WO2010022079A1

WO2010022079A1 - System and methods for diagnosis of epithelial lesions

Info

Publication number: WO2010022079A1
Application number: PCT/US2009/054196
Authority: WO
Inventors: James W. Tunnell; Narasimhan Rajaram
Original assignee: Board Of Regents, The University Of Texas System
Priority date: 2008-08-18
Filing date: 2009-08-18
Publication date: 2010-02-25

Abstract

Systems comprising an optical fiber switch connected to a light source and an optical fiber probe, the optical fiber probe comprising a first optical fiber connected to the optical fiber switch and a second optical fiber connected to a spectrophotometer. Methods for determining one or more tissue parameters comprising: emitting light from a first optical fiber into a tissue; collecting the light reemitted from the tissue with a second optical fiber; generating a spectra of the light reemitted from the tissue with a spectrophotometer; and utilizing a look-up table based algorithm to determine one or more tissue parameters, wherein the lookup-table based algorithm comprises the steps of: generating a look-up table by measuring the functional form of a reflectance measured by the spectrophotometer using one or more calibration standards with known optical properties; and implementing an iterative fitting routine based on the lookup- table.

Description

SYSTEMS AND METHODS FOR DIAGNOSIS OF EPITHELIAL LESIONS

CROSS-REFERENCE TO RELATED APPLICATIONS

The application claims the benefit of U.S. Provisional Patent Application Serial No. 61/089,736, filed August 18, 2008, the entire disclosure of which is hereby incorporated by reference.

BACKGROUND

The present disclosure, according to certain embodiments, generally relates to diagnostic systems and methods. In particular, the present disclosure provides, in certain embodiments, systems and methods for diagnosis of epithelial malignancies. Skin cancer, including both nonmelanoma and melanoma, is the most common malignancy worldwide. There are more than a million cases and greater than 10,000 deaths in the U.S. alone each year associated with skin cancer. As with other epithelial malignancies, early detection and subsequent treatment is paramount to improving prognosis. Early detection of melanoma can improve mortality rates, while the early detection of nonmelanoma can improve associated morbidity and cost. For both of these cutaneous malignancies, noninvasive detection strategies will improve mortality, morbidity, and associated costs.

The current early detection of skin cancers relies on a critical macroscopic visual analysis of the changes in the cutaneous lesions. Suspected malignancies are excised and analyzed using standard histopathology for diagnosis and treatment decisions. This early detection strategy has several limitations. First, diagnostic accuracy for the current clinical examination is inherently qualitative and depends largely on the experience of the physician. It has been shown that general practitioners often have a much lower diagnostic accuracy than expert dermatologist.[l] In addition, access to dermatologists can be limited by geography, financial barriers, and a shortage of supply. Second, the majority of cutaneous melanoma arise in atypical nevi which can easily go unnoticed because they appear as standard moles. In addition, for patients with familial and/or dysplastic nevus syndrome (>100 nevi), it is impossible to excise all suspected dysplastic nevi. Finally, although the sensitivity for the detection of melanoma has been improving (70-90%), the specificity is still quite low, resulting in a large number of unnecessary biopsies which increases costs and morbidity of the procedure. Therefore, a non-invasive method to inspect these lesions would be of great clinical importance.

Currently, when nonmelanoma skin cancers are removed, the surgeon is required to take an excess margin of skin around the lesion to account for nonclinically relevant spread of the tumor. This excess margin can result in a larger scar and greater cosmetic and functional deformity. Noninvasive techniques for limiting the size of these surgical excisions would potentially spare patients from requiring expensive grafting and reconstruction procedures.

Many current strategies for analyzing diffuse reflectance rely on the solution to the diffusion approximation of the radiative transport equation 24, or a modified form 6. However, the diffusion approximation is not valid at source-detector separations less than approximately one mean free path (l/(μS' + μa)) and in tissues with low albedo (μS7(μS' + μa)). In addition, many inverse solutions employing the diffusion approximation are computationally intensive. Several probe based systems for sampling shallow tissue depths have recently been proposed 25, 26. In addition, Mirabal et al. have shown that short source detector separations provide a higher diagnostic power compared to longer source-detector separations 27. Unfortunately, the diffusion approximation based inverse models are not accurate in many of these regimes. To overcome this limitation, several recent models based on Monte Carlo 28 or higher order approximations to radiative transport 29 have been proposed.

DRAWINGS

Some specific example embodiments of the disclosure may be understood by referring, in part, to the following description and the accompanying drawings.

Figure 1 shows the spectral diagnosis system used in Examples 1 and 2. L refers to plano-convex lenses; Ml - mirror; BS - beam splitter. Figure 2 shows the distal end of the fiber probe used in the system shown in Figure 1.

Scale bar is 3 mm.

Figure 3 shows (a) the white light spectrum from the xenon flash lamp reflecting off of a 20% reflectance standard; and (b) excitation pulses from the nitrogen laser at 337 nm and the dye laser at 445 nm. Figure 4 shows (a) the diffuse reflectance spectrum from a tissue phantom (μ_s'(λo) = 1.35 mm^"1 and [Hb] = 2 mg/ml) and the LUT-fit; (b) and (c) the optical parameters extracted from the experimental LUT (μ_s'((λo) = 0.75 mm^"1 (o), 1.35 mm^"1 (D) and 1.75 mm^"1 (Δ) with 1 μm beads. • indicates μ_s'((λo) = 1 mm^"1 with 2 μm beads).

Figure 5 shows data recorded from tissue phantoms with Stilbene 3 as fluorophore under test. The tissue phantoms contained a fixed scatterer concentration (μs' = 1 mm-1) and different concentrations of hemoglobin 0-1.45 mg/ml.

Figure 6 shows data recorded from tissue phantoms with FAD as fluorophore. Figure 6(a) shows measured intrinsic fluorescence spectrum of FAD and fluorescence spectra recorded from tissue phantoms with a fixed scatterer concentration (μ_s' = 1 mm^"') and different concentrations of hemoglobin 0-1.45 mg/ml. Figure 6(b) shows diffuse reflectance spectra from each of the tissue phantoms. Figure 6(c) shows recovery of intrinsic fluorescence spectrum of FAD from a turbid phantom containing 0.82 mg/ml of hemoglobin and comparison with the actual intrinsic fluorescence of the fluorophore. Figure 6(d) shows a comparison of extracted intrinsic fluorescence spectra for different tissue phantoms.

Figure 7 shows fluorophore concentrations of (a) Stilbene 3 and (b) FAD extracted using least-squares regression and comparison to actual values. The solid line indicates perfect agreement.

Figure 8 shows (a) spectrally resolved diffuse reflectance, [R(λ )] ( μ_s'(λ) = 1.69 - 3.56 mm^"1 and μ_a'(λ) = 0 - 5.33 mm^"1) from the calibration set; (b) diffuse reflectance as a sparse matrix mapped to optical property space, [R(μ_s'(λ), μ_a(λ))]; and (c) the resulting lookup table,

Figure 9 shows a schematic diagram of the LUT inverse model. Fit parameters are μ_s'(λo), B, [Hb] and α. Figure 10 shows (a) the diffuse reflectance spectrum (μ_s'(λo) = 2.49 mm^"1 and [Hb] = 2 mg/ml) and the LUT-fit from a tissue phantom (validation set); (b) and (c) scatter plot of the known versus measured values of μ_s'(λ) (b) and μ_a (λ) (c) for all tissue phantoms. The solid line indicates perfect agreement.

Figure 11 shows physical parameters extracted from the LUT inverse model (μ_s'(λo) = 0.91 mm^"1 (D), 1.83 mm^"1 (o) and 2.75 mm^"1 (0) ). The solid line indicates perfect agreement.

While the present disclosure is susceptible to various modifications and alternative forms, specific example embodiments have been shown in the figures and are described in more detail below. It should be understood, however, that the description of specific example embodiments is not intended to limit the invention to the particular forms disclosed, but on the contrary, this disclosure is to cover all modifications and equivalents as illustrated, in part, by the appended claims.

DESCRIPTION

The present disclosure, according to certain embodiments, generally relates to diagnostic systems and methods. In particular, the present disclosure provides, in certain embodiments, systems and methods for diagnosis of epithelial malignancies.

In certain embodiments, the present disclosure provides a system comprising an optical fiber switch connected to a light source and an optical fiber probe, the optical fiber probe comprising a first optical fiber connected to the optical fiber switch and a second optical fiber connected to a spectrophotometer. In certain embodiments, the present disclosure provides a system comprising an optical fiber switch connected to a light source and an optical fiber probe, the optical fiber probe comprising a first optical fiber connected to the optical fiber switch and a second optical fiber connected to a spectrophotometer; and a software interface connected to the spectrophotometer, wherein the software interface is capable of displaying a tissue parameter derived from a spectra generated by the spectrophotometer.

In certain embodiments, the present disclosure provides a method for assessing a tissue comprising: providing an optical fiber switch connected to a light source and an optical fiber probe, the optical fiber probe comprising a first optical fiber connected to the optical fiber switch and a second optical fiber connected to a spectrophotometer; providing a tissue disposed adjacent to the optical fiber probe; allowing light emitted from the first optical fiber into the tissue; collecting the light reemitted from the tissue with the second optical fiber.

In certain embodiments, the present disclosure provides a method for assessing a tissue comprising: providing an optical fiber switch connected to a light source and an optical fiber probe, the optical fiber probe comprising a first optical fiber connected to the optical fiber switch and a second optical fiber connected to a spectrophotometer; providing a software interface connected to the spectrophotometer, wherein the software interface is capable of displaying a tissue parameter derived from a spectra generated by the spectrophotometer; providing a tissue disposed adjacent to the optical fiber probe; allowing light emitted from the first optical fiber into the tissue; collecting the light reemitted from the tissue with the second optical fiber; generating a spectra of the light reemitted from the tissue with a spectrophotometer; and utilizing a look-up table based algorithm to determine one or more tissue parameters.

In certain embodiments, the present disclosure provides a method for determining one or more tissue parameters comprising: emitting light from a first optical fiber into a tissue; collecting the light reemitted from the tissue with a second optical fiber; generating a spectra of the light reemitted from the tissue with a spectrophotometer; and utilizing a look-up table based algorithm to determine one or more tissue parameters, wherein the lookup-table based algorithm comprises the steps of: generating a look-up table by measuring the functional form of a reflectance measured by the spectrophotometer using one or more calibration standards with known optical properties; and implementing an iterative fitting routine based on the lookup-table.

In certain embodiments, the systems of the present invention comprise an optical fiber probe, a spectrophotometer, an optical fiber switch, and a software interface. In certain embodiments, the optical fiber probe may comprise a plurality of optical fibers. In certain embodiments, the optical fiber probe may comprise seven optical fibers. These seven optical fibers may be spatially arranged in any suitable manner. One such arrangement is the "six around one" arrangement, in which six of the seven optical fibers are disposed around the outer diameter of the seventh optical fiber. In certain embodiments, one or more of the optical fibers may emit light, and the remaining fibers may collect the light emitted by one or more fibers.

In certain embodiments, the systems of the present invention further comprise a tissue. The tissue may be any tissue suitable for being analyzed with an optical fiber probe. In certain embodiments, the tissue may be an epithelial tissue, such as, but not limited to, skin, cervical, esophageal, breast, colon, or oral tissue. In certain embodiments, one or more fibers in the optical fiber probe may emit light into the tissue, and the remaining fibers may collect the light reemitted from the tissue.

The light emitted by the optical fiber probe may be one or more a variety of light types. Generally, the light may be any light type suitable for use in analyzing the optical properties of a tissue. Such light types may include, but are not limited to, laser light and white light. Specific examples of light that may be emitted by the optical fiber probe are laser light with a wavelength of about 337 nm, laser light with a wavelength of about 450 nm, and white light emitted from a xenon flashlamp. In certain embodiments, a single type of light may be emitted from a specified optical fiber within the optical fiber probe; thus, multiple fibers may be used to emit multiple light types. In certain embodiments, multiple light types may be emitted by a single optical fiber, hi certain embodiments, the "center fiber" in the "six around one" fiber arrangement (i.e. the one fiber around which the other six fibers are disposed) may be the optical fiber probe which emits the one or more light types. In certain embodiments, the remaining optical fibers (such as, but not limited to, the six fibers in the "six around one" fiber arrangement) may collect the light emitted from the single optical fiber.

Generally, light collected by the optical fibers is recorded by the spectrophotometer. Accordingly, any spectrophotometer capable of being operably connect to the optical fiber probes used in the present invention and recording light spectra for the light types used in the systems of the present invention may be used in the systems of the present invention. The optical fiber switch may be any optical fiber switch suitable for use with the optical fiber probe and spectrophotometer. The choice of a suitable optical fiber switch may depend upon, among other things, the type, source and/or number of sources of light to be emitted by the one or more optical fiber probes, the spectrophotometer chosen, and the tissue type. An example of an optical fiber switch which may be useful in certain embodiments of the systems and methods of the present invention is a FSM- 13 3x1 fiber optic switch, commercially available from Piezosystems Jena, Germany.

In certain embodiments, the software interface may display the light spectra generated by the optical fiber probe. In certain embodiments, such an interface may provide graphical plots for collected light from each light type used. In certain embodiments, the software interface may also a graphical plot of raw sample and calibration spectra taken from one or more calibration standards. In certain embodiments, the spectra generated by the spectrophotometer may be calibrated by the software interface up to or beyond a specified signal to noise ratio. In certain embodiments, such a signal to noise ratio may be about 17 dB. In certain embodiments, this calibration may be performed in a relatively short amount of time. In certain embodiments, the calibration may be performed in approximately one second or less. In certain embodiments, the software interface may determine and display a number of tissue parameters, including, but not limited to, tissue redox ratio, oxygen saturation, scattering parameter, blood concentration, melanin concentration, and collagen content. In certain embodiments, the systems of the present invention comprise an optical fiber probe comprising a plurality of optical fibers and a fiber tip. The plurality of optical fibers may be arranged and may function as described elsewhere in the present disclosure. The fiber tip, in certain embodiments, may be a detachable article which does not substantially interfere with the emission or collection of light by the optical fiber probe and which provides a sterile point of contact between the optical fiber probe and a subject. In certain embodiments, the fiber tip may be made of a suitable polymeric material, including, but not limited to, polystyrene. In certain embodiments, the fiber tip may be secured to the optical fiber probe via a locking mechanism. Such a locking mechanism, in certain embodiments, may comprise a flexible component on the fiber probe with a male element and a rigid component on the fiber tip with a female element. In such embodiments, the male and female elements may join to secure the fiber tip to the optical fiber probe.

In certain embodiments, the spectra generated by the spectrophotometer may be analyzed by a look-up table (LUT) based algorithm. In certain embodiments, the LUT based algorithm is a LUT-based inverse model that is valid for fiber-based probe geometries with close source- detector separations and tissues with low albedos. In certain embodiments, the LUT inverse model may comprise (1) generating a LUT by measuring the functional form of the reflectance using calibration standards with known optical properties and (2) implementing an iterative fitting routine based on the LUT. In certain embodiments, a nonlinear optimization fitting routine may be used to fit the reflectance spectra. Such a routine may comprise (1) constraining the reduced scatteπng coefficient to the form μs'(λ) = μs'(λθ) (λ/λO)-B where λO = 630 ran, (2) calculating an absorption coefficient using the absorption crosssections σπb and o^~ _Hbθ2 as μ_a(λ) = [Hb]*(ασHbθ2 + (1-α) O_Hb) + X, where α is the oxygen saturation of the tissue, Hb is the total hemoglobin concentration of the tissue, and X is adsorption coefficient of a chromophore (e g., melanin, beta-carotene, a dye (e.g , indocyanine green)). In certain embodiments, it may be assumed that the absorption in the visible range to be due to oxy- and deoxy-hemoglobin. In certain embodiments, depending on the type of tissue sampled and the wavelength range of interest, the expression for μa(λ) can be modified to include the absorption cross-sections of other absorbing chromophores In certain embodiments, the look-up algorithm may be used to determine the tissue parameters displayed by the software interface of the systems of the present invention. For example, in certain embodiments, laser excitation at 337 ran generates fluorescence from the metabolic coenzyme NADH and collagen, while laser excitation at 400 nm generates fluorescence from FAD Also, in certain embodiments, white light, such as light from xenon flashlamps, may be used to collect elastic scatteπng spectra. Both NADH and FAD are associated with tissue metabolism and can be used to determine the tissue redox ratio In certain embodiments, elastic scatteπng spectra can be fit to a diffusion theory model to extract the blood oxygen saturation, blood concentration, melanin concentration, and tissue scatteπng parameters. In certain embodiments, fluorescence spectroscopy may be used to extract biochemical properties Fluorescence photons are scattered and absorbed duπng their path to the tissue surface where they are collected via the optical fiber probe Therefore, the spectral features of the collected fluorescence can be significantly distorted, making the extraction of biochemical composition of the tissue from the measured signal difficult. This is a particular problem m the presence of strong tissue absorbers such as hemoglobin Intπnsic fluorescence spectroscopy (IFS) is a technique that extracts the fluorescence of the molecules unaffected by the absorption and scatteπng events from the bulk fluorescence Because diffuse reflectance undergoes similar absorption and scatteπng events as fluorescence, information contained within the reflectance spectra can be used to extract the intπnsic fluorescence from the collected fluorescence spectra Accordingly, in certain embodiments, a LUT-based inverse model may be used to measure the tissue optical properties and correct the acquired fluorescence

Therefore, the present invention is well adapted to attain the ends and advantages mentioned as well as those that are inherent therein While numerous changes may be made by those skilled in the art, such changes are encompassed within the spirit of this invention as illustrated, in part, by the appended claims.

EXAMPLES

Example 1 System Description

A representation of the spectral diagnosis system used in this example is shown in Figure 1. Three light sources were used: 1) a pulsed xenon flashlamp (L7684, Hamamatsu Photonics, Bridgewater, NJ) to collect white light reflectance; 2) a pulsed nitrogen laser (NL-100, Stanford Research Systems, Mountain View, CA) at 337 nm and 3) a nitrogen-pumped dye laser at 450 nm. Coumarin 450 (Exciton Inc., Dayton, OH) was used as the gain medium for the dye laser. A long pass filter (340 nm; Asahi Spectra, Torrance, CA) was placed in the optical path of the xenon flash lamp to minimize exposure to UV light. The specifications for each light source are listed in Table 1. The energy/pulse noted for each light source was the energy measured at the distal end of the output fiber.

Table 1

The white light and laser pulses are coupled into optical fibers and guided into a 3x1 fiber optic switch (FSM-13, Piezosystems Jena, Germany). The switch controls the excitation sequence and is triggered by TTL signals. The switch's output fiber is mated with the input fiber of the fiber optic probe. The distal end of the 2 m long bifurcated fiber optic probe (FiberTech Optica, Ontario, Canada) consists of 7 optical fibers arranged in a 6-around-l configuration (Figure 2b). The central fiber (NA = 0.22, 200 μm core diameter) is used for delivering the light to the skin while the six surrounding fibers (NA = 0.22, 200 μm core diameter) collect the remitted light. Because detection of early cancer lesions which originate in the epithelium is of interest in this example, a source-detector separation of 300 μm was chosen. This distance allowed sampling the skin superficially. The six collection fibers were arranged in a linear configuration and aligned parallel to the entrance slit of the spectrograph (SP-150, Princeton Instruments, Trenton, NJ). The spectrograph contained a 150g/mm grating blazed at 500 nm which disperses the collected light onto a cooled CCD (Photometries, Tucson, AZ). The CCD was cooled to a temperature of -30⁰C to minimize dark current. The CCD was gated (50 μs), to acquire data only during an excitation pulse. This enabled leaving the room lights on, making the system more clinically compatible. White light reflectance was recorded in the spectral range of 350- 700 nm. Binning was performed in a specific region along the slit axis corresponding to probe illumination and in groups of 3 pixels along the wavelength axis to increase the signal to noise ratio. An average over three acquisitions was taken for each light source to further improve the SNR. Instrument control and spectral calibration was automated using a personal computer running LABVIEW (National Instruments, Austin, TX) and MATLAB (Mathworks, Natick, MA). A timer/counter board (NI 2121, National Instruments) was used that interfaces with the computer to generate trigger pulses for the light sources, fiber optic switch and the spectrograph. The light sources are triggered in a sequence and the corresponding channels on the fiber optic switch are opened to allow for excitation.

The entire system was built on an optical breadboard and transferred to a portable rack (2'x2'x3'; PTRK-1426, Middle Atlantic Products) (Figure 2a). The rack houses the power supplies for all the system components as well as a personal computer. Adequate arrangements were made to ensure the proper ventilation of the rack and transfer of heat generated by the power supplies to the outside. System Calibration

A number of routines were employed to correct for system response and variations in source excitation intensity. The wavelength scale of the CCD is calibrated with a standard mercury-argon (HgAr) lamp. A background spectrum was recorded with the light sources turned off and subtracted from every reflectance spectrum. This eliminated the effects of CCD dark current and ambient light. In addition, white light reflectance from a 20% reflectance standard (Labsphere, North Sutton, NH) was recorded before the start of any measurement cycle. The background- corrected reflectance spectrum was then divided by the standard reflectance to obtain a relative diffuse reflectance measurement. To correct for variations in the white light intensity of the xenon flash lamp, the reflectance from a standard solution of polystyrene microspheres in water (0.12%; Polysciences, Warrington, PA) was measured. Diffuse reflectance spectra measured from phantoms are normalized with respect to standard solution of microspheres. The fluorescence spectra are corrected for the spectral response of the system using a NIST traceable tungsten calibration standard (LS-I-CAL, Ocean Optics, Dunedin, FL). The fluorescence from a Rhodamine B solution in water (0.01 g/1) was measured to calibrate both the nitrogen and the dye laser for variations in intensity.

Physical Tissue Model To test the ability of the system to collect reflectance and fluorescence spectra and recover various optical parameters, measurements were made on a matrix of tissue-simulating phantoms. The phantoms were laced in cylindrical vials (20 mm diameter by 15 mm depth) that were large enough to simulate semi-infinite media. The optical fiber probe was placed in contact with the phantom surface to collect diffuse reflectance and fluorescence spectra. Diffuse reflectance phantoms

Tissue phantoms were fabricated using polystyrene microspheres (diameter = 1 or 2 μm; Polysciences, Warrington, PA) and ferrous stabilized hemoglobin powder (Sigma, St. Louis, MO) dissolved in water to simulate scattering and absorption, respectively. Mie theory was used to calculate the reduced scattering coefficients (μ_s') of the tissue phantoms, and we measured the absorption coefficient (μ_a) of the stock Hb solution using a spectrophotometer (DU 720, Beckman Coulter, Fullerton, CA). An array of tissue phantoms were created with varying scattering (μ_s'(λo = 630 ran) = 0.25-3 mm^"1) and absorption parameters ([Hb] = 0 - 5mg/ml).

Intrinsic fluorescence phantoms Commercially available fluorophores were used to prepare tissue phantoms for measuring fluorescence. FAD (Sigma, St. Louis, MO) was available commercially and hence was used as the fluorophore under test with the dye laser (445 nm excitation). Stilbene 3 (Exciton, Dayton, OH) was chosen to simulate NADH fluorescence due to the similar position of its peak emission wavelength. The tissue phantoms were fabricated in three stages. Non- scattering solutions of the two fluorophores were first prepared to measure the intrinsic fluorescence. The fluorophore concentrations were selected so that the solutions were optically dilute. 0.64 μM of Stilbene 3 and 42.1 μM of FAD were used in the experiments. To examine the combined effect of scattering and absorption, hemoglobin was added in concentrations ranging from 0-1.45 mg/ml. A fixed concentration of beads (μ_s'(λ0) = 1 mm^"') was added to simulate scattering. LUT Inverse Model

Many current strategies for analyzing diffuse reflectance rely on the solution to the diffusion approximation of the radiative transport equation (1), or a modified form (2). However, the diffusion approximation is not valid at source-detector separations less than approximately one mean free path (l/(μ_s' + μ_a)) and in tissues with low albedo (μ_s7(μ_s' + μ_a)). In addition, many inverse solutions employing the diffusion approximation are computationally intensive. Several probe based systems for sampling shallow tissue depths have recently been proposed (3,4). In addition, Mirabal et al. have shown that short source detector separations provide a higher diagnostic power compared to longer source-detector separations (5). Unfortunately, the diffusion approximation based inverse models are not accurate in many of these regimes. To overcome this limitation, several recent models based on Monte Carlo (6) or higher order approximations to radiative transport (7) have been proposed. In order to fit the diffuse reflectance spectra generated in this example, a LUT-based inverse model was developed that is valid for fiber-based probe geometries with close source- detector separations and tissues with low albedos. The development of the LUT inverse model followed two steps. First, a LUT is generated by measuring the functional form of the reflectance using calibration standards with known optical properties. Second, an iterative fitting routine is implemented based on the LUT. This LUT was subsequently used in an iterative formulation of the inverse model. We used tissue phantoms to validate the accuracy of the LUT inverse model and consequently the system. A subset of tissue phantoms was used to create the LUT. No test samples used to validate the system had the same optical properties of the phantoms used to create the LUT. A nonlinear optimization fitting routine was implemented to fit the reflectance spectra.

The reduced scattering coefficient was constrained to the form μ_s'(λ) = μ_b'(λo).(λ/λo)^~B where λo = 630 nm. The absorption in the visible range was assumed to be due to oxy- and deoxy- hemoglobin. The absorption coefficient was calculated using the absorption crosssections (σπ_b and σ_Hbo₂) of these chromophores as μ_a(λ) = [Hb]⁺(UOHbO₂ + (1-α) o^"m_>), where α is the oxygen saturation and [Hb] is the total hemoglobin concentration. Depending on the type of tissue sampled and the wavelength range of interest, the expression for μ_a(λ) can be modified to include the absorption cross-sections of other absorbing chromophores. System Performance — Signal to noise ratio A typical white light spectrum reflecting off of a 20% reflectance standard is shown in Figure 3. The signal to noise ratio of a typical reflectance spectrum (Figure 4a) is ~34 dB. Also shown are laser excitation pulses from the nitrogen laser at 337 nm and the dye laser at 445 nm. The fluorescence spectra from both lasers also show excellent signal to noise (-40 dB).

System Performance — Diffuse reflectance The diffuse reflectance spectrum from a tissue phantom and its corresponding fit is plotted in Figure 4. The general shape of the curve is derived from the negative power law nature of μ_s' and depressions due to the characteristic absorption peaks of hemoglobin at 420,

542 and 577 nm. These depressions are quite prominent and become more pronounced as the hemoglobin concentration increases.

Figure 4 illustrates the results of fitting the measured spectra for the same phantom to the model. Each panel represents a particular optical property recovered. The LUT inverse model estimated the reduced scattering and absorption coefficients over a wide range (μ_s'(λ) = 0.56 - 3.36 mm^"1 and μ_a(λ) = 0 - 5.42 mm^'1) with mean RMS errors of 9.8% and 11.8% respectively (data not shown). Figure 4 illustrates extracted physical parameters for each test tissue phantom, demonstrating good agreement between the expected and the measured values of μ_s'(λo) and [Hb]. In all phantoms, the oxygen saturation was held a constant and did not vary by more than 2%. The average error in estimating the physical parameters was less than 10%. System Performance — Intrinsic fluorescence Figure 5 shows the results of fluorescence measurements on tissue-simulating phantoms with both scattering and absorption (hereafter referred to as turbid phantom). Fluorescence spectra from tissue phantoms with varying concentrations of hemoglobin were plotted and compared to the intrinsic fluorescence spectrum of Stilbene 3 (Figure 5a). The addition of hemoglobin introduces a distortion in the fluorescence spectrum around 420 nm. This can be attributed to absorption by hemoglobin in the Soret band. This is evident in the diffuse reflectance spectra for the same set of tissue phantoms which show a large depression in the Soret absorption band (Figure 5b). The intrinsic fluorescence was extracted using a photon migration model described by Mueller et al. (Figure 5c) (8). There is excellent agreement between the extracted and measured intrinsic fluorescence spectra (Figure 5d). We were able to recover the intensity and the line shape of the intrinsic fluorescence spectrum with an RMS error of less than 10%.

Figure 6 demonstrates the results of fluorescence measurements on turbid phantoms with FAD as the fluorophore. Although the fluorescence spectrum of FAD is distorted by hemoglobin absorption, the effect is not as prominent as that for 337 nm excitation (Figure 5a). This is probably because FAD emits in the Q-bands of hemoglobin where the absorption is not as high as in the Soret band (Figure 5b). This is seen in the corresponding diffuse reflectance spectra of these phantoms (Figure 5b) which show a relatively small depression in the Q-bands due to hemoglobin absorption. However, there was still a significant correction introduced in the measured fluorescence spectrum of FAD (Figure 5c). As with Stilbene 3, there was good agreement between the measured and extracted intrinsic fluorescence (RMS error less than 10%) (Figure 4d).

The concentrations of the two fluorophores were extracted with a least-squares regression technique using the following equation (9)

where β refers to intrinsic fluorescence from a turbid phantom and β_dii, the intrinsic fluorescence from an optically dilute solution of the fluorophore. The concentration of the phantom C was the free parameter that was extracted. Figure 7 shows a comparison of the actual and recovered values of the fluorophore concentrations. The average errors in estimating the concentrations of Stilbene 3 and FAD were 5.87% and 11.1 %, respectively. Example 2

The system we used to collect the diffuse reflectance is described in detail in Example 1. Briefly, a custom built clinical spectrometer system was used to collect steady state, spectrally resolved diffuse reflectance in the wavelength range of 350-700 nm. A pulsed xenon flash lamp (L7684, Hamamatsu Photonics, Bridgewater, NJ) was used as the light source. A six- around-one fiber optic probe configuration (diameter = 200 μm; NA = 0.22) was used where the central fiber illuminated the sample and six surrounding fibers collected the diffusely reflected light. A source-detector separation of 300 μm was employed. The light collected by the fiber optic probe was focused on the entrance slit of a spectrograph (SP-150, Princeton Instruments, Trenton, NJ) that dispersed the light onto a 12-bit CCD (CoolSnap, Photometries, Tucson, AZ).

A LUT was generated by measuring the functional form of the reflectance using tissue phantoms with known optical properties (a calibration set). These phantoms were fabricated using polystyrene microspheres (diameter = 1 μm; Polysciences, Wanϊngton, PA) and India ink (Salis International, Golden, CO) dissolved in water to simulate scattering and absorption, respectively. Mie theory was used to calculate μ_s' of the tissue phantoms, and μ_a of a stock

India ink solution was measured using a spectrophotometer (DU 720, Beckman Coulter,

Fullerton, CA). We created a matrix (4x6) of 24 tissue phantoms with varying scattering

(μ_s'(λ) = 0.22 - 7.1 mm^"1) and absorption parameters (μ_a(λ) = 0 - 5.33 mm^"1). The probe was placed in contact with the surface of the tissue phantoms and spectrally resolved diffuse reflectance spectra from the phantoms were recorded (Figure 8a). The mapping of the spectrally resolved diffuse reflectance (R) on to a unique LUT is shown in Figure 8. The spectral dependence of R results from the wavelength dependent optical properties, μ_s'(λ) and μ_a(λ). Because μ_s'(λ) and μ_a(λ) are known for the tissue phantoms, R can be mapped from wavelength space to the two dimensional optical property space. This mapping creates a sparse matrix (Figure 8b) for R. This sparse matrix was then interpolated to a grid of uniformly spaced data points of μ_s' and μ_a to obtain a LUT for diffuse reflectance (Figure 8c). The limits of the LUT correspond to the range of μ_s' and μ_a over which diffuse reflectance spectra were recorded.

To validate the performance of the LUT, a separate matrix of tissue phantoms (a validation set) was created. Because the LUT is generated with experimental data, a validation set with the same absorber as the calibration set might influence the inverse model while fitting the diffuse reflectance spectra. Therefore, tissue phantoms were used with different chromophores to create the LUT and validate the accuracy of the inverse model. Tissue phantoms for the validation set were fabricated using polystyrene microspheres and hemoglobin (Sigma, St. Louis, MO) as the absorber. A matrix (3x6) of 18 different tissue phantoms was then created for the validation set by varying the values of μ_s'(λo) and [Hb].

To fit our diffuse reflectance spectra and extract the optical properties, a nonlinear optimization fitting routine was implemented as described in Figure 9. We constrained the reduced scattering coefficient to the form μ_s'(λ) = μ_s'(λo).(λ/λo)^~B where λo = 630 nm. The absorption in the visible range was assumed to be due to oxy- and deoxy-hemoglobin. The absorption coefficient was calculated using the absorption cross-sections (σπ_b and σHbcc) of these chromophores as μ_a (λ) = [Hb] (ασκbθ2 + (l-α)σ_Hb), where α is the oxygen saturation and [Hb] is the total hemoglobin concentration. Depending on the type of tissue sampled and the wavelength range of interest, the expression for μ_a(λ) can be modified to include the absorption cross-sections of other absorbing chromophores.

The diffuse reflectance spectrum and corresponding fit from a sample validation phantom is shown in Figure 10a demonstrating excellent agreement between the model and the experimental data. Scatter plots of the extracted versus expected values of μ_s'(λ) (Figure 10b) and μ_a (λ) (Figure 10c) demonstrate a high degree of accuracy in extracting optical properties. The LUT inverse model estimated the reduced scattering and absorption coefficients over a wide range (μ_s'(λ) = 0.72 - 4.91 mm^"1 and μ_a (λ) = 0 - 2.29 mm^"1) with mean RMS percent errors of 5.9% and 11.6%, respectively. These scatter plots show the extracted μ_s'(λ) and μ_a (λ) for the entire validation set. Figure 11 illustrates extracted physical parameters for each tissue phantom of the validation set, demonstrating good agreement between the expected and the measured values of μ_s'(λo) and [Hb]. In all phantoms, the oxygen saturation was held a constant in the experiment and the fit did not vary by more than 2%. The average errors in estimating μ_s'(λo) and [Hb] over the entire validation set were 4.9% and 9.6%, respectively.

The performance of the LUT-based model was compared to a diffusion approximation (DA)-based model described by Farrell et al. (Table 1). At a source-detector separation of 300 μm, the LUT model represents a significant improvement in the recovery of the physical parameters over the DA model. The LUT model improved the accuracy in recovering scattering at 630 nm (μ_s'(λo) and hemoglobin concentration ([Hb]) by factors of 2.5 and 5.5, respectively. In addition, a fundamental limitation of analytical solutions such as the diffusion approximation is that scattering should dominate absorption by at least a factor of 10 (μ_s' > 10μ_a). This implies that the albedo must be greater than 0.9 for the diffusion approximation to be valid. This condition is usually violated in the visible region of light, especially the Soret absorption band (400-430 nm) of hemoglobin. However, at the lowest value of albedo seen in the validation set (0.35), the LUT model was able to estimate the μ_s'(λo) and [Hb] with errors of 6.2% and 8% respectively.

This analysis indicates that the errors for the LUT-based model are close to 10% for determining both scattering and absorption. A certain component of the error in the inverse model could arise from the uncertainty in optical properties of the calibration set used to generate the LUT. For example, the scattering coefficient of calibration set phantoms was calculated using Mie theory and the bead concentration reported by the manufacturer. The experimental error in this concentration is on the order of a few percent and will propagate through our final inverse solution. Other sources of error include knowledge of bead size, presence of electronic noise is the collected reflectance, knowledge dye extinction coefficient, and fabrication of phantoms. While these errors primarily contribute to the overall accuracy of the solution, they should not affect the precision, which is important when comparing tissue pathologies for disease diagnosis. Minimizing these errors could lead to a significant improvement in the accuracy of the LUT-based model.

Notwithstanding that the numerical ranges and parameters setting forth the broad scope of the invention are approximations, the numerical values set forth in the specific examples are reported as precisely as possible. Any numerical value, however, inherently contain certain errors necessarily resulting from the standard deviation found in their respective testing measurements. Therefore, the present invention is well adapted to attain the ends and advantages mentioned as well as those that are inherent therein. While numerous changes may be made by those skilled in the art, such changes are encompassed within the spirit of this invention as illustrated, in part, by the appended claims. References:

The following references are all incorporated by reference to the extent they provide information available to one of ordinary skill in the art regarding the implementation of the technical teachings of the invention.

1. Farrell, T.J., M.S. Patterson, and B. Wilson, A Diffusion-Theory Model Of Spatially Resolved, Steady-State Diffuse Reflectance For The Noninvasive Determination Of

Tissue Optical-Properties Invivo. Med. Phys. 19(4), 879-888 (1992).

2. Zonios, G., L.T. Perelman, V.M. Backman, R. Manoharan, M. Fitzmaurice, J. Van Dam, and M.S. FeId, Diffuse reflectance spectroscopy of human adenomatous colon polyps in vivo. Appl. Opt. 38(31), 6628-6637 (1999). 3. Wang, A., V. Nammalavar, and R. Drezek, Experimental evaluation of angularly variable fiber geometry for targeting depth-resolved reflectance from layered epithelial tissue phantoms. J. Biomed. Opt. 12(4), 044011 (2007).

4. Nieman, L., A. Myakov, J. Aaron, and K. Sokolov, Optical Sectioning Using a Fiber Probe with an Angled Illumination-Collection Geometry: Evaluation in Engineered Tissue Phantoms. Appl. Opt. 43(6), 1308- 1319 (2004).

5. Mirabal, Y.N., S.K. Chang, E.N. Atkinson, A. Malpica, M. Follen, and R. Richards- Kortum, Reflectance spectroscopy for in vivo detection of cervical precancer. J. Biomed. Opt. 7, 587 (2002).

6. Palmer, G.M. and N. Ramanujam, Monte Carlo-based inverse model for calculating tissue optical properties. Part I: Theory and validation on synthetic phantoms. Appl. Opt.

45(5), 1062-1071 (2006).

7. Hull, E.L. and T.H. Foster, Steady-state reflectance spectroscopy in the P-3 approximation. Journal of the Optical Society of America A 18(3), 584-599 (2001).

8. Muller, M.G., I. Georgakoudi, Q. G. Zhang, J. Wu, and M.S. FeId, Intrinsic fluorescence spectroscopy in turbid media: disentangling effects of scattering and absorption. Appl.

Opt. 40(25), 4633-4646 (2001).

9. Gardner, CM., S. L. Jacques, and A.J. Welch, Fluorescence spectroscopy of tissue: Recovery of intrinsic fluorescence from measured fluorescence. Appl. Opt. 35(10), 1780- 1792 (1996).

Claims

CLAIMSWhat is claimed is:

1. A system comprising an optical fiber switch connected to a light source and an optical fiber probe, the optical fiber probe comprising a first optical fiber connected to the optical fiber switch and a second optical fiber connected to a spectrophotometer.

2. The system of claim 1 further comprising a tissue disposed adjacent to the optical fiber probe.

3. The system of claim 1 wherein the second optical fiber is a plurality of optical fibers and the first optical fiber is disposed centrally to the second optical fiber.

4. The system of claim 1 wherein the second optical fiber is a plurality of optical fibers disposed around the outer diameter of the first optical fiber.

5. The system of claim 1 wherein the second optical fiber is a plurality of six optical fibers disposed around the outer diameter of the first optical fiber.

6. The system of claim 1 wherein the light source is a laser light, a white light, or both.

7. A system comprising: an optical fiber switch connected to a light source and an optical fiber probe, the optical fiber probe comprising a first optical fiber connected to the optical fiber switch and a second optical fiber connected to a spectrophotometer; and a software interface connected to the spectrophotometer, wherein the software interface is capable of displaying a tissue parameter derived from a spectra generated by the spectrophotometer.

8. The system of claim 7 wherein the software interface comprises a lookup-table based algorithm.

9. The system of claim 7 wherein the software interface comprises a lookup-table based algorithm, the lookup-table based algorithm comprising: generating a look-up table by measuring the functional form of a reflectance measured by the spectrophotometer using one or more calibration standards with known optical properties; and implementing an iterative fitting routine based on the lookup-table.

10. The system of claim 8 wherein the lookup-table based algorithm further comprises using a nonlinear optimization fitting routine to fit the spectra.

11. The system of claim 9 wherein the nonlinear optimization fitting routine comprises: constraining a reduced scattering coefficient to the form μ_s'(λ) = μ_s'(λo).(λ/λo)^~B where λo = 630 nm; and calculating an absorption coefficient using the absorption crosssections σ_Hb and o^"Hbθ2 as μ_a(λ) = [Hb]*(ασπb02 + (1-α) OHb) + X, where α is the oxygen saturation of the tissue, Hb is the total hemoglobin concentration of the tissue, and X is adsorption coefficient of a chromophore.

12. A method for assessing a tissue comprising: providing an optical fiber switch connected to a light source and an optical fiber probe, the optical fiber probe comprising a first optical fiber connected to the optical fiber switch and a second optical fiber connected to a spectrophotometer; providing a tissue disposed adjacent to the optical fiber probe; allowing light emitted from the first optical fiber into the tissue; and collecting the light reemitted from the tissue with the second optical fiber.

13. A method for assessing a tissue comprising: providing an optical fiber switch connected to a light source and an optical fiber probe, the optical fiber probe comprising a first optical fiber connected to the optical fiber switch and a second optical fiber connected to a spectrophotometer; providing a software interface connected to the spectrophotometer, wherein the software interface is capable of displaying a tissue parameter derived from a spectra generated by the spectrophotometer; providing a tissue disposed adjacent to the optical fiber probe; allowing light emitted from the first optical fiber into the tissue; collecting the light reemitted from the tissue with the second optical fiber; generating a spectra of the light reemitted from the tissue with a spectrophotometer; and utilizing a look-up table based algorithm to determine one or more tissue parameters.

14. The method of claim 13 wherein the light source is a laser light, a white light, or both.

15. The method of claim 13 wherein allowing light emitted from the first optical fiber into the tissue comprises: emitting laser light having a wavelength of about 337 nm; emitting laser light having a wavelength of about 450 nm; and emitting white light.

16. The method of claim 13 wherein the tissue comprises an epithelial lesion.

17. The method of claim 13 wherein utilizing a lookup-table based algorithm comprises: generating a look-up table by measuring the functional form of a reflectance measured by the spectrophotometer using one or more calibration standards with known optical properties; and implementing an iterative fitting routine based on the lookup-table.

18. The method of claim 13 wherein the lookup-table based algorithm further comprises the step of using a nonlinear optimization fitting routine to fit the spectra.

19. The method of claim 18 wherein the nonlinear optimization fitting routine comprises the steps of: constraining a reduced scattering coefficient to the form μ_s'(λ) = μ_s'(λo).(λ/λo)^"B where λo = 630 nm; and calculating an absorption coefficient using the absorption crosssections σ_Hb and σ_Hb02 as μ_a(λ) = [Hb]*(ασ_Hbθ2 + (1-α) σHb) + X, where α is the oxygen saturation of the tissue, Hb is the total hemoglobin concentration of the tissue, and X is adsorption coefficient of a chromophore

20. A method for determining one or more tissue parameters comprising: emitting light from a first optical fiber into a tissue; collecting the light reemitted from the tissue with a second optical fiber; generating a spectra of the light reemitted from the tissue with a spectrophotometer; and utilizing a look-up table based algorithm to determine one or more tissue parameters, wherein the lookup-table based algorithm comprises the steps of: generating a look-up table by measuring the functional form of a reflectance measured by the spectrophotometer using one or more calibration standards with known optical properties; and implementing an iterative fitting routine based on the lookup-table.