryo isolation eye. of different lipids (green and red cursors). (C) 3D intensity and FLIM images of a GFP expressing Thedifferent cursor lipids selections (green used and at red the cursors). phasor plot (C) enable3D intensity isolation and of diFLIMfferent images features, of asuch GFP asexpressing collagen

layers, GFP cells etc., due to lifetime heterogeneities, which cannot be seen in an intensity image. Methods Protoc. 2019, 2, x FOR PEER REVIEW 6 of 12

mouse embryo eye. The cursor selections used at the phasor plot enable isolation of different features, such as collagen layers, GFP cells etc., due to lifetime heterogeneities, which cannot be seen in an intensity image.

As was mentioned above, the DIVER detector is coupled with the Fast-FLIMbox which allows acquisition of 3D FLIM images, where various sample features are colored according to their fluorescence lifetimes. Examples of such 3D FLIM images are shown in Figure 4B,C, where features are highlighted by cursors on phasor plots, while intensity images do not reveal these differences. Methods Protoc. 2019, 2, 53 6 of 12 This feature is a quite unique option of the DIVER imaging system. Another example of the potential use of the DIVER microscope and its FLIM capability is applicationAs was to mentionedcomplex tissue above, diseases, the DIVER such detector as, for is instance, coupled withlung the pathology. Fast-FLIMbox Figure which 5A–F allows shows imagesacquisition of a large of 3D section FLIM images, (10 × 6 where × 1 mm) various of rat sample lung featuresembedded are colored in paraffin, according and tothis their sample fluorescence was also usedlifetimes. for traditional Examples microtome of such 3D cutting. FLIM images The images are shown were in Figureacquired4B,C, at wherelarge ~2 features mm FOV. are highlighted Figure 5D,E showsby cursors an enlarged on phasor region plots, of whileinterest intensity (ROI 1) images from Figure do not 5C reveal and these an intensity differences. profile This plot feature along is athe dashedquite uniqueline. These option images of the rely DIVER only imaging on two-phot system.on induced autofluorescence of NADH and do not requireAnother any staining example of ofa sample. the potential The useintensity of the profile DIVER plot microscope thus can and be its potentially FLIM capability used as is a structuralapplication parameter to complex for tissueanatomical diseases, characterization such as, for instance, of the septum lung pathology. thickness Figureor alveolar5A–F showscollapse, images of a large section (10 6 1 mm) of rat lung embedded in paraffin, and this sample was also which are common markers ×in lung× anatomy-pathology procedure, traditionally performed with used for traditional microtome cutting. The images were acquired at large ~2 mm FOV. Figure5D,E hematoxylin-eosin staining. In addition, the lifetime heterogeneity can be used to generate contrast shows an enlarged region of interest (ROI 1) from Figure5C and an intensity profile plot along the and identify modifications in the autofluorescence fingerprint of the lung (Figure 5F–H). The dashed line. These images rely only on two-photon induced autofluorescence of NADH and do paraffin lung block is perhaps one of the most challenging samples to obtain 3D images, because not require any staining of a sample. The intensity profile plot thus can be potentially used as a alveoli are bags of air that make this tissue sample very inhomogeneous and strongly distort images. structural parameter for anatomical characterization of the septum thickness or alveolar collapse, A which3D z-stack are common was acquired, markers and in orthogonal lung anatomy-pathology views are shown procedure, in the Figure traditionally 5I. As performeda result of withimage distortion,hematoxylin-eosin the image staining.resolution In decreases addition, the with lifetime imaging heterogeneity depth, however, can be usedthe DIVER to generate detector contrast is still ableand to identifyimage in modifications reasonably indeep the layers. autofluorescence fingerprint of the lung (Figure5F–H). The para ffin lungFigure block 5J is perhapsillustrates one the of the potential most challenging use of au samplestofluorescence to obtain 3Dimages images, and because FLIM-Phasors alveoli are to fingerprintbags of air lung that pathology make this by tissue differences sample veryin emissi inhomogeneouson lifetimes andfor healthy strongly and distort diseased images. tissues. A 3D In thisz-stack example was we acquired, show andthe difference orthogonal between views are a shown control in and the Figurea broncho-alveolar5I. As a result oflavage image injured distortion, lung. Usingthe imagethe DIVER resolution microscope decreases ability with imagingto image depth, at large however, FOV we the were DIVER able detector easily isto still acquire able to images image of fullin lung reasonably samples deep of 10 layers. × 6 mm size.

FigureFigure 5. 5.FLIM-PhasorFLIM-Phasor Lung Lung analysis ofof NADH NADH autofluorescence. autofluorescence. (A )( IllustrationA) Illustration of a ratof a lung rat andlung the and theregion region from from where where the the section section for para for ffiparaffinn inclusion inclusion was obtained. was obtained. (B) Picture (B) ofPicture the rat of lung the sectionrat lung sectionincluded included in para inffi paraffinn and placed and inplaced the dish in the for imaging.dish for imaging. (C) Large ( fieldC) Large of view field (FOV) of view FLIM (FOV) image FLIM of imagelung of using lung Zeiss using ECPlan-NEOFLUAR Zeiss ECPlan-NEOFLUAR 10x/0.3 objective.10x/0.3 objective. The region The of region interest of (ROI interest 1) is (ROI the region 1) is the regionshown shown in section in section D. (D )D. The (D dashed) The dashed line is showingline is showing the profile the of prof intensityile of tointensity quantify tothe quantify septum the thickness. (E) Septum thickness plot. (F) Pseudocolor image generated by the cursor selection at the

phasor plot in H. (G) Zoom in of the ROI 2 in (F). (H) FLIM-Phasor plot. The cursors select different lifetimes. (I) Orthogonal views of the lung 3D image. The arrows indicate an alveolus. (J) Pseudocolor images of control and injured rat lungs by broncho-alveolar lavage. The colors indicate differences in the lifetime. Methods Protoc. 2019, 2, x FOR PEER REVIEW 7 of 12 Methods Protoc. 2019, 2, 53 7 of 12 septum thickness. (E) Septum thickness plot. (F) Pseudocolor image generated by the cursor selection at the phasor plot in H. (G) Zoom in of the ROI 2 in (F). (H) FLIM-Phasor plot. The cursors select Figure5J illustrates the potential use of autofluorescence images and FLIM-Phasors to fingerprint different lifetimes. (I) Orthogonal views of the lung 3D image. The arrows indicate an alveolus. (J) lung pathology by differences in emission lifetimes for healthy and diseased tissues. In this example Pseudocolor images of control and injured rat lungs by broncho-alveolar lavage. The colors indicate we show the difference between a control and a broncho-alveolar lavage injured lung. Using the differences in the lifetime. DIVER microscope ability to image at large FOV we were able easily to acquire images of full lung samples of 10 6 mm size. 3.2. SHG and THG× Imaging 3.2. SHGHarmonic and THG generation Imaging imaging is a label-free imaging technique, wherein certain molecular structures can generate light of doubled (SHG) or tripled (THG) frequency of excitation light. In Harmonic generation imaging is a label-free imaging technique, wherein certain molecular biological tissue SHG can be used for imaging features such as collagen fibers, muscles, blood structures can generate light of doubled (SHG) or tripled (THG) frequency of excitation light. vessels and some other tissues [18,36–38]. THG, which is induced by features with sharp changes in In biological tissue SHG can be used for imaging features such as collagen fibers, muscles, blood vessels refraction index, can be used, for instance, for imaging lipid droplets and cell membranes [39,40]. and some other tissues [18,36–38]. THG, which is induced by features with sharp changes in refraction Both SHG and THG photons are, in most cases, intrinsically forward directed, therefore index, can be used, for instance, for imaging lipid droplets and cell membranes [39,40]. Both SHG and epi-detection is not efficient for this kind of imaging and requires much higher excitation laser THG photons are, in most cases, intrinsically forward directed, therefore epi-detection is not efficient power to induce sufficient photons for detection. Additionally , in conventional multiphoton for this kind of imaging and requires much higher excitation laser power to induce sufficient photons microscopes utilizing Ti:Sapphire lasers, which are tunable in the ~680–1060 nm range, THG for detection. Additionally, in conventional multiphoton microscopes utilizing Ti:Sapphire lasers, imaging becomes problematic, because of the short THG wavelength (350 nm max) that has to be which are tunable in the ~680–1060 nm range, THG imaging becomes problematic, because of the short passed through microscope optics, mainly designed for the visible wavelength range. That is why THG wavelength (350 nm max) that has to be passed through microscope optics, mainly designed most THG imaging is performed using special longer wavelength short pulsed lasers. for the visible wavelength range. That is why most THG imaging is performed using special longer The DIVER detector has a transmission geometry, where emission photons are induced from wavelength short pulsed lasers. one side of a sample and collected from the opposite side. That makes the DIVER microscope very The DIVER detector has a transmission geometry, where emission photons are induced from one suitable for SHG and THG imaging, since these photons naturally propagate toward the detector. side of a sample and collected from the opposite side. That makes the DIVER microscope very suitable The ability of the DIVER detector to collect photons from the wide area of scattering samples makes for SHG and THG imaging, since these photons naturally propagate toward the detector. The ability of it even more preferable for this type of imaging [8,37]. Since the DIVER detector has very few optical the DIVER detector to collect photons from the wide area of scattering samples makes it even more elements and its spectral response is defined by the PMT and filters used, it detects photons in the preferable for this type of imaging [8,37]. Since the DIVER detector has very few optical elements and ~320–650 nm range. Thus, with the DIVER detector THG imaging can be easily performed using its spectral response is defined by the PMT and filters used, it detects photons in the ~320–650 nm Ti:Sapphire lasers. range. Thus, with the DIVER detector THG imaging can be easily performed using Ti:Sapphire lasers. SHG imaging in combination with FLIM using the DIVER detector was successfully applied in SHG imaging in combination with FLIM using the DIVER detector was successfully applied in kidney fibrosis and diabetes studies [19,34] where progression of the disease could be detected by kidney fibrosis and diabetes studies [19,34] where progression of the disease could be detected by this this method in earlier stages than by standard histology. Figure 6A shows an example of SHG method in earlier stages than by standard histology. Figure6A shows an example of SHG images of images of collagen fibers in healthy and diseased kidney tissue samples. The normal and diseased collagen fibers in healthy and diseased kidney tissue samples. The normal and diseased tissues can be tissues can be also recognized by their FLIM autofluorescence signature and SHG on the phasor plot. also recognized by their FLIM autofluorescence signature and SHG on the phasor plot.

Figure 6. Second Harmonic Generation (SHG) (A) and FLIM (B) images of normal and diseased kidney samplesFigure 6. imaged Second with Harmonic the DIVER Generation detector. (SHG) FLIM images(A) and are FLIM colored (B) according images of to normal cursors and in the diseased phasor plotkidney (C). samples imaged with the DIVER detector. FLIM images are colored according to cursors in the phasor plot (C). Examples of THG imaging are shown in Figure7, where 1050 nm light was used for imaging and an UG11Examples ﬁlter used of THG to separate imaging 350 are nm shown THG in signal. Figure 7, where 1050 nm light was used for imaging and anTHG UG11 does filter not involveused to separate molecules 350 in nm excited THG states signal. and thus allow imaging of features that do not ﬂuoresce, for example, such as lipid deposition in label-free tissues and cells. Figure7A,B shows THG

Methods Protoc. 2019, 2, 53 8 of 12 images of mouse adipose tissue. The cell membranes emit THG photons when they cross a focal point of the scanning IR light. Scanning inside the cells induces THG at edges of the adipocyte, but not in the interior where the refractive index is homogeneous. In other examples of THG imaging we used sciatic nerve and liver from mouse (Figure7C–E). THG imaging reveals the e ﬀect of high fat and normal diets on a mouse liver. Using autoﬂuorescence and THG imaging of A549 cells it was possible to show the presence of lamellar bodies in the interior of in vivo cells (Figure7F). Lamellar bodies are subcellular organelles associated with the accumulation and secretion of pulmonary surfactant, a substance mainly composed of phospholipids associated with the reduction of pulmonary surface tension [28]. Figure7G–I shows THG images of label-free bacteria in vivo, where bacteria were in an Methodsagarose Protoc. matrix 2019 to, 2, avoid x FOR movement.PEER REVIEW 8 of 12

Figure 7. Third Harmonic Generation (THG) images. (A) Cell membranes and (B) interfaces between

cells in adipose tissue. (C) 3D image of a mouse sciatic nerve. (D) and (E) lipids deposition in liver from Figurenon-fat 7. or Third fat diet Harmonic mouse respectively. Generation ((THG)F) Overlap images. of THG (A) Cell signal membranes (green) and and NADH (B) interfaces autofluorescence between cellsfrom inin adipose vivo A549 tissue. cells. (C The) 3D green image structures of a mouse refer sciatic to lamellar nerve. bodies (D) and (LBs). (E) lipids (G–I) 3D-THGdeposition images in liver of fromvarious non-fat bacteria. or fat diet mouse respectively. (F) Overlap of THG signal (green) and NADH autofluorescence from in vivo A549 cells. The green structures refer to lamellar bodies (LBs). (G–I) 3.3. Hyperspectral Imaging Using Two Filters 3D-THG images of various bacteria. Hyperspectral imaging is a common technique in fluorescence microscopy to obtain the emission spectrumTHG atdoes each not pixel involve of an image.molecules However, in excited methods states toand obtain thus spectral allow imaging resolution of features based on that diff ractiondo not fluoresce,gratings or for integrated example, prisms such as work lipid poorly deposition when thein label-free sample is tissues strongly and scattering. cells. Figure Spectral 7A,B phasors shows THGwere recentlyimages of introduced mouse adipose as an alternative tissue. The to spectralcell membranes demixing, emit for theTHG determination photons when of FRETthey cross and for a focalthe measurement point of the scanning of dipolar IR relaxation light. Scanning of probes inside in membranes the cells induces [1–8]. TheTHG basic at edges concept of the in the adipocyte, spectral butphasor not approachin the interior is that thewhere entire the spectrum refractive is index not needed is homogeneous. for some of theIn classicalother examples spectral analysisof THG imagingtechniques we such used as sciatic demixing, nerve but and rather liver onlyfrom a mous few parameterse (Figure 7C,D,E). of the spectral THG imaging distribution reveals are the suffi effectcient offor high these fat calculations. and normal Using diets the on properties a mouse liver. of the Using spectral autofluorescence phasors we can resolveand THG spectral imaging components of A549 cellsand performit was possible the type to ofshow data the analyses presence that of are lamellar usually bodies performed in the ininterior hyperspectral of in vivo imaging. cells (Figure 7F). LamellarRecently bodies we haveare subcellular shown [32] thatorganelles spectral associ phasorated coordinates with the Gaccumulation and S can be directlyand secretion calculated of pulmonaryfrom intensity surfactant, images, a passed substance through mainly optical composed filters with of phospholipids transmittance associated spectra “similar” with the to reduction sine and ofcosine pulmonary functions, surface using tension the following [28]. Figure equations: 7G–I shows THG images of label-free bacteria in vivo, where bacteria were in an agarose matrix to avoid movement. Icos = (ΣλFcos(λ)I(λ))/ΣλI(λ) 3.3. Hyperspectral Imaging Using Two Filters I = (ΣλF (λ)I(λ))/ΣλI(λ) Hyperspectral imaging is a commonsin techniquesin in fluorescence microscopy to obtain the emission spectrum at each pixel of an image. However, methods to obtain spectral resolution based on diffraction gratings or integrated prisms work poorly when the sample is strongly scattering. Spectral phasors were recently introduced as an alternative to spectral demixing, for the determination of FRET and for the measurement of dipolar relaxation of probes in membranes [1–8]. The basic concept in the spectral phasor approach is that the entire spectrum is not needed for some of the classical spectral analysis techniques such as demixing, but rather only a few parameters of the spectral distribution are sufficient for these calculations. Using the properties of the spectral phasors we can resolve spectral components and perform the type of data analyses that are usually performed in hyperspectral imaging. Recently we have shown [32] that spectral phasor coordinates G and S can be directly calculated from intensity images, passed through optical filters with transmittance spectra “similar” to sine and cosine functions, using the following equations:

Methods Protoc. 2019, 2, x FOR PEER REVIEW 9 of 12 Methods Protoc. 2019, 2, x FOR PEER REVIEW 9 of 12

cos Σλ cos λ λ Σλ λ I = (ΣλF (λ)I(λ))/ΣλI(λ) Methods Protoc. 2019, 2, 53 Icos= ( Fcos( )I( ))/ I( ) 9 of 12 Isin= (ΣλFsin(λ)I(λ))/ΣλI(λ) Isin= (ΣλFsin(λ)I(λ))/ΣλI(λ)

G=2(Icos-FcosMIN)/(FcosMAX-FcosMIN)-1 G = 2(I G=2(Icos Fcos-FcosMIN)/(F)/(FcosMAX-FcosMINF )-1 ) 1 − cosMIN cosMAX − cosMIN − S=2(Isin-FsinMIN)/(FsinMAX-FsinMIN)-1, S = 2(I sinS=2(IFsinsinMIN-FsinMIN)/(F)/(FsinMAXsinMAX-FsinMINFsinMIN)-1,) 1, λ λ − − − λ where Fsin(λ) and Fcos(λ) are transmittance of “sine” and “cosine” filters at wavelength λ respectively where F sin((λ)) and and F Fcoscos( (λ) )are are transmittance transmittance of of “sine” “sine” and and “cosine” “cosine” filters filters at at wavelength wavelength λ respectivelyrespectively and I(λ) is the total intensity measured without filters. andand I( λ)) is is the the total total intensity measured without filters. filters. We used cyan and green plastic filters as “cosine” and “sine” filters respectively, which were WeWe used cyancyan andand greengreen plastic plastic filters filters as as “cosine” “cosine” and and “sine” “sine” filters filters respectively, respectively, which which were were cut cut and placed in the DIVER detector filter wheel. The empty position in the filter wheel was used cutand and placed placed in the in the DIVER DIVER detector detector filter filter wheel. wheel. The The empty empty position position in thein the filter filter wheel wheel was was used used for for measuring the total intensity images. Filters spectra are shown in Figure 8B. These filters are not formeasuring measuring the the total total intensity intensity images. images. Filters Filters spectra spec aretra shownare shown in Figure in Figure8B. These 8B. These filters filters are not are ideal; not ideal; however, they resemble cos-sin functions within the 400–600 nm spectral range and can be ideal;however, however, they resemble they resemble cos-sin cos-sin functions functions within thewithin 400–600 the 400–600 nm spectral nm rangespectral and range can beand corrected can be corrected for the non-ideal response using correction factors [32]. correctedfor the non-ideal for the non-ideal response usingresponse correction using correction factors [32 factors]. [32].

Figure 8. Ideal (A) and real (B) sin-cos filters response after normalization and shifting to give the FigureFigure 8. IdealIdeal ( (AA)) and and real real ( (BB)) sin-cos sin-cos filters ﬁlters response response after after normalization normalization and and shifting shifting to to give give the the range of the cosine and sine function. rangerange of of the the cosine cosine and and sine function.

FigureFigure 99 shows shows images images and and corresponding corresponding spectral spec phasorstraltral phasorsphasors of fluorescent ofof fluorescentfluorescent bead and beadbead tissue andand samples, tissuetissue µ μ µ μ samples,acquired acquired using cos-sin using filters. cos-sin Orange filters. Orange 10 m and 10 μ Yellow-Greenm and Yellow-Green 15 m fluorescent15 μm fluorescent beads arebeads clearly are clearlyseparated separated in the phasor in the phasor plot and plot colored and colored according according to cursors to incursors the corresponding in the corresponding image. We image. have Weshown have earlier shown [18 earlier] that collagen[18] that I collagen and collagen I and III collagen area of theIII area mouse of femurthe mouse bone femur sample bone (Figure sample9C), (Figure(Figurecan be separated 9C),9C), cancan usingbebe separatedseparated the FLIM usingusing and SHG thethe imagingFLIMFLIM andand approach. SHGSHG imagingimaging We were approach.approach. able to demonstrate WeWe werewere the ableable same toto demonstrateresult using sin-costhe same filters result and using spectral sin-cos phasor filtersfilters analysis. andand spectralspectral phasorphasor analysis.analysis.

Figure 9. Application of hyperspectral imaging using the sin-cos filters and the spectral phasor Figure 9. Application of hyperspectral imaging using the sin-cos filters and the spectral phasor Figure 9. Application of hyperspectral imaging using the sin-cos filters and the spectral phasor transformation. (A) 15μµm Yellow-Green and 10μµm Orange fluorescent beads in a highly scattering transformation. (A) 15 μm Yellow-Green and 10 μm Orange fluorescent beads in a highly scattering transformation.matrix. (B) Spectral (A) phasor15 m Yellow-Green plot of beads in and (A): 10 using m orangeOrange and fluorescent green cursors beads it in is possiblea highly toscattering separate matrix. (B) Spectral phasor plot of beads in (A): using orange and green cursors it is possible to matrix.different (B types) Spectral of beads. phasor (C )plot Images of beads of mouse in (A): femur using bone. orange (D )and FLIM green phasor. cursors (E) it Spectral is possible phasor to separate different types of beads. (C) Images of mouse femur bone. (D) FLIM phasor. (E) Spectral separateacquired different with sin-cos types filters. of beads. Both ( FLIMC) Images and spectralof mouse phasors femur allowbone. identifying(D) FLIM phasor. areas with (E) diSpectralfferent phasor acquired with sin-cos filters. Both FLIM and spectral phasors allow identifying areas with phasorkind of acquired collagens with (I and sin-cos III). filters. Both FLIM and spectral phasors allow identifying areas with different kind of collagens (I and III). different kind of collagens (I and III).

Methods Protoc. 2019, 2, 53 10 of 12

4. Discussion We have shown that the DIVER detector, which is simple to construct, may significantly improve the imaging depth in turbid samples. The DIVER detector utilizes two main principles: (1) use of wide area detector that allows collecting scattered emission photons from a wide surface area of a turbid sample; (2) the refraction index is matched in the optical path sample-PMT window to minimize signal attenuation due to reflections on optical interfaces. Implementation of these two principles allows harvesting emission photons much more efficiently than with an epi-detector. The DIVER detector becomes more sensitive than the epi-detector to low level signals with an increase in imaging depth and at certain depths scattering samples could be imaged only using the DIVER detector. The transmission geometry of the DIVER detector obviously leads to some limitations for its use, because emission photons, induced from one side of a sample, have to travel to other side of a sample to reach the detector. If a sample has strong absorption, these photons will be attenuated and not detected. However, for most biological tissues the absorption coefficient is low and emission photons are mainly attenuated due to scattering, rather than absorption. In this case they still have a chance to reach an opposite surface of a sample and be detected. It was shown that in multiple scattering media attenuation by scattering is linear with the distance, while in clear media the signal decays proportionally to the square of the distance [9,41]. For this reason, at certain depths the scattering may even “boost” the signal. We have shown that in the absence of strong absorption, scattering samples even of few cm thick can be imaged using the DIVER detector. On the other hand, the transmission geometry is very useful for SHG and THG imaging, because induced SHG and THG photons are directed toward the detector. This fact and the DIVER detector ability to collect scattered photons from a wide sample area make it orders of magnitude more sensitive for Harmonic Generation Imaging in turbid samples than an epi-detector [8]. Moreover, THG imaging is easily available with the DIVER detector using conventional Ti:Sapphire lasers for excitation, while with epi-detectors it will require special long-wavelength lasers or microscope optics for Ultraviolet (UV) range. We also demonstrated that using inexpensive plastic cyan-green filters we can obtain spectral phasors of samples and use them for hyperspectral image analysis. We note that there are no adjustable parameters involved in the procedure and the only information needed is the transmission of filters in the region of sensitivity of the detector. Fluorescence collected after passing through a strongly scattering media gives the same position of the spectral phasor as for transparent material. This hyperspectral imaging method using sin-cos filters is not specific for the DIVER detector and can be used potentially with any microscope and detector, however application to scattering media makes it a very useful addition to the DIVER system. Currently we are working on implementation of other, “perfect” sin-cos interference filters for hyperspectral imaging, as well as using adaptive optics technique to correct the point spread function distortion by turbid media and achieve higher imaging depths.

5. Patents Enrico Gratton, Alexander Dvornikov, Viera Crosignani. Apparatus and Method for Light Emission Detection for In-Depth Imaging of Turbid Media. U.S. Pat. 8692998.

Author Contributions: Conceptualization, A.D., L.M. and E.G.; software, E.G.; formal analysis, A.D. and L.M.; investigation, A.D. and L.M.; writing—original draft preparation, A.D. and L.M.; writing—review and editing, E.G.; visualization, A.D. and L.M. Funding: This research was funded by NIH P50 GM076516. NIH P41-GM103540 grants. Acknowledgments: The authors are indebted to Arturo Briva, Moshe Levi, Suman Ranjit, Rupsa Datta, Arunima Bhattacharjee for providing some of the samples and figures. LM was partially supported as a full time professor at the Universidad de la República-Uruguay. Conflicts of Interest: The authors declare no conflict of interest. Methods Protoc. 2019, 2, 53 11 of 12

References

1. O’Malley, D. Imaging in Depth: Controversies and Opportunities. In Methods in Cell Biology; Elsevier Inc.: Amsterdam, The Netherlands, 2008; Volume 89, pp. 95–128. 2. Theer, P.; Hasan, M.T.; Denk, W. Two-photon imaging to a depth of 1000 µm in living brains by use of a

Ti:Al2O3 regenerative amplifier. Opt. Lett. 2007, 28, 1022–1024. [CrossRef] 3. Kobat, D.; Durst, M.E.; Nishimura, N.; Wong, A.W.; Schaffer, C.B.; Xu, C. Deep tissue multiphoton microscopy using longer wavelength excitation. Opt. Express 2009, 17, 13354–13364. [CrossRef][PubMed] 4. Cicchi, R.; Pavone, F.S.; Massi, D.; Sampson, D.D. Contrast and depth enhancement in two-photon microscopy of human skin ex vivo by use of optical clearing agents. Opt. Express 2005, 13, 2337–2344. [CrossRef] 5. Riley, J.D.; Knutson, J.R.; Combs, C.A.; Smirnov, A.V.; Balaban, R.S.; Gandjbakhche, A.H. Optimization of multiphoton excitation microscopy by total emission detection using a parabolic light reflector. J. Microsc. 2007, 228, 330–337. 6. Crosignani, V.; Dvornikov, A.S.; Gratton, E. Enhancement of imaging depth in turbid media using a wide area detector. J. Biophotonics 2011, 4, 592–599. [CrossRef][PubMed] 7. Crosignani, V.; Dvornikov, A.; Aguilar, J.S.; Stringari, C.; Edwards, R.; Mantulin, W.W.; Gratton, E. Deep tissue fluorescence imaging and in vivo biological applications. J. Biomed. Opt. 2012, 17, 116023. [CrossRef] [PubMed] 8. Crosignani, V.; Jahid, S.; Dvornikov, A.S.; Gratton, E. A deep tissue fluorescence imaging system with enhanced SHG detection capabilities. Microsc. Res. Tech. 2014, 77, 368–373. [CrossRef][PubMed] 9. Dvornikov, A.; Gratton, E. Imaging in turbid media: A transmission detector gives 2-3 order of magnitude enhanced sensitivity compared to epi-detection schemes. Biomed. Opt. Express 2016, 7, 3747–3755. [CrossRef] 10. Campagnola, P.J.; Dong, C.Y. Second harmonic generation microscopy: Principles and applications to disease diagnosis. Laser Photonics Rev. 2011, 5, 13–26. [CrossRef] 11. Barad, Y.; Eisenberg, H.; Horowitz, M.; Silberberg, Y. Nonlinear scanning laser microscopy by third harmonic generation. Appl. Phys. Lett. 1997, 70, 922–924. [CrossRef] 12. Zipfel, W.R.; Williams, R.M.; Webb, W.W. Nonlinear magic: Multiphoton microscopy in the biosciences. Nat. Biotechnol. 2003, 21, 1369–1377. [CrossRef][PubMed] 13. Tserevelakis, G.J.; Megalou, E.V.; Filippidis, G.; Petanidou, B.; Fotakis, C.; Tavernarakis, N. Label-free imaging of lipid depositions in C. elegans using third-harmonic generation microscopy. PLoS ONE 2014, 9, e84431. [CrossRef][PubMed] 14. Chen, X.; Raggio, C.; Campagnola, P.J. Second-harmonic generation circular dichroism studies of osteogenesis imperfecta. Opt. Lett. 2012, 37, 3837–3839. [CrossRef][PubMed] 15. Nadiarnykh, O.; LaComb, R.B.; Brewer, M.A.; Campagnola, P.J. Alterations of the ECM in Ovarian Carcinogenesis Studied by Second Harmonic Generation Imaging Microscopy. BMC Cancer 2010, 10. [CrossRef][PubMed] 16. Ajeti, V.; Nadiarnykh, O.; Ponik, S.M.; Keely, P.J.; Eliceiri, K.W.; Campagnola, P.J. Structural changes in mixed Col I/Col V collagen gels probed by SHG microscopy: Implications for probing stromal alterations in human breast cancer. Biomed. Opt. Express 2011, 2, 2307–2316. [CrossRef][PubMed] 17. Plotnikov, S.V.; Kenny, A.M.; Walsh, S.J.; Zubrowski, B.; Joseph, C.; Scranton, V.L.; Kuchel, G.A.; Dauser, D.; Xu, M.; Pilbeam, C.C.; et al. Measurement of muscle disease by quantitative second-harmonic generation imaging. J. Biomed. Opt. 2008, 13, 044018. [CrossRef] 18. Ranjit, S.; Dvornikov, A.; Stakic, M.; Hong, S.H.; Levi, M.; Evans, R.M.; Gratton, E. Imaging Fibrosis and Separating Collagens using Second Harmonic Generation and Phasor Approach to Fluorescence Lifetime Imaging. Sci. Rep. 2015, 5, 13378. [CrossRef] 19. Ranjit, S.; Dvornikov, A.; Levi, M.; Furgeson, S.; Gratton, E. Characterizing fibrosis in UUO mice model using multiparametric analysis of phasor distribution from FLIM images. Biomed. Opt. Express 2016, 7, 3519–3530. [CrossRef] 20. Aguilar-Arnal, L.; Ranjit, S.; Stringari, C.; Orozco-Solis, R.; Gratton, E.; Sassone-Corsi, P. Spatial dynamics of SIRT1 and the subnuclear distribution of NADH species. Proc. Natl. Acad. Sci. USA 2016, 113, 12715–12720. [CrossRef] 21. Ma, N.; Digman, M.A.; Malacrida, L.; Gratton, E. Measurements of absolute concentrations of NADH in cells using the phasor FLIM method. Biomed. Opt. Express 2016, 7, 2441–2452. [CrossRef] Methods Protoc. 2019, 2, 53 12 of 12

22. Bastiaens, P.I.H.; Squire, A. Fluorescence lifetime imaging microscopy: Spatial resolution of biochemical processes in the cell. Trends Cell Biol. 1999, 9, 48–52. [CrossRef] 23. Van Munster, E.B.; Gadella, T.W.J. Suppression of photobleaching-induced artifacts in frequency-domain FLIM by permutation of the recording order. Cytom. Part A 2004, 58, 185–194. [CrossRef][PubMed] 24. Becker, W.; Bergmann, A.; Hink, M.A.; König, K.; Benndorf, K.; Biskup, C. Fluorescence Lifetime Imaging by Time-Correlated Single-Photon Counting. Microsc. Res. Tech. 2004, 63, 58–66. [CrossRef][PubMed] 25. Jameson, D.M.; Gratton, E.; Hall, R.D. The Measurement and Analysis of Heterogeneous Emissions by Multifrequency Phase and Modulation Fluorometry. Appl. Spectrosc. Rev. 1984, 20, 55–106. [CrossRef] 26. Digman, M.A.; Caiolfa, V.R.; Zamai, M.; Gratton, E. The Phasor Approach to Fluorescence Lifetime Imaging Analysis. Biophys. J. 2008, 94, L14–L16. [CrossRef][PubMed] 27. Fereidouni, F.; Bader, A.N.; Gerritsen, H.C. Spectral phasor analysis allows rapid and reliable unmixing of ﬂuorescence microscopy spectral images. Opt. Express 2012, 20, 12729–12741. [CrossRef][PubMed] 28. Malacrida, L.; Astrada, S.; Briva, A.; Bollati-Fogolín, M.; Gratton, E.; Bagatolli, L.A. Spectral phasor analysis of LAURDAN ﬂuorescence in live A549 lung cells to study the hydration and time evolution of intracellular lamellar body-like structures. Biochim. Biophys. Acta-Biomembr. 2016, 1858, 2625–2635. [CrossRef] 29. Malacrida, L.; Gratton, E.; Jameson, D.M. Model-free methods to study membrane environmental probes: A comparison of the spectral phasor and generalized polarization approaches. Methods Appl. Fluoresc. 2015, 3, 047001. [CrossRef][PubMed]

30. Malacrida, L.; Gratton, E. LAURDAN fluorescence and phasor plots reveal the effects of a H2O2 bolus in NIH-3T3 fibroblast membranes dynamics and hydration. Free Radic. Biol. Med. 2018, 128, 144–156. [CrossRef] 31. Sameni, S.; Malacrida, L.; Tan, Z.; Digman, M.A. Alteration in Fluidity of Cell Plasma Membrane in Huntington Disease Revealed by Spectral Phasor Analysis. Sci. Rep. 2018, 8, 734. [CrossRef] 32. Dvornikov, A.; Gratton, E. Hyperspectral imaging in highly scattering media by the spectral phasor approach using two filters. Biomed. Opt. Express 2018, 9, 4833–4840. [CrossRef][PubMed] 33. Ayers, F.; Grant, A.; Kuo, D.; Cuccia, D.J.; Durkin, A.J. Fabrication and characterization of silicone-based tissue phantoms with tunable optical properties in the visible and near infrared domain. Proc. SPIE 2008, 6870. [CrossRef] 34. Ranjit, S.R.; Dvornikov, A.; Dobrinskikh, E.; Wang, X.; Luo, Y.; Levi, M.; Gratton, E. Measuring the effect of a Western diet on liver tissue architecture by FLIM autofluorescence and harmonic generation microscopy. Biomed. Opt. Express 2017, 8, 371–378. 35. Negrean, A.; Mansvelder, H.D. Optimal lens design and use in laser-scanning microscopy. Biomed. Opt. Express 2014, 5, 1588–1609. [CrossRef][PubMed] 36. Suhalim, J.L.; Chung, C.Y.; Lilledahl, M.B.; Lim, R.S.; Levi, M.; Tromberg, B.J.; Potma, E.O. Characterization of cholesterol crystals in atherosclerotic plaques using stimulated Raman scattering and second-harmonic generation microscopy. Biophys. J. 2012, 102, 1988–1995. [CrossRef] 37. Ranjit, S.; Dobrinskikh, E.; Montford, J.; Dvornikov, A.; Lehman, A.; Orlicky, D.J.; Nemenoff, R.; Gratton, E.; Levi, M.; Furgeson, S. Label-free fluorescence lifetime and second harmonic generation imaging microscopy improves quantification of experimental renal fibrosis. Kidney Int. 2016, 90, 1123–1128. [CrossRef][PubMed] 38. Tjin, G.; Xu, P.; Kable, S.H.; Kable, E.P.W.; Burgess, J.K. Quantification of collagen I in airway tissues using second harmonic generation. J. Biomed. Opt. 2014, 19, 36005. [CrossRef] 39. Sun, C.-K.; Yu, C.-H.; Tai, S.-P.; Kung, C.-T.; Wang, I.-J.; Yu, H.-C.; Huang, H.-J.; Lee, W.-J.; Chan, Y.-F. In vivo and ex vivo imaging of intra-tissue elastic fibers using third-harmonic-generation microscopy. Opt. Express 2007, 15, 11167–11177. 40. Débarre, D.; Supatto, W.; Pena, A.M.; Fabre, A.; Tordjmann, T.; Combettes, L.; Schanne-Klein, M.C.; Beaurepaire, E. Imaging lipid bodies in cells and tissues using third-harmonic generation microscopy. Nat. Methods 2006, 3, 47–53. [CrossRef] 41. Theer, P.; Denk, W. On the fundamental imaging-depth limit in two-photon microscopy. J. Opt. Soc. Am. A 2006, 23, 3139–3149. [CrossRef]

© 2019 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).