Open Access
Add to BibSonomyAbstract
We report the design and implementation of a multispectral imaging Fourier transform endospectroscopy (IFTES) system. The IFTES system employs a flexible fiber bundle catheter coupled to a home-built imaging Fourier transform spectroscope. The instrument enables the performance of non- or minimally invasive subsurface imaging and multispectral imaging at the cellular level in vivo and in situ. A maximum spectral resolution of 0.2 nm at 632.8 nm and a lateral resolution of 4.4 μm were proved. Preliminary results of a standard resolution target, ex-vivo small animal tissue, single wavelength laser, fluorescence solution, in-vivo mouse skin, microspheres mixture, and in-vivo transgenic mouse brain were given to demonstrate the potential of the technique.
©2012 Optical Society of America
1. Introduction
Intravital optical imaging is an essential research tool to use to understand cellular and molecular activities and functions, and furthermore to understand fundamental biological processes in a native anatomical context. However, traditional optical imaging in vivo is usually achieved under a fluorescence microscope, where the bulky objective lens does not permit access to any internal organs in the body [1,2]. There is an urgent need to develop new technologies and instruments to acquire images with high resolution in vivo for visualization of cellular morphology and architecture in situ through non- or minimally invasive ways.
Fluorescence endoscopic technologies, such as single fiber scanning endoscopy [3–5], nonlinear fiber endoscopy [6], fiber bundle microendoscopy [7], and so on, have been developed to more effectively detect and localize critical, early pathologic changes and cell activities occurring in the epithelial and subepithelial regions in vivo and in situ. Among these endoscopic technologies, fiber bundle microendoscopy is very attractive since it, unlike other means based on a single fiber, eliminates the requirement of a scanning mechanism at the distal end of a fiber bundle catheter. Lateral resolution of fiber bundle microendoscopy is determined by the distance between adjacent fiber cores of the fiber bundle [7], which is enough to detect the presence of disease at the subcellular level [8]. A miniaturized outer diameter allows the fiber bundle catheter to be easily compatible with conventional biopsy and therapy operations.
Technical advances of fiber bundle microendoscopy focus on improving optical imaging quality and adding more functions. Several different illumination modes, such as structure illumination [9], hybrid illumination [10,11], selective plane illumination [12], and dynamic speckle illumination [13], have been introduced to obtain optical sectioning capability. Besides intensity imaging, fiber bundle microendoscopy has also been demonstrated to achieve fluorescence lifetime imaging (FLIM) [14–17] and spectral imaging [18–21] to extend the applications of fiber bundle microendoscopy. Endoscopic FLIM [14–17] can distinguish fluorophores by their lifetime difference and can monitor cellular or molecular interactions and local environmental fluctuations [22]. However, lifetime measurement requires expensive imaging systems and long data acquisition time of more than 100 s [14,15], which is prohibitive in most applications in vivo.
Multispectral endoscopic imaging, relatively economical and rapid, is capable of observing and discriminating various fluorescent labels simultaneously by relying on spectral information [23–25]. Florence Jean et al. [19] developed a fiber-bundle-based system with dual-wavelength excitation and dual detection channels coupled with a commercial spectrometer. This system can obtain a full spatial profile in two channels and spectral information only averaged over the entire field of view. Timothy J. Muldoon et al. [8] presented fiber optic microendoscopy with the ability of imaging and distinguishing three kinds of fluorescent labels within a single sample. The system can sequentially acquire images with three separate emission filters or simultaneously image the sample with a CCD camera’s Bayer mask. But the system was only suitable for spectrally distinct fluorescent labels due to limited spectral resolution. Gmitro et al. [18,20,21] demonstrated their fiber-bundle-based system capable of providing detailed spectral information associated with every image point. However, the dispersive component in their system spread the signals of different wavelengths at different spatial positions, which induced a non-uniform spectral resolution. Image mapping spectroscopy (IMS) [26] has been introduced into a multispectral endoscope [27] to collect spectral imaging information in a single snapshot. In this technology the number of spectral imaging data cube voxels is limited by the number of pixels of the camera. Spectral imaging based on Fourier transform spectral reconstruction is another potential solution for spectral imaging in vivo. This solution has several advantages because of the adoption of Fourier transform spectral measurements [28]: high, variable, and unlimited spectral resolution; wide spectral range; high throughput; simultaneous measurement of all of the frequencies; and internal calibration. To our knowledge, a flexible endoscopic version of spectral imaging based on Fourier transform spectral measurement has not been demonstrated.
In this paper, we propose a method of imaging Fourier transform endospectroscopy (IFTES) that is able to measure fluorescence emission spectra simultaneously of all points in the field of view (FOV) in vivo and in situ by transmitting the emission signal through a fiber bundle into an imaging Fourier transform spectroscope. As the proof of principle, preliminary multispectral imaging results acquired by our system are presented. The results show that this instrument can enhance the flexibility of choosing multiple spectral-overlapping fluorescent probes and extend the potential applications of fiber bundle microendoscopy.
2. Principle and setup
2.1 Setup
Figure 1 depicts a schematic of the IFTES system. The system consisted of a light source, a full field epi-illumination configuration, a fiber bundle catheter, a Michelson interferometer, and an imaging detector. A 488 nm laser (CYAN 488-50 CDRH, Spectra-Physics) provided collimated fluorescent excitation illumination and was transmitted into a full-field epi-illumniation configuration, which was composed of a tube lens, a 500 nm cutoff dichroic mirror (DM1, Semrock), and a 10×/0.25 NA infinity-corrected objective lens (Olympus). A beam expander located before the tube lens expanded the beam size to satisfy the incident condition of the objective lens. The objective lens coupled the beam into the proximal end of a flexible fiber bundle catheter. Excitation light was transferred coherently to the distal end of the catheter and illuminated the sample. Fluorescence emission from the sample was collected back through the catheter, the objective lens, and DM1 and then sent into a self-built Michelson interferometer. The beam was divided by a beam splitter into two parts. One part was transmitted toward a fixed aluminum-coated mirror, M2. Another part was reflected toward a movable mirror, M1, which was fixed on a frictionless, high-precision flexure piezoelectric translation (PZT) stage (P-629.1CD, Physik Instrumente). A large sequence of imaging interferogram frames from the output port of the interferometer was obtained by a cooled digital sCMOS camera (ORCA-Flash 2.8, Hamamastu), which provided the multiple benefits of high resolution, high readout speed, and low noise all in one. The image captured by the camera was optically conjugated to the fluorescent emission on the proximal end of the fiber bundle. The imaging magnification, M, was equal to the focal length ratio of the camera lens and the objective lens. After Fourier transform calculation and relevant data processing, an x/y/λ data cube of multispectral imaging was reconstructed.
Fig. 1 Imaging Fourier transform endospectroscopy (IFTES). DM: dichroic mirror; M: mirror; Obj: Objective lens; LP: long pass filter.
The images of the tissue surface were acquired through the fiber bundle in our system. The FOV is determined by the image circle diameter of the fiber bundle catheter. The image guide (Fujikura, FIGH-30-850N) contains 30,000 fibers, with an overall outer diameter of 950 μm and an active image diameter of 790 μm. The effective area of the camera is 6.97 mm (H)×5.23 mm (V). The pixel number of the camera is 1920 (H)×1440 (V), and the pixel size is 3.63 μm (H)×3.63 μm (V). According to the Nyquist sampling theorem, at least 2 pixels are needed to represent a corresponding fiber core; then, the maximum FOV is 1584 μm [8]. The actual relay magnification, M, is chosen as 6.6×. Therefore, there are 4 pixels presenting per fiber core at the camera, which is sufficient for sampling individual fibers.
In order to monitor the movement of M1, a small portion of the excitation beam was reflected by a beam pickoff and another 500 nm cutoff dichroic mirror, DM2, and transmitted into the same interferometer. The interference fringes were projected into an amplified Si photo detector (PDA36A-EC, Thorlabs). The monitor channel guarantees the accuracy of the Fourier transform spectroscopy with long-term stability and eliminates extra spectral calibrations [29,30]. Furthermore, this arrangement avoids extra reference signals and makes the whole layout compact and cost-effective.
When the optical path difference (OPD) between the interferometer arms is scanned, the intensity of the emission light after the interferometer is modulated by interference. The OPD scan is achieved by the translation of the PZT stage. A maximum-stage travel range of 1500 μm enables us to provide a theoretical highest spectral resolution of ~4.02 cm−1 (~0.14 nm at 600 nm) [31]. Adopting beam-folding technology can improve the spectral resolution twofold or better [32]. In theory, the spectral resolution is unlimited. A full FOV frame rate of up to 45 fps also allows the IFTES system to provide real-time full FOV gray-scale images, providing a potential of monitoring dynamic processes of biological applications. The system can directly switch between gray-scale imaging and multispectral imaging modes without any extra adjustment of the configuration.
2.2 Imaging approach
Technically, IFTES can be implemented by combining Fourier transform spectroscopy with both wide-field and confocal imaging approaches through a coherent fiber bundle. However, considering the resolution, complexity, and stability of the system as well as the potential of enhancing the imaging throughput of the system, of the two approaches the former is superior to the latter for IFTES. First, the lateral spatial resolution of IFTES is irrelevant to the imaging approach and is determined by the core-to-core spacing between two neighboring fibers in the bundle catheter [7], which is <4 μm in this case. A wide-field approach reduces the optical sectioning strength of IFTES in comparison with truly confocal techniques, which results in increased collection of background light and reduced image contrast [33]. However, contact imaging and bright fluorophores enable the wide-field IFTES approach to achieve high image quality [8]. IFTES only involves one mirror movement of OPD scanning in the wide-field approach, while three scanning mirrors have to be driven and synchronized simultaneously in the confocal approach: two for x/y scanning and the third one for high-speed OPD scanning. This results in an increase in complexity and a decrease in stability of IFTES. Compared with pixel-by-pixel collection of confocal IFTES, the parallel acquisition of the wide-field approach [34] makes enhancing the image throughput of IFTES without sacrificing acquisition speed easy. Spatial image ranges can be readily broadened at the same frame rate owing to the technological advances of sCMOS and associated equipment. For example, the maximum image range at 100 fps can be expanded from 1920 (H)×600 (V) pixels by the sCMOS camera used here to 2048 (H)×2048 (V) pixels by the new style camera (ORCA-Flash 4.0, Hamamastu). To attain the same improvement on image throughput in the confocal approach, IFTES has to shorten pixel dwelling time or prolong total data acquisition time.
3. Results
3.1 Spatial resolution
Technical implementation of IFTES is composed of microendoscopic imaging and spectral detection. A 3D spectral data cube, named as λ-cube or λ-stack, is acquired to present the spatial profile (x/y) of the sample and the spectral characteristic (λ) of each pixel in the x/y plane. Here, the spatial imaging ability of the IFTES system was validated first. To evaluate the effective lateral resolution of IFTES, the US Air Force target (Edmund Optics) was put in contact with the distal end of the fiber bundle [Fig. 2(a) ]. The illumination wavelength of IFTES is 488 nm, but the detection channel is restricted from 500 nm to 700 nm by dichroic mirrors and emission filters. Since the target is not a fluorescence sample, a cold light source (LG-PS2, Olympus) was employed to illuminate from the rear side of the target for imaging. Minimally resolved bars are of the group 6, element 6 at the lower right corner of the image [Fig. 2(b)], corresponding to the actual resolution of about 4.4 μm.
Fig. 2 (a) Image of the USAF target used to determine spatial resolution illuminated by a cold light source from the rear side. (b) Enlarged detail of the red rectangle in (a). Scale bar in (a) is 100 μm.
Imaging experiments of excised animal tissue were carried out to demonstrate the spatial imaging ability of IFTES for fluorescence samples. Excised liver, pancreas, kidney, and peritonaeum tissue samples of normal Kunming mice were stained with 0.1% acriflavine [35] and imaged immediately. A Gauss profile of laser excitation illumination brought about the brighter central parts of the images compared with the edge parts. The non-uniform brightness of the images can be corrected by calibration as shown in Fig. 3 . In Fig. 3(a), hepatocytes contacting with one another are arranged linearly into hepatic cords. These cords are separated by adjacent vascular sinusoids. Figure 3(b) shows that pancreatic cells are evenly distributed, and the microvasculature of the pancreas can be easily identified. The kidney in Fig. 3(c) exhibits microstructural features of the renal cortex. In Fig. 3(d), the long and narrow structures in the image are muscle fibers of peritoneum, and the bright spots are the nuclei of peritoneum cells.
Fig. 3 Images from excised normal mouse tissues: (a) liver, (b) pancreas, (c) kidney, and (d) peritoneal wall. Scale bar is 100 μm.
3.2 Spectral resolution
The spectrum of a He-Ne laser (Gangdong) was measured by our system to demonstrate the highest spectral resolution at maximum OPD scanning. The results are shown in Fig. 4 . Figure 4(a) shows an arbitrary interferogram frame of the fiber bundle back illuminated by a He-Ne laser. Figure 4(b) is an enlarged detail of Fig. 4(a). To achieve the highest spectral resolution, a maximum OPD of 1500 μm was scanned and a total number of 20k frames were acquired at 45 fps. Figure 4(c) plots its spectrum reconstructed by the Fourier transform, where the intensity peak occurs accurately at 632.8 nm. The spectral resolution at full width at half-maximum (FWHM) intensity is 0.2 nm at wavelength 632.8 nm, which is close to the theoretical resolution of 0.16 nm at wavelength 632.8 nm. This tiny difference is because of a small scanning misalignment of the interferometer. Though the coherence of the He-Ne laser induced the speckles on the image of the fiber bundle in Figs. 4(a) and 4(b), it did not influence the measurement of the best spectral resolution of the IFTES system.
Fig. 4 Spectral imaging of a He-Ne Laser. (a) Interferogram frame of the fiber bundle illuminated by a He-Ne laser (false color by ImageJ software; the same below). (b) Enlarged detail of the blue rectangle in (a). (c) Spectrum of the red pixel pointed out by the blue arrow in (b). Scale bar is 100 μm.
3.3 Spectral imaging rate
It is important to achieve fast spectral imaging in most intravital biological applications. Data acquisition time of spectral imaging by IFTES T is the time of recording a whole interference fringe over OPD scanning, which is dependent on OPD scanning amplitude d, OPD scanning step ∆d, and the frame rate of the camera fr:
The OPD scanning amplitude d is determined by the desired spectral resolution ∆λ, as shown here [31],
where λ is the detection wavelength. By applying Eqs. (1) and (2), data acquisition time T can be written asFrom Eq. (3) at a certain OPD scanning step ∆d, data acquisition time T is inversely proportional to the desired spectral resolution ∆λ and the frame rate fr of the camera in the IFTES system.
From Eq. (3), the data acquisition time of achieving the highest spectra resolution by the IFTES system has to reach the order of minutes, limited by the full resolution frame rate of the camera. This is unacceptable for intravital biological applications. Moreover, the highest spectral resolution of about 0.1 nm by the IFTES system is too high for most multi-labeled biological experiments. Emission spectra of most spectrally distinct fluorescent proteins and dyes have FWHMs of about 20 nm or more, and the spectral peak distance between different fluorescent labels for the same experiment are usually chosen at more than 20 nm. Reducing spectral resolution appropriately can decrease desired scanning amplitude and save data acquisition time. A tradeoff between spectral resolution and data acquisition time can be made according to different application requirements.
To verify this point, emission spectra of a fluorescein sodium solution as an instance were measured by IFTES with different spectral resolutions. This kind of fluorescent dye is popular and safe as a clinical image contrast medium approved by the U.S. Food and Drug Administration (FDA) [36]. The fiber bundle catheter was immersed into a solution of 0.1% fluorescein sodium and imaged in full FOV. Since spectral resolutions ∆λ of a Fourier transform measurement are randomly tunable, several typical spectral resolutions ∆λ of 21.8 nm, 10.9 nm, 2.2 nm, and 0.2 nm at 521 nm in wavelength were tested in our system. The corresponding OPD scanning amplitudes are: 7.5 μm, 15 μm, 75 μm, and 750 μm, repectively. For oversampling, the scanning step ∆d was equal to 75 nm; thus, the actual data acquisition times T were 2.2 s, 4.4 s, 22.2 s, and 222.2 s, shown as red circles in Fig. 5(a) , respectively. The corresponding reconstructed emission spectra with different spectral resolutions are presented as the blue lines in Figs. 5(b)–5(e). All the lines are compared with the red line representing the measurement result by a commercial spectrometer (USB2000, Ocean Optics) in Figs. 5(b)–5(e). The FWHM and peak wavelength of the emission spectrum of fluorescein sodium are 23.6 nm and 521 nm, respectively. Since the spectral resolution of 21.8 nm is more than half of the FWHM of the spectrum, there is an obvious difference between the lines in Fig. 5(b), but the other three measurement results of different spectral resolutions by IFTES are consistent with the spectrum by the commercial spectrometer in Figs. 5(c)–5(e). It implies that spectral resolution equal to or less than half of the FWHM of the spectrum is enough to identify the spectral characteristic of fluorescein sodium in this experiment. The data acquisition time can be reduced to several seconds, even sub-seconds, which is short enough for most intravital biological applications. In other words, the IFTES system can balance spectral resolution and data acquisition time according to different fluorescent applications. Also, the results indicate that the spectral transmission characteristics of the fiber bundle used for the IFTES system evidently do not influence the spectral measurement for visible wavelength.
Fig. 5 (a) Data acquisition time T using various spectral resolution ∆λ (blue line). Red circles represent corresponding T by ∆λ of 21.8 nm, 10.9 nm, 2.2 nm, and 0.2 nm at 521 nm in wavelength measured by IFTES system, respectively. (b)–(e) Corresponding normalized emission spectra of fluorescein sodium solution with spectral resolution ∆λ of 21.8 nm, 10.9 nm, 2.2 nm and 0.2 nm at 521 nm in wavelength measured by IFTES system (blue lines) and commercial spectroscopy (red lines), respectively.
From Eq. (3), it can be found that another potential method to enhance the speed of spectral imaging in the IFTES system is to decrease the number of pixels to be read out and increase the frame rate of the camera with a sub-array function. In the sub-array readout mode, it is possible to enhance the frame rate from 60 fps [1920 (H)×1080 (V)] to 1273.6 fps [1920 (H)×8 (V)] without sacrificing the readout noise performance. All of them are faster than the full resolution readout speed of 45.5 fps [1920 (H)×1440 (V)]. Therefore, only obtaining the spectral imaging data set from the region of interest (ROI) in a full resolution image can further shorten the data acquisition time of the spectral imaging. Furthermore, another advantage is to avoid unwanted photo damage, photo bleaching, and photo toxicity effects on the samples due to long time exposure.
To demonstrate that choosing the ROI appropriately to shorten data acquisition time is feasible for biological applications, a spectral imaging experiment of healthy mouse skin in situ was performed. The fiber bundle used here and in experiments below was IGN-037/10 (Sumitomo) that has an image circle diameter of 333 μm. To narrow the excitation illumination range on the proximal tip of the fiber bundle, a 20× objective lens (Olympus) substituted for the 10× objective lens in the IFTES system. The C57 mouse was fixed in a custom-made mount after removing hair, and then topical acridine orange (0.15%) was applied to the skin surface. Figure 6(a) shows a full FOV image. The vascular network distribution is clearly visible, and the bright spots are the nuclei of keratinocytes. A sub-array frame of the cells size of 256×256 pixels was chosen as the ROI enclosed by the red rectangle in Fig. 6(a). The OPD scanning step is 120 nm for oversampling. The OPD scanning range is 20 μm, and the corresponding spectral resolution is about 10.8 nm at 600 nm. A group of interferogram frames of ROI were obtained at 200 fps in 0.83 s. The emission spectrum of the pixel pointed out by a red cross in Fig. 6(a) is shown in Fig. 6(b) as a blue line that has an intensity peak at about 529 nm. Though the spectral resolution here is not as high as possible, the curve is still consistent with the measurement result of the fluorescent labeling solution by the commercial spectrometer plotted as a red line in Fig. 6(b).
Fig. 6 Spectral imaging results from C57 mouse skin in situ. (a) Full FOV image acquired by the fiber bundle in the living C57 mouse. Red rectangle represents ROI range. Scale bar is 50 μm. (b) Emission spectra of the bright spot pointed out by the red cross in (a) by the IFTES system and commercial spectroscopy are plotted as blue and red lines, respectively.
3.4 Spectral imaging
Spectral imaging experiments were performed by the IFTES system to validate its ability to distinguish multi-label objects. In order to efficiently excite some dyes, a 561 nm laser (Spectra-Physics) was added into the excitation light source unit. Then the DM1 and LP filters were replaced by a dual-edge laser-flat dichroic beam splitter (Di01-R488/561-25×36, Semrock) and a dual-band-passing filter (Em01-R488/568-15, Semrock), respectively, which resulted in a block around 561 nm in the following spectra measured by the IFTES system.
A slide containing a mixture of 10.0–14.0 μm yellow and purple fluorescent microspheres (Spherotech) was used in the experiment. Since a coverslip blocked contact imaging of the beads with the fiber tip, two infinity-corrected 20×/0.75 NA objectives (Olympus) were mounted back-to-back and inserted between the distal end of the fiber bundle and the slide for a 1:1 relay. The experimental parameters of OPD scanning and data acquisition were the same as the ones of the last experiment. Figure 7(a) shows an arbitrary sub-array interferogram frame size of 256×256 pixels in which three microspheres link together and each one covers about four fibers in diameter. Figure 7(b) presents a λ-cube of the microspheres acquired by the IFTES system. The data set spanned the wavelength range of 501 nm to 690 nm and contained 20 images. As shown in Fig. 7(b), the left and right beads are visible in short wavelength channels yet are not visible in long wavelength channels; but, with the middle one the opposite is the case. The emission spectra of these two kinds of beads were reconstructed as shown in Fig. 7(c). The difference in the Fig. 7(b)–7(c) indicates the left and right microspheres are yellow beads, and the middle one is a purple bead. Figure 7(d) is a false-color composite image of the mixture of yellow beads and a purple bead.
Fig. 7 Spectral imaging from a mixture of 10-14 μm dual-color fluorescent beads. (a) Interferogram frame acquired by IFTES. (b) Spectral images at wavelength from 501 nm to 690 nm. (c) Spectra associated with yellow (green line) and purple (red line) beads measured by IFTES, respectively. (d) False-color composite image of the mixture. Green and red represent yellow and purple beads, respectively. Scale bars are 10 μm.
A spectral imaging experiment of dual-color cerebral cortex neural cells in vivo was performed to demonstrate the cerebral spectral imaging ability of IFTES. In this experiment, Thy1-eYFP-H transgenic mouse [37] labeled with Sulforhodamine 101 (SR 101) by a multicellular bolus loading technique [38,39], was adopted as the animal model. In the barrel cortex of this animal model, the V layer neurons expressed YFP reporter protein, while SR 101 only marked astrocytes [40]. Simultaneously detecting interactions between neurons and astrocytes is critical for gaining insight into various cellular processes and mechanisms of neurologic disorders in the central nervous system [41]. About 20 minutes after the dye injection, the fiber bundle was inserted into the barrel cortex (down to 300–400 μm), as shown in Fig. 8(a) . The small diameter of the fiber bundle provided a minimally invasive way to observe neural cells in vivo. A full FOV image is shown in Fig. 8(b) in which axons in FOV are clearly visible, as marked by red arrows. A ROI with an image resolution of 256×256 pixels was chosen, as enclosed by a red rectangle in Fig. 8(b). An OPD scanning range of 120 μm was achieved, and then the corresponding spectral resolution was 1.3 nm at 500 nm, 1.8 nm at 600 nm, and 2.5 nm at 700 nm. The non-uniformity of the spectral resolution here is better than the result in Ref [18]. A set of interferogram images of ROI were captured within 5 s. Figure 8(c) is a false-color composite image of ROI according to the reconstructed x/y/λ data set. There are two kinds of signals in the ROI, and Fig. 8(d) plots the corresponding spectra of these two kinds of signals. One has an intensity peak at 530 nm, while the intensity peak of another locates at 608 nm. From spectral imaging results in Figs. 8(c) and 8(d), it can show that the objects in the top of the ROI are neurons expressing YFP, while the signal of the middle part of the ROI is from cortical astrocytes. This technique has the potential to extend intravital spectral imaging to the deep brain, which is impossible to observe under traditional microscope in vivo.
Fig. 8 Spectral imaging from YFP-transgenic mouse labeled with SR101: (a) Photograph of YFP-transgenic mouse with inserting fiber bundle into cerebral cortex. (b) Fluorescence image of mouse brain in full FOV acquired by IFTES. (c) False-color composite image of ROI enclosed by the red rectangle in (b). Green and red represent neuron and astrocyte, respectively. (d) Emission spectra associated with neuron (green line) and astrocyte (red line) measured by IFTES, respectively. Scale bars in (b) and (c) are 50 μm and 10 μm, respectively.
4. Conclusion
We have presented an endoscopic spectral imaging system based on a flexible fiber bundle catheter and a home-built imaging Fourier transform spectroscope. The system retains several advantageous features of the Fourier transform spectral measurement method including high, variable, and unlimited spectral resolution; wide spectral range; high throughput; and sensitive, simultaneous measurement of all of the frequencies and internal calibration. The flexible spectral resolution and ROI of spectral imaging according to various biological applications were demonstrated in a series of experiments. The results indicate that the instrument is a potential prototype to perform non- or minimally invasive subsurface in vivo and in situ imaging and multispectral imaging at the cellular level.
Acknowledgments
The authors thank Xiuli Liu, Meng Wang, Zhihong Zhang, Liang Wang, and Haiming Luo for the preparation of samples. Additionally, the authors acknowledge Beijing Hamamtsu for a free trial of the ORCA-Flash 2.8s CMOS camera and relevant technical support. This work was funded by the National Major Scientific Research Program of China (No. 2011CB910401), the National Key Technology R & D Program (No. 2011BAI12B06), the Science Fund for Creative Research Group of China (No. 61121004), the National Natural Science Foundation of China (Nos. 61205197 and 61178077), the Wuhan Youth Science and Technology Program (No. 201271031424), the Fundamental Research Funds for the Central Universities (HUST: No. 2012QN109), and the Scientific Research Foundation for the Returned Overseas Chinese Scholars, State Education Ministry. The authors also thank the Analytical and Testing Center (Huazhong University of Science and Technology) for spectral measurements of the fluorescein sodium solution.
References and links
1. B. A. Wilt, L. D. Burns, E. T. W. Ho, K. K. Ghosh, E. A. Mukamel, and M. J. Schnitzer, “Advances in light microscopy for neuroscience,” Annu. Rev. Neurosci. 32(1), 435–506 (2009). [CrossRef] [PubMed]
2. R. K. Jain, L. L. Munn, and D. Fukumura, “Dissecting tumor pathophysiology using intravital microscopy,” Nat. Rev. Cancer 2(4), 266–276 (2002). [CrossRef] [PubMed]
3. C. M. Lee, C. J. Engelbrecht, T. D. Soper, F. Helmchen, and E. J. Seibel, “Scanning fiber endoscopy with highly flexible, 1 mm catheter scopes for wide-field, full-color imaging,” J. Biophotonics 3(5-6), 385–407 (2010). [CrossRef] [PubMed]
4. B. A. Flusberg, E. D. Cocker, W. Piyawattanametha, J. C. Jung, E. L. M. Cheung, and M. J. Schnitzer, “Fiber-optic fluorescence imaging,” Nat. Methods 2(12), 941–950 (2005). [CrossRef] [PubMed]
5. L. Giniūnas, R. Juskaitis, and S. V. Shatalin, “Endoscope with optical sectioning capability,” Appl. Opt. 32(16), 2888–2890 (1993). [CrossRef] [PubMed]
6. L. Fu, A. Jain, H. K. Xie, C. Cranfield, and M. Gu, “Nonlinear optical endoscopy based on a double-clad photonic crystal fiber and a MEMS mirror,” Opt. Express 14(3), 1027–1032 (2006). [CrossRef] [PubMed]
7. A. F. Gmitro and D. Aziz, “Confocal microscopy through a fiber-optic imaging bundle,” Opt. Lett. 18(8), 565–567 (1993). [CrossRef] [PubMed]
8. T. J. Muldoon, M. C. Pierce, D. L. Nida, M. D. Williams, A. Gillenwater, and R. Richards-Kortum, “Subcellular-resolution molecular imaging within living tissue by fiber microendoscopy,” Opt. Express 15(25), 16413–16423 (2007). [CrossRef] [PubMed]
9. N. Bozinovic, C. Ventalon, T. Ford, and J. Mertz, “Fluorescence endomicroscopy with structured illumination,” Opt. Express 16(11), 8016–8025 (2008). [CrossRef] [PubMed]
10. S. Santos, K. K. Chu, D. Lim, N. Bozinovic, T. N. Ford, C. Hourtoule, A. C. Bartoo, S. K. Singh, and J. Mertz, “Optically sectioned fluorescence endomicroscopy with hybrid-illumination imaging through a flexible fiber bundle,” J. Biomed. Opt. 14(3), 030502–030503 (2009). [CrossRef] [PubMed]
11. T. N. Ford, D. Lim, and J. Mertz, “Fast optically sectioned fluorescence HiLo endomicroscopy,” J. Biomed. Opt. 17(2), 021105–021107 (2012). [CrossRef] [PubMed]
12. C. J. Engelbrecht, F. Voigt, and F. Helmchen, “Miniaturized selective plane illumination microscopy for high-contrast in vivo fluorescence imaging,” Opt. Lett. 35(9), 1413–1415 (2010). [CrossRef] [PubMed]
13. J. Yin, Y. Shao, J. Qu, H. Lin, and H. Niu, “A fluorescence sectioning endoscopy using dynamic speckle illumination,” in Endoscopic Microscopy III, (SPIE, 2008), 68510N.
14. G. O. Fruhwirth, S. Ameer-Beg, R. Cook, T. Watson, T. Ng, and F. Festy, “Fluorescence lifetime endoscopy using TCSPC for the measurement of FRET in live cells,” Opt. Express 18(11), 11148–11158 (2010). [CrossRef] [PubMed]
15. Y. Sun, J. Phipps, D. S. Elson, H. Stoy, S. Tinling, J. Meier, B. Poirier, F. S. Chuang, D. G. Farwell, and L. Marcu, “Fluorescence lifetime imaging microscopy: in vivo application to diagnosis of oral carcinoma,” Opt. Lett. 34(13), 2081–2083 (2009). [CrossRef] [PubMed]
16. D. S. Elson, J. A. Jo, and L. Marcu, “Miniaturized side-viewing imaging probe for fluorescence lifetime imaging (FLIM): validation with fluorescence dyes, tissue structural proteins and tissue specimens,” New J. Phys. 9(5), 127 (2007). [CrossRef] [PubMed]
17. J. Siegel, D. S. Elson, S. E. D. Webb, K. C. B. Lee, A. Vlandas, G. L. Gambaruto, S. Lévêque-Fort, M. J. Lever, P. J. Tadrous, G. W. H. Stamp, A. L. Wallace, A. Sandison, T. F. Watson, F. Alvarez, and P. M. W. French, “Studying biological tissue with fluorescence lifetime imaging: microscopy, endoscopy, and complex decay profiles,” Appl. Opt. 42(16), 2995–3004 (2003). [CrossRef] [PubMed]
18. H. Makhlouf, A. F. Gmitro, A. A. Tanbakuchi, J. A. Udovich, and A. R. Rouse, “Multispectral confocal microendoscope for in vivo and in situ imaging,” J. Biomed. Opt. 13(4), 044016–044019 (2008). [CrossRef] [PubMed]
19. F. Jean, G. Bourg-Heckly, and B. Viellerobe, “Fibered confocal spectroscopy and multicolor imaging system for in vivo fluorescence analysis,” Opt. Express 15(7), 4008–4017 (2007). [CrossRef] [PubMed]
20. A. R. Rouse, “Multi-spectral confocal microendoscope for in-vivo imaging,” Ph.D Thesis, University of Arizona (2004).
21. A. R. Rouse and A. F. Gmitro, “Multispectral imaging with a confocal microendoscope,” Opt. Lett. 25(23), 1708–1710 (2000). [CrossRef] [PubMed]
22. J. R. Lakowicz, Principles of Fluorescence Spectroscopy (Springer, 2006).
23. T. Zimmermann, J. Rietdorf, and R. Pepperkok, “Spectral imaging and its applications in live cell microscopy,” FEBS Lett. 546(1), 87–92 (2003). [CrossRef] [PubMed]
24. Y. Garini, I. T. Young, and G. McNamara, “Spectral imaging: principles and applications,” Cytometry A 69A(8), 735–747 (2006). [CrossRef]
25. L. L. Nuffer, P. A. Medvick, H. P. Foote, and J. C. Solinsky, “Multispectral/hyperspectral image enhancement for biological cell analysis,” Cytometry A 69(8), 897–903 (2006). [CrossRef] [PubMed]
26. L. A. Gao, R. T. Kester, N. Hagen, and T. S. Tkaczyk, “Snapshot image mapping spectrometer (IMS) with high sampling density for hyperspectral microscopy,” Opt. Express 18(14), 14330–14344 (2010). [CrossRef] [PubMed]
27. R. T. Kester, N. Bedard, L. Gao, and T. S. Tkaczyk, “Real-time snapshot hyperspectral imaging endoscope,” J. Biomed. Opt. 16(5), 056005 (2011). [CrossRef] [PubMed]
28. S. Wartewig, IR and Raman Spectroscopy (Wiley, 2003).
29. J. T. Motz, D. Yelin, B. J. Vakoc, B. E. Bouma, and G. J. Tearney, “Spectral- and frequency-encoded fluorescence imaging,” Opt. Lett. 30(20), 2760–2762 (2005). [CrossRef] [PubMed]
30. L. Peng, J. T. Motz, R. W. Redmond, B. E. Bouma, and G. J. Tearney, “Fourier transform emission lifetime spectrometer,” Opt. Lett. 32(4), 421–423 (2007). [CrossRef] [PubMed]
31. J. Kauppinen and J. Partanen, Fourier Transforms in Spectroscopy (Wiley, 2001).
32. R. K. Y. Chan, P. K. Lim, X. Z. Wang, and M. H. Chan, “Fourier transform ultraviolet-visible spectrometer based on a beam-folding technique,” Opt. Lett. 31(7), 903–905 (2006). [CrossRef] [PubMed]
33. V. Dubaj, A. Mazzolini, A. Wood, and M. Harris, “Optic fiber bundle contact imaging probe employing a laser scanning confocal microscope,” J. Microsc. 207(Pt 2), 108–117 (2002). [CrossRef] [PubMed]
34. N. Hagen, R. T. Kester, L. Gao, and T. S. Tkaczyk, “Snapshot advantage: a review of the light collection improvement for parallel high-dimensional measurement systems,” Opt. Eng. 51(11), 111702 (2012). [CrossRef] [PubMed]
35. G. D. De Palma, “Confocal laser endomicroscopy in the “in vivo” histological diagnosis of the gastrointestinal tract,” World J. Gastroenterol. 15(46), 5770–5775 (2009). [CrossRef] [PubMed]
36. D. I. Gheonea, A. Saftoiu, T. Ciurea, C. Popescu, C. V. Georgescu, and A. Malos, “Confocal laser endomicroscopy of the colon,” J. Gastrointestin. Liver Dis. 19(2), 207–211 (2010). [PubMed]
37. C. Porrero, P. Rubio-Garrido, C. Avendaño, and F. Clascá, “Mapping of fluorescent protein-expressing neurons and axon pathways in adult and developing Thy1-eYFP-H transgenic mice,” Brain Res. 1345, 59–72 (2010). [CrossRef] [PubMed]
38. M. A. Busche, G. Eichhoff, H. Adelsberger, D. Abramowski, K.-H. Wiederhold, C. Haass, M. Staufenbiel, A. Konnerth, and O. Garaschuk, “Clusters of hyperactive neurons near amyloid plaques in a mouse model of Alzheimer’s disease,” Science 321(5896), 1686–1689 (2008). [CrossRef] [PubMed]
39. C. Stosiek, O. Garaschuk, K. Holthoff, and A. Konnerth, “In vivo two-photon calcium imaging of neuronal networks,” Proc. Natl. Acad. Sci. U.S.A. 100(12), 7319–7324 (2003). [CrossRef] [PubMed]
40. A. Nimmerjahn, F. Kirchhoff, J. N. D. Kerr, and F. Helmchen, “Sulforhodamine 101 as a specific marker of astroglia in the neocortex in vivo,” Nat. Methods 1(1), 31–37 (2004). [CrossRef] [PubMed]
41. E. E. Benarroch, “Neuron-astrocyte interactions: partnership for normal function and disease in the central nervous system,” Mayo Clin. Proc. 80(10), 1326–1338 (2005). [CrossRef] [PubMed]
