OSA's Digital Library

Virtual Journal for Biomedical Optics

Virtual Journal for Biomedical Optics

| EXPLORING THE INTERFACE OF LIGHT AND BIOMEDICINE

  • Editor: Gregory W. Faris
  • Vol. 3, Iss. 1 — Jan. 29, 2008
« Show journal navigation

Investigation of signal dependence on tissue thickness in near infrared spectral imaging

Bevin Lin, Victor Chernomordik, Amir Gandjbakhche, Dennis Matthews, and Stavros Demos  »View Author Affiliations


Optics Express, Vol. 15, Issue 25, pp. 16581-16595 (2007)
http://dx.doi.org/10.1364/OE.15.016581


View Full Text Article

Acrobat PDF (1476 KB)





Browse Journals / Lookup Meetings

Browse by Journal and Year


   


Lookup Conference Papers

Close Browse Journals / Lookup Meetings

Article Tools

Share
Citations

Abstract

The signal intensity in near infrared autofluorescence and polarization sensitive light scattering imaging is explored as a function of tissue thickness using homogeneous porcine cardiac tissue samples as a model system. Eight images are recorded from each tissue sample including two autofluorescence images obtained under 408 nm and 633 nm excitation and six light scattering images acquired with alternating linear polarization orientations (parallel or perpendicular) under 700 nm, 850 nm, and 1000 nm linearly polarized illumination. The mean image intensity of each sample for each imaging method is plotted as a function of tissue thickness. The experimental results indicate a strong dependence of the detected signal on tissue thickness up to approximately 2 mm. Furthermore, the intensity of the spectral ratio images also exhibit thickness-dependent changes up to about 3 mm. The behavior of the light scattering experimental data was reproduced using a mathematical model based on a modified version of the random walk theory of photon migration.

© 2007 Optical Society of America

1. Introduction

An array of optical methods has been explored in recent years as a means to characterize tissue parameters for the detection and evaluation of disease phenotypes. One of the main tools for tissue characterization is based on its autofluorescence (AF) under ultra violet and visible (UV-VIS) excitation wavelengths arising from endogenous tissue fluorophores such as flavins (FAD), collagen, elastin, and porphyrins [1

1. R. R. Alfano, D. Tata, J. Cordero, P. Tomashefsky, F. Longo, and M. Alfano, “Laser-induced fluorescence spectroscopy from native cancerous and normal tissue,” IEEE J. Quantum Electron 20, 1507–1511 (1984). [CrossRef]

4

4. S. G. Demos, A. J. Vogel, and A. H. Gandjbakhche, “Advances in Optical Spectroscopy and Imaging of Breast Lesions,” J. of Mammary Gland Biol. 11, 165–181 (2006). [CrossRef]

]. Light scattering methods have also been extensively investigated for tissue characterization utilizing the changes of the optical properties of the tissue components due to onset and progression of disease [5

5. J. R. Mourant, I. J. Bigio, J. Boyer, R. L. Conn, T. Johnson, and T. Shimada “Spectroscopic diagnosis of bladder cancer with elastic light scattering,” Lasers Surg. Med. 17, 350–357 (1995). [CrossRef] [PubMed]

9

9. J. R. Mourant, J. P. Freyer, A. H. Hielscher, A. A. Eick, D. Shen, and T. M. Johnson, “Mechanisms of light scattering from biological cells relevant to noninvasive optical-tissue diagnostics,” Appl. Opt. 37, 3586–3593 (1998) [CrossRef]

].

Research that probes imaging at depths beyond the penetration of the UV-VIS has explored the use of near infrared (NIR) wavelengths. Cross-polarized light scattering has been investigated in surface and subsurface imaging [10

10. S. G. Demos and R. R. Alfano, “Optical polarization imaging,” Appl. Opt. 36, 150–155 (1997). [CrossRef] [PubMed]

, 11

11. S. G. Demos, H. B. Radousky, and R. R. Alfano, “Deep subsurface imaging in tissues using spectral and polarization filtering,” Opt Express 7, 23–28 (2000). [CrossRef] [PubMed]

]. Multispectral imaging of skin lesions has been used to obtain functional information [12

12. G. N. Stamatas, M. Southall, and N. Kollias, “In vivo monitoring of cutaneous edema using spectral imaging in the visible and near infrared,” J. Invest. Dermatol. 126, 1753–1760 (2006). [CrossRef] [PubMed]

, 13

13. M. Hassan, R. F. Little, A. Vogel, K. Aleman, K. Wyvill, R. Yarchoan, and A. H. Gandjbakhche, “Quantitative assessment of tumor vasculature and response to therapy in Kaposi’s sarcoma using functional noninvasive imaging,” Technol. Cancer Res. T 3, 451–457 (2004).

]. Such information can be extracted using Monte Carlo simulations to model light interaction in tissues and analyze the concentration of various chromophores. These techniques can be valuable for the diagnosis and staging of disease as well as for monitoring progression during treatment. It is therefore imperative to better understand how to analyze and accurately quantify the autofluorescence or light scattering signal arising from various tissue layers with varying depths especially in the NIR spectral region, which provides information of deeper tissue and is less affected by the properties of blood and its derivatives. Understanding how the measured signal changes with the thickness of the tissue is a key parameter that, to the best of our knowledge, is still not well understood.

In this work, NIR autofluorescence and polarization sensitive light scattering spectral imaging are explored to investigate the dependence of the measured signal on tissue thickness. The experiments were performed using ex vivo myocardial tissue, a system that provides sufficient tissue uniformity to isolate thickness-dependent changes. The experimental results are modeled based on the random walk theory of photon migration.

2. Experimental materials and methods

Porcine cardiac tissue was the tissue model in this work. Specifically, intact porcine hearts obtained from a local supermarket were sliced by hand to produce samples with homogeneous surface areas and thickness avoiding irregularities such as large vascular structures. This maximized sample uniformity allowing the tissue thickness to be the experimental variable. Each sliced sample was placed individually in phosphate buffered saline (PBS) solution and ice for preservation during experimentation. For image acquisition, each sample was removed from the PBS solution and pat dry on a kim wipe to remove excess blood and PBS solution. The samples were positioned between glass slides at the sample holder and the thickness was measured with a micrometer. The thickness of the glass slides was subtracted so that only the thickness of each tissue sample was considered during analysis. The tissue sample was measured after slicing, rather than attempting to slice a specific thickness. Sample sizes varied with thickness but were each large enough such that the intensity of the central region was free of edge effects. For example, a 0.28 mm thick sample was approximately 15 mm x 12 mm, and a 7.36 mm thick sample was approximately 44 mm x 38 mm in their lateral dimensions, shown in Fig. 1.

Fig. 1. The 700 nm perpendicular polarization raw image of a 7.36 mm in thickness tissue sample. Note the tissue variability visible from the surface. The blue line in the center of the tissue outlines the region of interest used for data analysis.

A detailed description of the imaging system is provided elsewhere [14

14. C. A. Lieber, S. Urayama, N. Rahim, R. Tu, R. Saroufeem, B. Reubner, and S. G. Demos, “Multimodal near infrared spectral imaging as an exploratory tool for dysplastic esophageal lesion identification,” Opt Express 14, 2211–2219 (2006). [CrossRef] [PubMed]

] and shown in Fig. 2. In brief, AF images were acquired using excitation at 408 nm and 633 nm provided by compact laser sources. Using appropriate optical elements, the beams were expanded so that the illumination intensity within the imaged area was nearly uniform. The images were acquired using a 690 nm long wavelength pass filter in front of the camera in order to use the same spectral range of the emission in the NIR for image formation under both excitation wavelengths. In addition, polarization-sensitive light scattering images were acquired under white light illumination using three different narrow-band interference filters of 40 nm bandwidth at full width half maximum (FWHM) centered at 700 nm, 850 nm, and 1000 nm. Linear polarization elements were attached in the illumination and imaging paths so that both the parallel and perpendicular light scattering image components could be recorded. The polarizer P’ in front of the lens was removable and could be rotated to enable acquisition of the parallel and perpendicular image components. The exposure times (40 seconds for the AF images and 0.2 seconds for the light scattering images) were experimentally determined to maximize imaging intensity without saturation of the charge coupled device (CCD).

Fig. 2. Schematic layout of the experimental system used for the acquisition of the multispectral autofluorescence and light scattering images

A white piece of paper was placed in the sample holder and imaged in the same manner as the tissue samples for the fluorescence and light scattering measurements. This recorded the spatial distribution of the illumination intensity (illumination profile image) for each imaging mode. After the background (dark images with no illumination) was subtracted, each tissue image was divided by the corresponding illumination profile image to eliminate artifacts from non-uniformity in the illumination intensity. The average image intensity was then estimated across a user specified central region of the tissue image. Lastly, the intensity from the small piece of masking tape positioned in the upper right corner of the sample holder (see Fig. 1) was used to monitor the average illumination/excitation intensity in each imaging method during the course of the experiment. These values yielded a multiplier for each image, which varied very little between images of the same imaging method since the illumination was very stable except for the 408 nm laser excitation (which varied within 10 %).

Image processing was performed using Matlab, (MathWorks, Natick, MA). Overall, the mean intensity of the 8 different images from 135 tissue samples was calculated. The average intensity in the 2 AF images of each tissue sample was normalized as follows: First, the average intensity of samples greater than 3 mm thick was calculated. Above this thickness, the intensity appears to exhibit no dependence on tissue thickness. Then, the intensities of all samples were divided by this average value for normalization. Thus, intensity for the 408 nm and 633 nm images are plotted in arbitrary units as a function of thickness.

The average intensity in the 6 light scattering images of each tissue sample was normalized against a WS-1 Diffuse Reflectance Standard, a diffusing material in a circular anodized aluminum encasement with reflective outer diameter of 32 mm obtained from Ocean Optics, (Dunedin, FL). The diffusing material produces more than 98% reflection across the surface of the diffusing material from 250 nm–1500 nm. The normalization procedure was the following: First, the WS-1 Diffuse Reflectance Standard was imaged and normalized in the same manner as the 6 light scattering imaging process under 0.2 second exposure. Then, the sum of the intensities of the parallel and perpendicular image components for each of the three imaging wavelengths (700 nm, 850 nm, and 1000 nm) was calculated in order to obtain the total scattering intensity of the reference standard sample at each imaging wavelength. Subsequently, the average image intensity from each tissue sample using each of the 6 light scattering imaging methods was divided by the total scattering intensity of the reference sample at the corresponding imaging wavelength. This allowed us to obtain an accurate estimate of the amount of backscattered light from each sample as a function of its thickness for each imaging method.

The light scattering data are combined with random walk theory of photon migration to present a mathematical method towards developing a tool to predict the signal dependence on thickness in various experimental configurations and future applications. In order to test the ability of this modeling approach to predict the photon propagation characteristics in our experiment, we performed additional measurements to quantify the amount of light that penetrates the tissue as a function of tissue thickness. A 600 µm diameter fiber was used to transmit illumination from a Mikropack DH-2000 halogen light source (Ocean Optics, Dunedin, FL) to the sample after passing through a 500 nm long wavelength pass filter. The output tip of the fiber was secured to a digital micrometer (not shown) positioned directly above the sample in order to achieve contact of the tip of the fiber with the tissue and simultaneously measure tissue thickness. A collection fiber was secured to a magnetic base and aligned directly beneath the illumination fiber and just below a glass slide where the tissue sample is placed. The collection fiber delivered the transmitted signal to a Triax Series 320 spectrometer (Jobin Yvon Inc., Edison, NJ) before detection at a liquid nitrogen cooled CCD (Princeton Instruments, Inc., Trenton, NJ). This experimental set-up is shown in Fig. 3.

Fig. 3. Schematic of transmission spectroscopy experimental system.

The acquisition time was adjusted in each measurement to achieve sufficient signal to noise ratio (SNR) for all tissue thickness without saturating the detector. The signal was kept above approximately 1000 counts per spectral pixel (depending on the tissue thickness and exposure time) while the noise was on the order of 15 counts. Four transmission spectra were acquired per sample within an approximate 1 mm–3 mm range of each other within the central region of the sample. These four spectra were binned to produce one average spectrum per tissue sample for data analysis. Acquired spectra were normalized in a similar manner described above - subtracting a dark noise spectrum and dividing against the WS-1 Diffuse Reflectance Standard spectrum. The central region of the tissue was used for each spectrum to avoid tissue surface area imperfections and edge effects.

3. Experimental results

3.1 NIR autofluorescence intensity vs. thickness

The average intensities of the AF images of each sample obtained under 408 nm laser excitation as a function of the tissue thickness are shown in Fig. 4(a). The normalized intensity appears to increase as tissue thickness increases before stabilizing around 2 mm. The corresponding results obtained from the AF images under 633 nm laser excitation are shown in Fig. 4(b). Similar to the results shown in Fig. 4(a), the intensity appears to increase with tissue thickness before stabilizing around 2 mm.

Fig. 4. The average intensities of the autofluorescence images of each sample obtained under a) 408 nm and b) 633 nm laser excitation as a function of the tissue thickness.

3.2. NIR scattering intensity vs. thickness

The average intensities of the backscattered polarized image components after passing through the narrow-band interference filter centered at 700 nm of each sample are shown in Fig. 5(a). The normalized intensity appears to increase for both the parallel and perpendicular components before stabilizing around 2 mm. Similar results were obtained using the narrow-band interference filters centered at 850 nm and 1000 nm shown in Fig. 5(b) and Fig. 5(c) respectively.

From the light scattering data shown in Fig. 5, additional information may be obtained by analyzing the polarization properties of the recorded signal and the variation in intensity as a function of imaging wavelength. As it has been previously discussed [10

10. S. G. Demos and R. R. Alfano, “Optical polarization imaging,” Appl. Opt. 36, 150–155 (1997). [CrossRef] [PubMed]

] the perpendicular signal component represents half the intensity of the depolarized signal component while the difference between the parallel (I) and perpendicular (I) components represents the intensity of the polarization-preserving component. Therefore, subtracting the I from the I signal components allow us to evaluate in our measurements the intensity of the polarization-preserving component. Figure 6 shows the intensity of the polarization-preserving component (I - I) as a function of tissue thickness. The results indicate that, for the 700 nm and 1000 nm data sets, the intensity of the polarization-preserving component remains constant for all tissue thickness with an average value at approximately 0.03. However, the intensity of the polarization-preserving component for the 850 nm data set is somewhat higher and may depend on tissue thickness.

Fig. 5. The average intensities of the backscattered polarized image components after passing through the narrow-band interference filter centered at a) 700 nm, b) 850 nm, and c) 1000 nm.
Fig. 6. The measured intensity of the polarization-preserving component (I - I) at a) 700 nm, b) 850 nm, and c) 1000 nm as a function of tissue thickness.

The spread in the experimental data shown in all figures can be attributed to localized variations in the optical properties of the different tissue samples. The spectral information embedded in the light scattering images from each sample can provide additional information that relates to the spectral characteristics of the reflected light as a function of tissue thickness. To best depict this spectral information, we calculated the ratio of intensities of the scattering images of each sample. This approach also normalizes the value of the ratio with respect to the absolute intensity of the reflected light from each tissue sample. Figures 7(a), 7(b), and 7(c) show the ratio of the average intensities of the perpendicular image components at 700 nm over that at 1000 nm, 700 nm over that at 850 nm and, 850 nm over that at 1000 nm, respectively. The corresponding ratios obtained from the parallel image components are shown in Figs. 8(a), 8(b), and 8(c) respectively.

Fig. 7. The ratio of the average intensities of the perpendicular image components at a) 700 nm over 1000 nm, b) 700 nm over 850 nm, and c) 850 nm over 1000 nm.
Fig. 8. The ratio of the average intensities of the parallel image components at a) 700 nm over 1000 nm, b) 700 nm over 850 nm, and c) 850 nm over 1000 nm.

The perpendicular ratiometric results appear to increase with tissue thickness up to approximately 3 mm. The parallel ratiometric results exhibit complex behavior presumably due to the presence of the polarization-preserving component that does not depend on tissue thickness.

3.3. NIR transmission intensity vs. thickness

Typical transmission spectra are shown in Fig. 9(a) for tissue sample thickness of approximately 1.0 mm, 3.0 mm and 5.0 mm. These exemplary transmission spectra were normalized at 920 nm to allow simultaneous display and comparison. The average intensities of the transmitted signal at 600 nm, 650 nm, and 1000 nm of each sample are shown as representative results in Fig. 9(b) plotted on a semi-logarithmic scale. The normalized intensity is seen to decrease exponentially as tissue thickness increases for all wavelengths. These results will be discussed later in more detail.

Fig. 9. Transmission spectra from a) 1.0 mm, 3.0 mm and 5.0 mm thick tissue samples normalized to 920 nm, and b) the average intensity of the transmitted signal at 600 nm, 650 nm, and 1000 nm as a function of tissue thickness.

3.4. Mathematical representation of photon migration

Analysis of the data on cardiac tissue reflectivity has been performed in several steps. First, the difference in I and I of backscattered light was attributed to the contribution of the specular reflection at the interface of the sample. Figure 6 demonstrates an increase of I by approximately 0.03 compared to that of I for the entire range of tissue thickness. Therefore, this change may be attributed to the specular reflectivity contribution [(n-1)/(n+1)]2~0.03 at the air-tissue interface (for ntissue ~1.4). The results shown in Fig. 6 are in good agreement with this assumption at λ=700 nm and 1000 nm. For λ=850 nm, the agreement is worse, but still reasonable within the experimental variation of the data. This assumes that in all cases, multiple scattering in the cardiac tissue samples causes nearly complete depolarization of the detected light except that from specular reflections. Consequently, we assumed the total diffuse reflection of tissue samples to be the sum of the measured parallel and perpendicular polarized light components minus the estimated contribution of specular reflection (I+I- R0) with R0≈0.03. We then explored the ability of the random walk theory model (RW) [16

16. G. H. Weiss, A. H. Gandjbakhche, and J. Masoliver, “Isotropization length for random-walk models of photon migration in turbid media,” J. Mod. Opt. 42, 1567–1574 (1995). [CrossRef]

] to describe our experimental results. We have chosen RW for analysis of our data because of its simplicity It is known that RW gives quantitative results very similar to diffusion model, and is substantiated by experimental data (see for example, paper [17

17. R. Cubeddu, A. Pifferi, P. Taroni, A. Torricelli, and G. Valentini, “Experimental test of theoretical models for time-resolved reflectance,” Med. Phys. 23, 1625 (1996). [CrossRef] [PubMed]

], where theoretical models of time-resolved reflectance are compared, showing that RW is by no means inferior to standard diffusion model for the slab geometry).

The diffuse reflectance within the random walk formula can be written as:

R(μ)=exp(2μ)24μ[1exp(24μ)+2(1cosh(24μ))exp(L24μ)1]
(1)

where μ=μaμs', µa, µs and µs=µs(1-g) are the absorption, scattering, and transport-corrected optical coefficients respectively, g is the anisotropy factor, L=s/√2+1 and, D is the sample thickness. It should be noted that the diffuse reflectance increases with increasing slab thickness, and reaches for large thickness L≫1/√(24µ) an asymptotic value of:

R(μ)=exp(2μ)24μ[1exp(24μ)]
(2)

This asymptotic value depends only on the ratio of absorption and transport-corrected scattering coefficients µ. With decreasing slab thickness, the observed reflectivity is decreasing much faster than is expected in the diffusion-like model, described by Eq. (1) because of the presence of the ballistic and quasiballistic photons that do not contribute to the backscattering signal. Thus, as a first approximation to model the reflected signal from relatively thin slabs, Eq. (2) may be modified to exclude these non-diffusive (non-contributing) photons by introducing a correction factor of (1- exp(-µatD)). This empiric coefficient is used to take into account quasi-ballistic photons transmitted through thin tissue samples that experience only a few scattering events and do not contribute to the detected backscattered light.

The estimates of the optical coefficients, obtained by fitting the experimental data with this simplified model, empirically corrected to account for the quasi-ballistic photons under the assumption of µatas’ are given in Table 1. Comparison between the total diffuse reflection (I+I- 0.03) as estimated from our experimental data and best fit with the model is illustrated Fig. 10 for λ=700 nm, 850 nm, 1000 nm, respectively. Conventional Marquardt-Levenberg algorithm [18

18. W. H. Press, S. A. Teukolsky, W. T. Vetterling, and B. P. Flannery, Numerical Recipes in C: The Art of Scientific Computing (Cambridge University Press, New York, 1992).

] realized in SigmaPlot curve fitter (Systat Software, Inc.), with two fitting parameters was applied for data analysis. Fitting produces a unique set of optical properties because the cross talk between free parameters in the model µ at and asymptotic value of reflectance R(µ) is small. R(µ) is obtained from large thickness (“asymptotic”) data, while µ at is determined mainly by small thickness data. Standard errors of the estimates of µ at and R(µ) prove to be ≤5%. Table 1 summarizes the fitting parameters and presents values of the product of µaµs’ because they can be estimated independently.

Several recent papers have provided estimates of the optical parameters of the cardiac tissues in the visible and near infrared spectral range [19

19. A. H. Gandjbakhche, R. F. Bonner, A. E. Arai, and R. S. Balaban, “Visible-light photon migration through myocardium in vivo,” Am. J. Physiol-Heart C. 277, H698–H704 (1999).

21

21. V. K. Ramshesh and S. B. Knisley, “Spatial localization of cardiac optical mapping with multiphoton excitation,” J. Biomed. Opt. 8, 253–259 (2003). [CrossRef] [PubMed]

]. Different experimental approaches were used to extract these parameters. The optical coefficients of cardiac tissues extracted from modeling our experimental results (Table 1) seem to be larger (by ~1.5–2 X) compared to the published data on this type of tissues which are summarized in Table 2. The reason for this discrepancy is not clear. It may partially be due to different procedures of sample preparation. It is worth noting that the slope of the transport corrected scattering coefficient shown in Fig. 11 is estimated from our model to be ~-0.59, compared to a slope of ~1.18, obtained in [20

20. J. Swartling, S. Palsson, P. Platonov, S. B. Olsson, and S. Andersson-Engels, “Changes in tissue optical properties due to radio-frequency ablation of myocardium,” Med. Biol. Eng. Comput. 41, 403–409 (2003). [CrossRef] [PubMed]

] over a narrower spectral range. This may be indicative of larger Mie equivalent diameters of the scatterers in our samples.

Fig. 10. Comparison between the total diffuse reflection (I+I- 0.03) as estimated from our experimental data and best fit with the model for a) 700 nm, b) 850 nm, and c) 1000 nm.

Table 1. The value of the optical parameters as estimated using our modeling approach

table-icon
View This Table
| View All Tables

Table 2. Optical Coefficients, reported in the literature

table-icon
View This Table
| View All Tables
Fig. 11. The slope of transport corrected scattering in log-log graph.
Fig. 12. Estimate of µaµs’ plotted with coefficient reflectivity measurements.

The transmission measurements shown in Fig. 9(b) were used to verify the values of the optical coefficients of the tissue using our modeling approach. According to the random walk model [16

16. G. H. Weiss, A. H. Gandjbakhche, and J. Masoliver, “Isotropization length for random-walk models of photon migration in turbid media,” J. Mod. Opt. 42, 1567–1574 (1995). [CrossRef]

], the transmitted intensity for thick tissue samples is proportional to

exp(6μaμs'L)=exp(3μaμs'D)
(3)

Thus, the slopes of log intensity linear fit versus thickness can be used to estimate the product of µaµs’. Fig. 12 shows the estimated values of the µaµs’ product obtained using the transmission measurements (shown in Fig. 9(b)) in the spectral range between 600 nm and 1000 nm. For comparison, the estimated values of µaµs’ obtained from the light scattering imaging measurements for λ=700 nm, 850 nm, and 1000 nm, (see Table 1) are also shown. The observed dependence of µaµs’ on the wavelength agrees qualitatively with our reflection data as well as the literature data. In particular, steep increase of this parameter at λ ~600 nm is due to considerable increase of hemoglobin absorption [20

20. J. Swartling, S. Palsson, P. Platonov, S. B. Olsson, and S. Andersson-Engels, “Changes in tissue optical properties due to radio-frequency ablation of myocardium,” Med. Biol. Eng. Comput. 41, 403–409 (2003). [CrossRef] [PubMed]

]. Increased values of µaµs’ for λ>900 nm may be due to lipids and/or additional absorption of water (peak at ~950 nm).

4. Discussion

The results shown in Fig. 4(a) and Fig. 4(b) reveal similar behavior for both excitations, increasing in intensity as thickness increases up to approximately 2 mm. This similarity may be surprising since the photon penetration thickness is very different at these two wavelengths. However, it should be recognized that the emission spectral range (690 nm and longer) for the AF imaging experiments was the same for both, 408 nm and 633 nm excitation wavelengths. Therefore, these results may suggest that this observed behavior arises primarily from the light transport characteristics of the emitted photons. Porphyrins are a likely candidate responsible for this detected NIR emission. Various porphyrins absorb light at 633 nm, exhibit peak absorption near 405 nm, and emit in the NIR part of the spectrum [22

22. S. G. Demos, R. Gandour-Edwards, R. Ramsamooj, and R. D. White, “Near-infrared autofluorescence imaging for detection of cancer,” J. Biomed. Opt. 9, 587–592 (2004). [CrossRef] [PubMed]

]. The emission tail of various other chromophores could also constitute a component of the NIR emission under 408nm excitation. Upon photoexcitation, a large portion of the emitted NIR photons will propagate in the forward direction in a similar fashion as in the case of light scattering under NIR illumination. This will lead to a dependence of the AF intensity as a function of thickness (shown in Fig. 4) that is nearly identical to the light scattering profiles (shown in Fig. 5) since both behaviors are governed by the same mechanism.

The normalized NIR scattering intensities for the three illumination wavelengths shown as a function of thickness in Fig. 5 demonstrate that, although both the parallel and perpendicular detected scattering intensities depend on tissue thickness up to approximately 2 mm, the parallel component is more intense than the perpendicular component. This is not surprising based on earlier work that established the parallel image component carries the polarized photons and half of the depolarized photons, while the perpendicular image component carries only the second half of the depolarized photons [10

10. S. G. Demos and R. R. Alfano, “Optical polarization imaging,” Appl. Opt. 36, 150–155 (1997). [CrossRef] [PubMed]

, 23

23. S. L. Jacques, J. R. Roman, and K. Lee, “Imaging superficial tissues with polarized light,” Laser Surg. Med. 26, 119–129 (2000). [CrossRef]

]. The difference in mean intensity between the parallel and perpendicular components was analyzed by determining the polarization-preserving component (I - I), shown in Fig. 6. This exhibits a small change with thickness for the 850 nm data set but it is approximately equivalent to the specular reflection component for the 700 nm and 1000 nm data sets.

The results shown in Fig. 7 and Fig. 8 demonstrate that the spectral ratio exhibits a dependence on tissue thickness up to approximately 3 mm. This can be attributed to the change in the optical properties as a function of wavelength. Ratiometric analysis of the intensity from the same tissue samples at different illumination wavelengths utilizes the spectral information as a method to normalize for variations in absolute scattering intensity between tissue samples. We postulate that this in turn provides for increased sensitivity to tissue thickness. Only the intensity of the spectral ratio of the perpendicular polarized image components shown in Figs. 7(a), 7(b), and 7(c) exhibit a monotonically increasing profile, which may be used to extrapolate tissue thickness. The corresponding spectral ratio from the parallel polarized image component shown in Figs. 8(a), 8(b), and 8(c) exhibits more complex behavior that may be attributed to the presence of the polarization-preserving component (I - I). This indicates that polarization sensitive imaging to avoid artifacts such as specular reflections can provide benefit in better assessing the relative scattering intensity originating from different tissue depths in various applications such as functional imaging using multispectral imaging.

Our modeling approach based on a modified random walk theory was able to reasonably approximate our experimental data. Spread in the data can be attributed to tissue preparation, storage, and handling, but mainly to inhomogeneity of tissue composition. An additional analysis approach might include the consideration of muscle striation and its effect on light propagation and depolarization. Overall, the signal as measured and analyzed exhibits a dependence on tissue thickness up to approximately 3 mm. The results suggest that such methods may be useful for the evaluation of tissue thickness or lesion thickness. An adaptation of this method may include fiber optic components that may be suitable for incorporation into various types of catheters or endoscopes. Furthermore, these results highlight the importance of the tissue or lesion thickness in the analysis of multispectral NIR images for achieving an accurate quantitative assessment such as for retrieving functional information.

Acknowledgments

This work performed in part under the auspices of the U.S. Department of Energy by Lawrence Livermore National Laboratory under Contract DE-AC52-07NA27344.

This research is supported by funding from the Center for Biophotonics, an NSF Science and Technology Center, managed by the University of California, Davis, under Cooperative Agreement No. PHY 0120999.

References and links

1.

R. R. Alfano, D. Tata, J. Cordero, P. Tomashefsky, F. Longo, and M. Alfano, “Laser-induced fluorescence spectroscopy from native cancerous and normal tissue,” IEEE J. Quantum Electron 20, 1507–1511 (1984). [CrossRef]

2.

R. Richards-Kortum, “Quantitative Optical Spectroscopy for Tissue Diagnosis,” Annu. Rev. Phys. Chem. 47, 555–606 (1996). [CrossRef] [PubMed]

3.

G. A. Wagnieres, W. M. Star, and B. C. Wilson, “In vivo fluorescence spectroscopy and imaging for oncological applications,” Photochem. Photobiol. 68, 603–632 (1998). [PubMed]

4.

S. G. Demos, A. J. Vogel, and A. H. Gandjbakhche, “Advances in Optical Spectroscopy and Imaging of Breast Lesions,” J. of Mammary Gland Biol. 11, 165–181 (2006). [CrossRef]

5.

J. R. Mourant, I. J. Bigio, J. Boyer, R. L. Conn, T. Johnson, and T. Shimada “Spectroscopic diagnosis of bladder cancer with elastic light scattering,” Lasers Surg. Med. 17, 350–357 (1995). [CrossRef] [PubMed]

6.

Y. L. Kim, Y. Liu, R. K. Wali, H. K. Roy, M. J. Goldberg, A. K. Kromin, K. Chen, and V. Backman, “Simultaneous measurement of angular and spectral properties of light scattering for characterization of tissue microarchitecture and its alteration in early precancer,” IEEE J. Sel. Top. Quantum Electron 9, 243–256 (2003). [CrossRef]

7.

R. Drezek, M. Guillaud, T. Collier, I. Boiko, A. Malpica, C. Macaulay, M. Follen, and R. Richards-Kortum, “Light scattering from cervical cells throughout neoplastic progression: influence of nuclear morphology, DNA content, and chromatin texture,” J. Biomed. Opt. 8, 7–16 (2003). [CrossRef] [PubMed]

8.

G. Zonios, L. T. Perelman, V. M. Backman, R. Manoharan, M. Fitzmaurice, J. Van Dam, and M. S. Feld, “Diffuse reflectance spectroscopy of human adenomatous colon polyps in vivo,” Appl. Opt. 38, 6628–6637 (1999) [CrossRef]

9.

J. R. Mourant, J. P. Freyer, A. H. Hielscher, A. A. Eick, D. Shen, and T. M. Johnson, “Mechanisms of light scattering from biological cells relevant to noninvasive optical-tissue diagnostics,” Appl. Opt. 37, 3586–3593 (1998) [CrossRef]

10.

S. G. Demos and R. R. Alfano, “Optical polarization imaging,” Appl. Opt. 36, 150–155 (1997). [CrossRef] [PubMed]

11.

S. G. Demos, H. B. Radousky, and R. R. Alfano, “Deep subsurface imaging in tissues using spectral and polarization filtering,” Opt Express 7, 23–28 (2000). [CrossRef] [PubMed]

12.

G. N. Stamatas, M. Southall, and N. Kollias, “In vivo monitoring of cutaneous edema using spectral imaging in the visible and near infrared,” J. Invest. Dermatol. 126, 1753–1760 (2006). [CrossRef] [PubMed]

13.

M. Hassan, R. F. Little, A. Vogel, K. Aleman, K. Wyvill, R. Yarchoan, and A. H. Gandjbakhche, “Quantitative assessment of tumor vasculature and response to therapy in Kaposi’s sarcoma using functional noninvasive imaging,” Technol. Cancer Res. T 3, 451–457 (2004).

14.

C. A. Lieber, S. Urayama, N. Rahim, R. Tu, R. Saroufeem, B. Reubner, and S. G. Demos, “Multimodal near infrared spectral imaging as an exploratory tool for dysplastic esophageal lesion identification,” Opt Express 14, 2211–2219 (2006). [CrossRef] [PubMed]

15.

S. G. Demos, H. Savage, A. S. Heerdt, S. Schantz, and R. R. Alfano, “Time resolved degree of polarization for human breast tissue,” Opt. Commun. 124.439–442 (1996). [CrossRef]

16.

G. H. Weiss, A. H. Gandjbakhche, and J. Masoliver, “Isotropization length for random-walk models of photon migration in turbid media,” J. Mod. Opt. 42, 1567–1574 (1995). [CrossRef]

17.

R. Cubeddu, A. Pifferi, P. Taroni, A. Torricelli, and G. Valentini, “Experimental test of theoretical models for time-resolved reflectance,” Med. Phys. 23, 1625 (1996). [CrossRef] [PubMed]

18.

W. H. Press, S. A. Teukolsky, W. T. Vetterling, and B. P. Flannery, Numerical Recipes in C: The Art of Scientific Computing (Cambridge University Press, New York, 1992).

19.

A. H. Gandjbakhche, R. F. Bonner, A. E. Arai, and R. S. Balaban, “Visible-light photon migration through myocardium in vivo,” Am. J. Physiol-Heart C. 277, H698–H704 (1999).

20.

J. Swartling, S. Palsson, P. Platonov, S. B. Olsson, and S. Andersson-Engels, “Changes in tissue optical properties due to radio-frequency ablation of myocardium,” Med. Biol. Eng. Comput. 41, 403–409 (2003). [CrossRef] [PubMed]

21.

V. K. Ramshesh and S. B. Knisley, “Spatial localization of cardiac optical mapping with multiphoton excitation,” J. Biomed. Opt. 8, 253–259 (2003). [CrossRef] [PubMed]

22.

S. G. Demos, R. Gandour-Edwards, R. Ramsamooj, and R. D. White, “Near-infrared autofluorescence imaging for detection of cancer,” J. Biomed. Opt. 9, 587–592 (2004). [CrossRef] [PubMed]

23.

S. L. Jacques, J. R. Roman, and K. Lee, “Imaging superficial tissues with polarized light,” Laser Surg. Med. 26, 119–129 (2000). [CrossRef]

OCIS Codes
(000.1430) General : Biology and medicine
(110.3080) Imaging systems : Infrared imaging
(170.1610) Medical optics and biotechnology : Clinical applications
(170.3880) Medical optics and biotechnology : Medical and biological imaging
(170.4580) Medical optics and biotechnology : Optical diagnostics for medicine

ToC Category:
Medical Optics and Biotechnology

History
Original Manuscript: October 5, 2007
Revised Manuscript: November 27, 2007
Manuscript Accepted: November 28, 2007
Published: November 29, 2007

Virtual Issues
Vol. 3, Iss. 1 Virtual Journal for Biomedical Optics

Citation
Bevin Lin, Victor Chernomordik, Amir Gandjbakhche, Dennis Matthews, and Stavros Demos, "Investigation of signal dependence on tissue thickness in near infrared spectral imaging," Opt. Express 15, 16581-16595 (2007)
http://www.opticsinfobase.org/vjbo/abstract.cfm?URI=oe-15-25-16581


Sort:  Author  |  Year  |  Journal  |  Reset  

References

  1. R. R. Alfano, D. Tata, J. Cordero, P. Tomashefsky, F. Longo, M. Alfano, "Laser-induced fluorescence spectroscopy from native cancerous and normal tissue," IEEE J. Quantum Electron 20, 1507-1511 (1984). [CrossRef]
  2. R. Richards-Kortum, "Quantitative Optical Spectroscopy for Tissue Diagnosis," Annu. Rev. Phys. Chem. 47, 555-606 (1996). [CrossRef] [PubMed]
  3. G. A. Wagnieres, W. M. Star, and B. C. Wilson, "In vivo fluorescence spectroscopy and imaging for oncological applications," Photochem. Photobiol. 68, 603-632 (1998). [PubMed]
  4. S. G. Demos, A. J. Vogel, and A. H. Gandjbakhche, "Advances in Optical Spectroscopy and Imaging of Breast Lesions," J. Mammary Gland Biol. 11, 165-181 (2006). [CrossRef]
  5. J. R. Mourant, I. J. Bigio, J. Boyer, R. L. Conn, T. Johnson, and T. Shimada "Spectroscopic diagnosis of bladder cancer with elastic light scattering," Lasers Surg. Med. 17, 350-357 (1995). [CrossRef] [PubMed]
  6. Y. L. Kim, Y. Liu, R. K. Wali, H. K. Roy, M. J. Goldberg, A. K. Kromin, K. Chen, and V. Backman, "Simultaneous measurement of angular and spectral properties of light scattering for characterization of tissue microarchitecture and its alteration in early precancer," IEEE J. Sel. Top. Quantum Electron 9, 243-256 (2003). [CrossRef]
  7. R. Drezek, M. Guillaud, T. Collier, I. Boiko, A. Malpica, C. Macaulay, M. Follen, and R. Richards-Kortum, "Light scattering from cervical cells throughout neoplastic progression: influence of nuclear morphology, DNA content, and chromatin texture," J. Biomed. Opt. 8, 7-16 (2003). [CrossRef] [PubMed]
  8. G. Zonios, L. T. Perelman, V. M. Backman, R. Manoharan, M. Fitzmaurice, J. Van Dam, and M. S. Feld, "Diffuse reflectance spectroscopy of human adenomatous colon polyps in vivo," Appl. Opt. 38, 6628-6637 (1999) [CrossRef]
  9. J. R. Mourant, J. P. Freyer, A. H. Hielscher, A. A. Eick, D. Shen, and T. M. Johnson, "Mechanisms of light scattering from biological cells relevant to noninvasive optical-tissue diagnostics," Appl. Opt. 37, 3586-3593 (1998) [CrossRef]
  10. S. G. Demos and R. R. Alfano, "Optical polarization imaging," Appl. Opt. 36, 150-155 (1997). [CrossRef] [PubMed]
  11. S. G. Demos, H. B. Radousky, and R. R. Alfano, "Deep subsurface imaging in tissues using spectral and polarization filtering," Opt Express 7, 23-28 (2000). [CrossRef] [PubMed]
  12. G. N. Stamatas, M. Southall, and N. Kollias, "In vivo monitoring of cutaneous edema using spectral imaging in the visible and near infrared," J. Invest. Dermatol. 126, 1753-1760 (2006). [CrossRef] [PubMed]
  13. M. Hassan, R. F. Little, A. Vogel, K. Aleman, K. Wyvill, R. Yarchoan, and A. H. Gandjbakhche, "Quantitative assessment of tumor vasculature and response to therapy in Kaposi's sarcoma using functional noninvasive imaging," Technol. Cancer Res. T 3,451-457 (2004).
  14. C. A. Lieber, S. Urayama, N. Rahim, R. Tu, R. Saroufeem, B. Reubner, and S. G. Demos, "Multimodal near infrared spectral imaging as an exploratory tool for dysplastic esophageal lesion identification," Opt Express 14, 2211-2219 (2006). [CrossRef] [PubMed]
  15. S. G. Demos, H. Savage, A. S. Heerdt, S. Schantz, and R. R. Alfano, "Time resolved degree of polarization for human breast tissue," Opt. Commun. 124. 439-442 (1996). [CrossRef]
  16. G. H. Weiss, A. H. Gandjbakhche, and J. Masoliver, "Isotropization length for random-walk models of photon migration in turbid media," J. Mod. Opt. 42, 1567-1574 (1995). [CrossRef]
  17. R. Cubeddu, A. Pifferi, P. Taroni, A. Torricelli, and G. Valentini, "Experimental test of theoretical models for time-resolved reflectance," Med. Phys. 23, 1625 (1996). [CrossRef] [PubMed]
  18. W. H. Press, S. A. Teukolsky, W. T. Vetterling, B. P. Flannery, Numerical Recipes in C: The Art of Scientific Computing (Cambridge University Press, New York, 1992).
  19. A. H. Gandjbakhche, R. F. Bonner, A. E. Arai, R. S. Balaban, "Visible-light photon migration through myocardium in vivo," Am. J. Physiol-Heart C. 277, H698-H704 (1999).
  20. J. Swartling, S. Palsson, P. Platonov, S. B. Olsson, and S. Andersson-Engels, "Changes in tissue optical properties due to radio-frequency ablation of myocardium," Med. Biol. Eng. Comput. 41, 403-409 (2003). [CrossRef] [PubMed]
  21. V. K. Ramshesh and S. B. Knisley, "Spatial localization of cardiac optical mapping with multiphoton excitation," J. Biomed. Opt. 8, 253-259 (2003). [CrossRef] [PubMed]
  22. S. G. Demos, R. Gandour-Edwards, R. Ramsamooj, and R. D. White, "Near-infrared autofluorescence imaging for detection of cancer," J. Biomed. Opt. 9, 587-592 (2004). [CrossRef] [PubMed]
  23. S. L. Jacques, J. R. Roman, and K. Lee, "Imaging superficial tissues with polarized light," Laser Surg. Med. 26, 119-129 (2000). [CrossRef]

Cited By

Alert me when this paper is cited

OSA is able to provide readers links to articles that cite this paper by participating in CrossRef's Cited-By Linking service. CrossRef includes content from more than 3000 publishers and societies. In addition to listing OSA journal articles that cite this paper, citing articles from other participating publishers will also be listed.


« Previous Article  |  Next Article »

OSA is a member of CrossRef.

CrossCheck Deposited