Method and devices for laser induced fluorescence attenuation spectroscopy

Abstract

The Laser Induced fluorescence Attenuation Spectroscopy (LIFAS) method and apparatus preferably include a source adapted to emit radiation that is directed at a sample volume in a sample to produce return light from the sample, such return light including modulated return light resulting from modulation by the sample, a first sensor, displaced by a first distance from the sample volume for monitoring the return light and generating a first signal indicative of the intensity of return light, a second sensor, displaced by a second distance from the sample volume, for monitoring the return light and generating a second signal indicative of the intensity of return light, and a processor associated with the first sensor and the second sensor and adapted to process the first and second signals so as to determine the modulation of the sample. The methods and devices of the inventions are particularly well-suited for determining the wavelength-dependent attenuation of a sample and using the attenuation to restore the intrinsic laser induced fluorescence of the sample. In turn, the attenuation and intrinsic laser induced fluorescence can be used to determined a characteristic of interest, such as the ischemic or hypoxic condition of biological tissue.

Description

BACKGROUND OF THE INVENTION

The present invention is directed to methods and devices for determining a spectroscopic characteristic of a sample utilizing laser induced fluorescence attenuation spectroscopy (“LIFAS”). More particularly, the invention is directed to methods and devices for measurement of the wavelength-dependent attenuation of the sample and subsequent restoration of the intrinsic laser induced fluorescence (“LIF”) for physiological monitoring, biological tissue characterization and biochemical analysis.

Conventionally, samples have been characterized by determining the attenuation and laser induced fluorescence (“LIF”). Once the attenuation and LIF of a sample have been determined, these spectroscopic properties can be utilized to determine a physical or physiological property of the sample. For example, the attenuation of a sample can be used to determine the concentrations of mixture components or turbidity of a fluid. Similarly, the LIF of a sample has been used in fields such as analytical chemistry, environmental monitoring, industrial inspection and medical diagnosis. In the medical field, for example, LIF spectroscopic techniques have been used for tissue characterization, malignant tumor identification, atherosclerotic plaque diagnosis, metabolism evaluation, and the like.

Conventionally, the attenuation or the absorption of a sample is determined by placing the sample between a light source and a detector and measuring any reduction in the intensity of the light as it passes through the sample. In order to obtain a measurement having an acceptable signal-to-noise ratio using these conventional techniques, it is important to transmit the incident light with sufficient intensity. Thus, the thickness of the sample and the wavelength of the incident light are important factors affecting the reliability of the resulting measurement. Moreover, because it is necessary to place the sample between the light source and the detector, it is difficult to perform attenuation measurements on certain types of samples, such as living tissue.

More recently, fiber optic techniques have been developed for the measurement of attenuation in which an optical fiber is used to guide the incident light to illuminate a sample inside a small chamber at the tip of a probe. A reflector is placed on the opposite side of the chamber to reflect the incident light into a second fiber which is associated with a detector. Unfortunately, these techniques find limited application with materials, such as fluids, that can readily pass into the chamber in the tip of the probe.

Conventionally, laser induced fluorescence (“LIFS”) spectroscopic techniques utilize various optical configurations in which a laser is directed at a sample using an optical fiber and the LIF from the sample is collected using a second optical fiber. Alternatively, the same fiber can be used for excitation of the sample and the collection of LIF. In either case, the LIF collected by the fiber is modulated by the sample, e.g. by the wavelength-dependent absorption and scattering of constituents of the sample. Therefore, existing LIFS methods are limited by the fact that the “intrinsic” or “true” fluorescence of the sample's fluorophores cannot be determined. Recent reports have suggested measuring the diffuse reflectance spectrum of the tissue as a means for correcting the intrinsic LIF using Monte-Carlo mathematical formulations. However, such correction methods are critically dependent on the backscattering characteristics of the tissue. Furthermore, backscattering does not account for the effects of absorption and scattering suffered by the intrinsic fluorescence prior to its measurement.

In biological LIF techniques, lasers are used to cause fluorophores in the sample to emit fluorescence. The main fluorophores in normal biological tissue are tryptophan, collagen and elastin. Other fluorophores, such as NAD (nicotinamide adenine dinucleotide) and FAD (flavin adenine dinucleotide), are also normally present, but at much lower concentrations. Under certain conditions, the contribution of the various fluorophores to the LIF can change. For example, during an ischemic or hypoxic event, the tissue is deprived of oxygen and anerobic respiration takes place. Consequently, the weak fluorophore NAD will be converted into the strong fluorophore reduced nicotinamide adenine dinucleotide (“NADH”). As a result, the LIF collected from the sample will reflect an increased contribution from NADH. This change is typically observed as a rise in the intensity of the LIF spectrum in the region of peak NADH emission, from about 470 to 490 nm. Thus, the metabolic state of the tissue can be determined by measuring the relative change in LIF intensity at the wavelength of peak NADH emission as compared to the relative change of the LIF intensity of fluorophores that are normally present in the tissue, such as elastin or collagen.

Unfortunately, the LIF of biological tissue is heavily modulated in the 390-450 nm range by the peak absorption of the main tissue chromophore, hemoglobin. Thus, although the intrinsic LIF spectra of normal tissue should approximately resemble that of the pure fluorophore components of the tissue, the measured LIF spectra has a valley in the 400-450 nm region that is associated with the hemoglobin absorption. As a result, the measured LIF spectrum tissue appears to have a double peak instead of the single peak spectrum associated with the pure fluorophores of the tissue. Hence, the LIF spectrum of normal tissue begins to resemble the spectrum associated with tissue suffering hypoxia or ischemia, which makes it more difficult to identify tissue abnormalities.

Furthermore, since the optical properties of biological tissue are influenced by its hemoglobin concentration, the measured LIF will vary with the level of blood perfusion throughout the cardiac cycle. The LIF of contractile tissue, such as the myocardium, is highly dependent on its state of contraction. Contraction increases the concentration of the fluorophore NADH and, hence, its contribution to tissue fluorescence. During contraction, however, blood is pumped out of the tissue, thereby reducing its hemoglobin concentration and, hence, light attenuation. Thus, a contracted myocardium retains less blood (i.e., a lower hemoglobin concentration) and, therefore, exhibits lower light absorption, than when relaxed.

Organs such as the brain, heart and kidney are the most sensitive to oxygen deficiency and can suffer permanent damage following an ischemic or hypoxic event. During open-heart surgery, for example, continuous monitoring of kidney perfusion, i.e., ischemia, is required. Similarly, since the success of a transplantation surgery is highly dependent on the level of organ perfusion at harvest and during preservation of the tissue, continuous monitoring of the organ is typically required.

Ischemia and hypoxia are both conditions that deprive tissue of oxygen, leading to anaerobic metabolism and the accumulation of the metabolic coenzyme NADH. The coenzyme NADH is a fluorescent molecule. Therefore, ischemia and hypoxia can be indirectly detected using LIF techniques by sensing increased concentrations of NADH and interpreting its elevation as a sign of oxygen deficiency. A common indicator of oxygen deficiency is the ratio between the LIF intensity at wavelengths associated with the peak fluorescence emission of NADH, collagen and elastin. However, such methods have not been practically applied for the detection of ischemia because of several complications. First, these methods cannot determine whether the elevated NADH concentration is caused by ischemia, hypoxia or hypermetabolism. Second, scarred or fibrosed tissue would be detected as normal because of their low NADH concentration. Finally, the indicator ratios are calculated by normalizing the intensity of NADH peak fluorescence by that of collagen or elastin. Although the fluoresence of the structural proteins elastin and collagen does not vary with tissue oxygenation, their fluorescence (top row is an elevation view; middle row is a side view; bottom row is a bottom view)11
I_c2(λ)^c=I_T(λ)e^{−α(λ).(R+y2′)}   (2)

The attenuation coefficient α(λ) can be calculated independent of the value of “R” from (1) and (2) as follows:
α(λ)={1/{y2′−y1′}}ln{I_c1(λ)^c/I_c2(λ)^c}  (3)

Where the difference {y2′−y1′} is a known constant, “e” is the natural exponential function and “ln” is the natural logarithm. It will be appreciated by those of ordinary skill in the art that the natural exponential and logarithm can be replaced by the common exponential and logarithm to the base 10, respectively. Similarly, for the probe configuration in FIG. 2(c), the attenuation coefficient can be calculated as follows:
α(λ)={1/{z2′−z1′}}ln{I_c1(λ)^c/I_c2(λ)^c}  (4)

Once the attenuation coefficient α(λ) is determined, the intrinsic fluorescence, I_T(λ), can be restored from either of the signals I_c1(λ)^c or I_c2(λ)^c (preferably I_c2(λ)^c where y2′<<y1′) by assuming an average effective range “R” from which most of the intrinsic fluorescence is collected. For biological tissue, the constant “R” is approximately about 0.2 mm at 308 nm excitation radiation. Therefore, the intrinsic fluorescence I_T(λ) can be obtained by substituting the measured I_c2(λ)^c into equation (2) and solving for I_T(λ) using the constant R and the known value y2′:
I_T(λ)=I_c2(λ)^ce^{α(λ).(R+y2′)}   (5)

Whether the measured attenuation accounts for absorption and/or scattering is mainly determined by the wavelength band of interest and the nature of the sample. The optical properties of biological tissue and the effects of tissue on LIF are greatly different for the wavelengths below and above approximately 600 nm. Below about 600 nm, the optical attenuation of biological tissue is primarily due to absorption and, hence, the attenuation coefficient α(λ) will represent absorptivity
%T(λ)=100·{I_c1(λ)^c/I_c2(λ)^c}  (7)

In an alternative embodiment of the LIFAS system 310, shown in FIG. 3(a), the source 311 emits laser radiation 312 at a wavelength and intensity capable of inducing fluorescence of the sample 314. The laser radiation 312 is reflected by a dichroic mirror 326 into an excitation-collection waveguide 321, which transmits the laser radiation 312 to the sample volume 313 where the laser radiation 312 excites local fluorophores to emit intrinsic fluorescence 329. Furthermore, the intensity of the laser 312 should be sufficient to create a sample volume 313 that overlaps with the numerical aperture 319 of the collection-only waveguide 322. The dimensions of the sample volume 313 will also depend on the numerical aperture of the excitation-collection waveguide 321, on the wavelength and intensity of the excitation radiation 312 and on the optical properties of the sample 314. The intrinsic fluorescence 329 is modulated, for example, by the absorption and scattering of the local chromophores (not shown) and scatterers (not shown) of the sample 314.

The excitation-collection waveguide 321 collects a first portion 328 of the return light 320 directly from the sample volume 313. The first portion 328 of the return light 320 is transmitted through the dichroic mirror 326 to the first sensor 316. The first sensor 316 generates a first signal 324a representing the intensity I_xc(λ) of the first portion 328 of the return light 320 at a plurality of wavelengths within a predetermined wavelength band.

As shown in FIG. 3(a), the collection-only waveguide 322 may be positioned such that the aperture 322a is laterally displaced from the excitation-collection waveguide 321 by the distance x3 and axially displaced from the aperture 321a of the excitation-collection waveguide 321 by the range y3. Preferably, the lateral distance x3 is zero while the axial distance y3 is non-zero. Also, the apertures 321a and 322a of the waveguides 321 and 322 are preferably positioned in close proximity to or in contact with the sample 314 during use of the LIFAS system 310, as shown in FIG. 3(a). Meanwhile, the numerical aperture 319 of the collection-only waveguide 322 is selected to include at least a portion of the sample volume 313.

The collection-only waveguide 322 collects a second portion 330 of the return light 320 which is transmitted to a second sensor 318. The second sensor 318 generates a second signal 325a representing the intensity I_co(λ) of the second portion 330 of the return light 320 at a plurality of wavelengths preferably within the same wavelength bands as the signal 324a generated by the first sensor 316. Furthermore, a longpass filter is placed in front of the aperture of each of the sensors 316 and 318 to selectively block backscattered excitation radiation.

In the embodiment shown in FIG. 3(a), the first and second portions 328 and 330 of the return light 320 experience different attenuation effects due to the unequal pathlengths traversed by the return light 320 from the sample volume 313 to the apertures 321a and 322a of the waveguides 321 and 322, respectively. The second portion 330 of the return light 220 collected by the aperture 322a travels an additional path-length through the tissue as compared to the first portion 328 of the return light 320 which is collected by the aperture 321a directly from the sample volume 313. Hence, the first portion 328 of the return light 320 suffers less path-length-dependent attenuation as compared to the second portion 330 of the return light 320. Thus, the signals 324a and 325a representing intensity I_xc(λ) and I_co(λ) at different wavelengths will exhibit different levels of modulation caused by the sample 314.

The signals 324a and 325a will also exhibit wavelength-dependent modulations caused by the instrumental effects. The wavelength-dependent modulations due to instrumental effects can be determined and, in turn, compensated for by conducting a calibration of the LIFAS system 310. System calibration is performed using light from a standard lamp (Qaurtz Halogen Lamp, Model No. 63358, Oriel Instruments, Stratford, Conn.) having a predetermined continuous spectrum to measure the wavelength-dependent instrumental effects of the LIFAS system 310. Once these instrumental effects are known, the processor 323 can be adapted to correct the measured intensities I_xc(λ) and I_co(λ) for modulations caused by the wavelength-dependent instrumental effects. The corrected intensities I_xc(λ)^c and I_co(λ)^c representing the intensity of the first and second portions 328 and 330 of the return light 320 at different wavelengths can then be used by the processor 323 to determine the wavelength-dependent attenuation α(λ) of the sample 314. As discussed in detail, below, once the attenuation is known, either of the signals 324a or 325a, preferably 324a, can be corrected for the effects of attenuation to restore the intrinsic LIF 329 of the fluorophores in the sample volume 313.

In the preferred embodiment, the source 311 is an Xe—Cl excimer laser emitting ultraviolet excitation radiation at a wavelength of 308 nm. The waveguides 321 and 322 are advantageously comprised of optical fibers or optical fiber bundles, which can be integrated into a probe for ease of use and durability. A 1.4 mm diameter optical bundle and a 0.4 mm optical fiber are used as the excitation-collection and the collection-only waveguides, respectively. Preferably, the collection-only waveguide is a plurality of 0.4 mm optical fibers disposed about the periphery of the excitation-collection waveguide. The optical bundle and fibers are made of fused silica which is transparent to the 308 nm ultraviolet radiation. Furthermore, the probe may be configured within, for example, the shaft of a hypodermic needle, so that the in-vivo physiological or pathological properties of biological tissue can be determined. Each of the sensors 316 and 318 is a spectrograph (Model FF250, ARIES Inc., Concorde, Mass.) associated with a 1024 element intensified photo diode array (PDA) to image the resolved spectrum. Each PDA is connected to an optical multichannel analyzer (OMA III, EG&G Princeton Applied Research Corporation, Princeton, N.J.) which measures the intensity of the light spectrum imaged by the PDA and produce the signals 324a and 325a. The processor 323 is a personal computer that is networked to each of the sensors 316 and 318 via the signal paths 324 and 325 to receive the signals 324a and 325a, respectively.

The method of determining the attenuation coefficient α(λ) and the intrinsic fluorescence I_T(λ) of a sample using the embodiment of FIG. 3(a) in accordance with the current invention is as follows. In the embodiment shown in FIG. 3(a), “D” represents the effective range, from which the collection waveguide 321 collects a majority of the portion 328 of the return light 320. The portion 328 of the return light 320 is collected directly from the sample volume 313 by the excitation-collection waveguide 321. Therefore, the wavelength-dependent intensity I_xc(λ)^c of the portion 328 of the return light 320 collected by the excitation-collection waveguide 321 will be substantially similar to the wavelength-dependent intensity of the return light 320, represented as I_o(λ). However, the return light 320 with the intensity I_o(λ) will suffer additional wavelength-dependent attenuation as it travels the extra path-length to reach the aperture 322a and is collected as the second portion 330 of the return light 320 with the intensity I_co(λ)^c. The lateral and axial distances x3 and y3 will be predetermined by the configuration of the waveguides 321 and 322. In the preferred embodiment, the lateral distance x3=0 and the axial distance y3>0. Thus, the wavelength-dependent intensity I_co(λ)^c can be approximated by the following equations:
I_xc(λ)^c≈I_o(λ)=I_T(λ)e^−α(λ)·D  (8)
I_co(λ)^c≈I_o(λ)e^{−α(λ)·y3}  (9)
≈I_xc(λ)^ce^{−α(λ)·y3}   (9)
α(λ)={1/y³}ln{I
Similarly, for the alternative probe configuration shown in FIG. 3(b), where the lateral distance x3′>0 and the axial distance y3′=0, the attenuation coefficient can be approximated as follows:
α(λ)={1/x3′}ln{I_xc(λ)^c/I_co(λ)^c}  (11)
Once the attenuation coefficient α(λ) is determined, the intrinsic fluorescence, I_T(λ), can be restored from either of the signals I_xc(λ)^c or I_co(λ)^c, preferably I_xc(λ)^c, by assuming an average effective range “D” from which most of the intrinsic fluorescence is collected. For biological tissue, the constant “D” is approximately about 0.2 mm at 308 nm excitation radiation. By substituting the measured I_xc(λ)^c into equation (7) and solving for I_T(λ):
I_T(λ)=I_xc(λ)^ce^α(λ).D   (12)

As discussed in connection with the previous embodiment, the measured attenuation may account for absorption and/or scattering depending on the wavelength band of interest and the nature of the sample. Below about 600 nm, the optical attenuation of biological tissue is primarily due to absorption and, hence, the attenuation coefficient α(λ) will represent absorptivity
%T(λ)={I_co(λ)^c/I_xc(λ)^c}·100   (14)

In another alternative embodiment shown in FIG. 4, a LIFAS system 410 in accordance with the present invention has been adapted for biomedical applications. In this particular embodiment, the LIFAS system 410 has been adapted to determine the optical attenuation of biological tissue. The laser 411 emits radiation 412 at a wavelength capable of exciting the tissue 414 to emit fluorescence. The radiation 412 is directed through an iris 440 at a dichroic mirror 442 which reflects the radiation 412 onto a lens 444 which focuses the radiation 412 onto the proximal tip 445 of an optical fiber 446. The adjustable iris 440 is advantageously used to reduce the energy of the radiation 412. Alternatively, the adjustable iris 440 can be replaced by any suitable attenuator.

The probe 448 includes a central optical fiber 446 that is used as the excitation-collection fiber and peripheral optical fibers 450a-h are used as the collection-only fibers. The distal end of the optical fibers 446 and 450a-h are incorporated into the optical probe 448. The aperture 448a the optical fiber probe 448 is placed in proximity to the tissue 414 so that the aperture 446a of the excitation-collection fiber 446 and apertures 451a-h of the collection-only fibers 450a-h are in contact with the tissue 414. As shown in a partial perspective view in FIG. 5(a), the optical fiber probe 448 includes a central optical fiber 446 and a plurality of optical fibers 450a-h disposed about the periphery of the central optical fiber 446. The apertures 451a-h of the distal ends of the collection-only fibers 450a-h are axially displaced by a small distance y3 with respect to the aperture 446a of the excitation-collection fiber 446. The return light 420 collected by apertures 451a-h of the collection-only fibers 450a-h pool into the aperture 418a of the sensor 418. It will be appreciated by those in the art that the optical fibers 446 and 450a-h can advantageously be replaced by optical fiber bundles in order to obtain greater flexibility and durability. For example, as shown in FIG. 5(b), the central optical fiber 446 can be replaced with an optical fiber bundle 454.

The excitation radiation 412 is transmitted through the optical fiber 446 to the tissue 414 to induce intrinsic fluorescence of the fluorophores of the tissue 414. The intrinsic fluorescence is modulated, for example, by the chromophores and/or scatterers of the tissue 414. A first portion 428 of the return light 420 is collected by the optical fiber 446 from the tissue volume that is directly irradiated by the excitation radiation 412 and transmitted through the optical fiber 446 to a first lens 444 where the return light is directed through the dichroic mirror 442 and then focused by a second lens 449 onto the aperture 416a of a first sensor 416. Similarly, the collection-only optical fibers 450a-h collect a second portion 430 of the return light 420 and transmit the second portion 430 of the return light 420 to the second sensor 418. Preferably, longpass filters are placed in front of the apertures 416a and 418a of the sensors 416 and 418 to selectively block backscattered excitation radiation. A first signal 424a and a second signal 425a representing the intensity of the first and second portions 428 and 430, respectively, of the return light 420 are generated by the first and second sensors 416 and 418 and transmitted via signal paths 424 and 425 to the processor 423. The processor 423 uses the first and second signals 424a and 425a generated by the detectors 416 and 418 to determine the wavelength-dependent attenuation of the sample 414 using equations (8), (9) and (10) as described in the previous embodiment.

In a particularly preferred embodiment of the LIFAS probe 448 shown in FIG. 8, an illumination source 490 emitting visible light 492 is mounted on the probe 448 to act as a spotlight for the operator. Because room illumination can contaminate the spectral measurements of the system 410 by adding background light, use of the illumination source 490 will allow the operator to see and accurately position the probe 448 under low-light conditions. The illumination source 490 is configured to illuminate the sample at all times, except when the LIFAS system is monitoring return light from the sample 414.

The biological electrical signal arising in nerves or contractile tissue, such as muscle, is known as the “action potential” and is caused by sudden changes in the ion conductivity of the cell membrane. The occurrence of an action potential in contractile tissue initiates a contraction. For example, in cardiac tissue, the action potential propagates and spreads in a wave-like manner to induce a local myocardial contraction wherever it travels.

An action potential propagating in the tissue local to the aperture 448a of the probe 448 can be detected by the electrode 464 which is incorporated into the probe 448 as shown, for example, in FIG. 6(a). Voltage alternations caused by the occurrence of the action potential are picked up by the electrode 464 in reference to the common electrode 466 and are transmitted to the amplifier 460. The common electrode 466 provides the electrical ground for the amplifier 460 by maintaining contact with the tissue of interest 414 or the whole body. As discussed in connection with FIG. 7, below, the amplified action potential 468 is transmitted to the processor 423 to trigger the acquisition process of the LIFAS system 410 at a pre-selected phase of the tissue contraction or of the cardiac cycle, whichever is applicable.

In an alternative probe arrangement, a plurality of electrodes or fibers with a conductive coating can be distributed circumferentially about the tip of the optical fiber probe 448. For example, as shown in FIG. 6(b), the probe 448 can be equipped with three electrodes 470, 472 and 474 arranged in a triangular configuration. The action potentials measured by each of the electrodes 470, 472 and 474 are amplified through separate channels of the amplifier 460 and transmitted to the processor 423. As illustrated in FIG. 7, the processor 423 processes the received signals to determine the direction of propagation of the contraction vector 462. In particular, the processor detects the phase lead/lag between the action potentials collected by the electrodes 470, 472 and 474 to determine the orientation of the contraction vector 462 with respect to the location of the electrodes 470, 472 and 474.

For example, the contraction vector 462 shown in FIG. 7(a) is propagating from the tissue site 476a to the site 478a and, hence, as shown in FIG. 7(c), the action potential 474a arrives before the action potential 472a which in turn arrives before the action potential 470a. Moreover, the phase difference or time delay between a pair of action potentials indicates how the contraction vector 462 is centered between the location of the corresponding pair of electrodes. For example, as shown in FIG. 7(b), the contraction vector 462 propagates from the tissue site 476b to the site 478b. As a result, the corresponding time of arrival of action potentials 470b, 472b and 474b, as shown in FIG. 7(d), will vary. The processor 423 processes the signals and indicates the direction of the propagation of the contraction to the system operator. The direction of propagation can be indicated, for example, by a circular array 496 of light emitting diodes (LED) 498 mounted circumferentially on the probe 448, as shown in FIG. 8. The processor 423 transmits a signal to the LED array 496 so that, for example, only the LED element pointing in the direction of contraction propagation would glow.

It will be appreciated that the amplifier 460 can be replaced by any device that can measure the electrical activity of biological tissue such as a differential amplifier, an electrocardiogram (ECG), an electromyogram (EMG), an electroencephalogram (EEG), depending on the LIFAS application. Furthermore, it will be appreciated that optical fibers with a metallic or electrically conductive coating can be used in place of one or all of the electrode 464, 470, 472 or 474 to measure the action potential of the tissue. It should be noted that, conventional ECG is generally not suitable for triggering the acquisition of LIFS or LIFAS systems, since it does not accurately indicate the instantaneous state of myocardial contraction at the sample volume. However, customary ECG using limb or chest leads can be used to trigger data acquisition of the LIFAS system 410 where the sample 414 is non-contractile tissue.

In the LIFAS system 410 shown in FIG. 4, the light source 411 is preferably a lamp or a laser that emits ultraviolet, visible or infrared radiation. In the preferred embodiment, the source 411 is an XeCl excimer laser emitting pulses of ultraviolet excitation radiation at 308 nm. Where ultraviolet radiation is used, it is advisable that the optical components used in the acquisition system be made of synthetic quartz (fused silica) to ensure maximal ultraviolet transmission and minimal instrumental fluorescence. Alternatively, a nitrogen laser, a helium-cadmium laser, a frequency-multiplied laser, a solid-state laser, an arc lamp or a light-emitting diode can be used as the light source 411. The energy of the excitation light 412 is typically between 0.001-10 m Joules. However, it will be appreciated that the selected energy level should be low enough to avoid tissue ablation and/or photobleaching while still being adequate to produce detectable LIF.

In the preferred embodiment, the sensors 416 and 418 are each comprised of a spectrograph (Model FF250, ARIES Inc., Concorde, Mass.) associated with a 1024 element intensified photo diode array (PDA) detector. An optional low fluorescence, long-pass filter (not shown) with a cutoff wavelength above 308 nm, preferably 335 nm (Schott WG335), is placed before the entrance slit of each spectrograph to selectively block any backscattered excitation radiation from reaching the sensors 416 and 418. The entrance slit of the spectrograph preferably has a width of 100 micrometers. The spectrograph uses a 150 lines per millimeter diffraction grating to disperse the incoming return light 420 into its spectral components.

The spectrum formed by the spectrograph is imaged by a detector, preferably an intensified linear photodiode array (Model 1420, EG&G Princeton Applied Research Corporation, Princeton, N.J.) facing the output port of the spectrograph. The photodiode array generates a plurality of electrical signals representing the intensity of the return light 420 at wavelengths within predetermined wavelength bands. Alternatively, the sensors can be constructed of any suitable materials, such as individual light-sensitive diodes with appropriate band-pass filters for the analysis of spectral bands of the return light or an optical spectrum analyzer (“OSA”) for analysis of a broader spectrum. Selection of the return light monitoring device will depend on a variety of factors, including cost, accuracy, resolution, and whether the user is interested in monitoring a single wavelength, a wavelength band or an entire spectrum.

Although in the preferred embodiments shown in FIGS. 4 and 5, the optical fiber probe includes a central excitation-collection optical fiber or optical fiber bundle with a plurality of collection-only optical fibers disposed around its periphery, it will be appreciated by those of ordinary skill in the art that the probe can take many forms. Furthermore, it will be appreciated that the central optical fiber can be used as the collection-only waveguide while all or some of the peripheral fibers can be used for excitation-collection or collection-only. The latter arrangement is preferred when testing highly-attenuating samples to achieve a better signal-to-noise ratio.

FIG. 9 shows three alternative geometrical configurations of the collection-only optical fibers
The matrices “S” and “R” holds the input and output training vectors as their columns, respectively. The superscripts ^T and ⁻¹ indicate matrix transpose and inversion, respectively, and “I” is the identity matrix. The parameter η is initially set to 0.98; however, it can assume any value between 0 and 1 depending on the noise of the system. Following the training stage, the output “r” for an unknown input “s” can be readily calculated from the dot product:
r=M(η).s  (18)
Finally, an appropriate transfer function is used to assign the output “r” into one of several predetermined classification categories.

In the present invention, the MAM classifier is initially trained with LIFA (or absorbance) spectra acquired from normal or ischemic tissue as the training inputs. The corresponding normal or ischemic state can be encoded as, for example, “−1” or “1,” respectively and used as the training outputs. Therefore, the training LIFA spectra are placed as columns of the input matrix “S” while their corresponding state-coded values are arranged in the same order as elements of the output row vector “R.” The number of columns in “S” and elements in “R” are equal to the number of available training sets. A trained MAM matrix can determine whether an unknown LIFA spectrum “s_?” has been acquired from normal or ischemic tissue by calculating the dot product r_?=M(η), s_? and passing the scalar result “r_?” to a hard limit transfer function. The hard limit transfer function then converts any negative or positive values of “r_?” into “0” or “1” indicating a normal or ischemic classification, respectively.

In a similar fashion, the MAM classifier can be applied for the detection of hypoxia and the discrimination between normal, ischemic and hypoxic tissue. It should be apparent that common or intrinsic LIF spectra can replace LIFA spectra as the classifier input. In addition, this input can be an entire spectrum, a re-sampled version of a spectrum or a set of statistical parameters or features characterizing a spectrum. Similarly, actual ischemia or hypoxia levels can be used as the classifier outputs instead of the binary coded output values employed in the above demonstration. In this case a linear transfer function may be used to categorize the output “r” of the MAM classifier.

The MAM technique outperformed the commonly-used artificial neural network (ANN) classifier in accurately classifying LIF/LIFA spectra resulting from normal and ischemic tissue. The superiority of the MAM classifier is most probably due to its insensitivity to spectral noise that might be present in LIF/LIFA spectra measured from biological systems. Although the foregoing MAM classifier is currently applied to discriminate spectral data for the purpose of ischemia or hypoxia detection, it is understood that those skilled in the art may apply it in various ways for different classification purposes.

Tissue Characterization

It will also be appreciated that the LIFAS devices and methods can be applied to tissue characterization, i.e., to differentiate between normal and diseased tissue, for tissue diagnostics and malignancy detection. Current LIFS techniques use the intensity spectrum of modulated LIF to identify malignant (cancerous and pre-cancerous) tissue and classify its type. LIFAS techniques offer a unique tissue characterizing capability not offered by conventional LIFS, based upon measurement of the attenuation spectrum.

A simple demonstration of LIFAS diagnostic capability is shown in FIGS. 17(a)-(d). Although normal kidney and heart tissue are different in nature, their common LIF spectra, shown in FIGS. 17(a) and (b) are almost identical and, hence, are not so useful for classification purposes. However, heart and kidney LIFA spectra, shown in FIGS. 17(c) and d are different in terms of both shape and peak attenuation values. Thus, LIFAS techniques offer better tissue identification power than conventional LIFS techniques.

Other biomedical applications of the LIFAS methods and devices will include laser removal of decorative tattoos, detection of the in-vivo glucose level, assessment of the degree of burn trauma, detection of atherosclerotic plaque, angioplasty, measurement of acidity or alkalinity, pH measurement, the analysis of biochemical fluids, and the like. For example, measurement of the in vivo skin absorption using LIFAS methods and devices in accordance with the present invention can aid in the selection of optimal laser wavelengths for removing tattoos of different colors. Furthermore, since LIFAS techniques can be used to determine absorbance from a hypodermic sample volume, the skin color and the depth of the tattoo dye can be more accurately characterized than in surface reflectance techniques. Similarly, burn injury assessment can be accomplished by using LIFAS techniques to measure the depth of burn by probing for the presence of blood perfusion at varying locations within the tissue. LIFAS techniques can also be used to determine the absorbance or turbidity of a liquid in-situ without the necessity of extracting a sample for use in a spectrometer.

It will be understood by those of ordinary skill in the art that, although the LIFAS system and method is shown in the exemplary method as applied to biological tissue, it is also readily applicable to chemical and industrial material. For example, LIFAS devices and methods can be used to measure the absorbance and/or turbidity of materials and mixtures in medical, food, beverage, detergent, plastic, glass, oil, paint, textile, and semiconductor applications. Furthermore, the concentration of the pure components of the mixture can be determined from the absorbence spectrum using chemometric techniques such as multivariate regression (MLR), partial least squares (PLS) or artificial neural networks described above.

Although the foregoing discloses preferred embodiments of the present invention, it is understood that those skilled in the art may make various changes to the preferred embodiments without departing from the scope of the invention. The invention is defined only by the following claims.

Claims

1. A spectroscopic method of analyzing a sample, comprising:

irradiating a sample with radiation to produce fluorescence from the sample, wherein the fluorescence is modulated by the sample;

monitoring a first portion of the modulated fluorescence at a first distance from the sample;

monitoring a second portion of the modulated fluorescence at a second distance from the sample, the second distance being different from the first distance; and

comparing the first and second portions of the modulated fluorescence to each other to determine a modulation characteristic of the sample.

2. The method of claim 1, wherein the radiation comprises substantially monochromatic light.

3. The method of claim 1, wherein the radiation comprises laser light.

4. The method of claim 1, wherein irradiating the sample comprises directing radiation at the sample using a waveguide.

5. The method of claim 4, wherein the waveguide is an optical fiber.

6. The method of claim 4, wherein the waveguide is an optical fiber bundle.

7. The method of claim 1, wherein monitoring of the modulated fluorescence comprises:

collecting a portion of the modulated fluorescence; and

determining the intensity of the collected portion of modulated fluorescence.

8. The method of claim 7, wherein the first portion of the modulated fluorescence is collected with a first waveguide and the second portion of the modulated fluorescence is collected with a second waveguide.

9. The method of claim 8, wherein the first waveguide is an optical fiber.

10. The method of claim 8, wherein the first waveguide is an optical fiber bundle.

11. The method of claim 8, wherein the second waveguide is an optical fiber.

12. The method of claim 8, wherein the second waveguide is an optical fiber bundle.

13. The method of claim 1, wherein irradiating the sample comprises directing radiation to the sample using a first waveguide and wherein the fluorescence is monitored using the first waveguide.

14. The method of claim 7, wherein the intensity of the collected portion of the fluorescence is determined with a sensor.

15. The method of claim 7, wherein the intensity of the first portion of the modulated fluorescence is determined with a sensor.

16. The method of claim 7, wherein the intensity of the second portion of the modulated fluorescence is determined with a sensor.

17. The method of claim 7, wherein the intensity of the first portion of the modulated fluorescence is determined with a first sensor and the intensity of the second portion of the modulated fluorescence is determined with a second sensor.

18. The method of claim 7, wherein the first and second portions of the modulated fluorescence are measured consecutively.

19. The method of claim 7, wherein the first and second portions of the modulated fluorescence are measured simultaneously.

20. The method of claim 11 1, wherein the method further includes determining the intrinsic fluorescence of the sample.

21. The method of claim 1, wherein the sample is biological material.

22. The method of claim 21, wherein the biological material is living tissue.

23. The method of claim 21, wherein the method further includes determining a physiological property of the biological material using the modulation characteristic.

24. The method of claim 21, wherein the method further includes determining a pathological property of the biological material using the modulation characteristic.

25. The method of claim 22, wherein the method further includes determining a physiological property of the living tissue using the modulation characteristic.

26. The method of claim 25, wherein the physiological property of the tissue is tissue oxygenation.

27. The method of claim 22, wherein the method further includes determining a pathological property of the tissue using the modulation characteristic.

28. The method of claim 27, wherein the pathological property of the tissue is the malignant condition of the tissue.

29. The method of claim 1, wherein either but not both of the distances is substantially zero.

30. A spectroscopic method of analyzing a sample, comprising:

irradiating a sample with radiation to produce return radiation from the sample, wherein the return radiation is modulated by the sample;

monitoring a first portion of the modulated return radiation at a first distance from the sample;

monitoring a second portion of the modulated return radiation at a second distance from the sample;

processing the first and second portions of the modulated return radiation to determine a modulation characteristic of the sample,

wherein the return radiation is modulated by attenuation.

31. The method of claims 30, wherein the return radiation is attenuated by scattering.

32. The method of claim 30, wherein the return radiation is attenuated by adsorption.

33. The method of claim 30, wherein the modulation characteristic of the sample is attenuation.

34. The method of claim 30, wherein the modulation characteristic of the sample is absorption.

35. The method of claim 34, wherein the method further includes determining transmittance.

36. The method of claim 30, wherein the modulation characteristic of the sample is optical rotation.

37. A spectroscopic method of analyzing a sample, comprising:

irradiating a sample with radiation to produce return radiation from the sample, wherein the return radiation is modulated by the sample;

monitoring a first portion of the modulated return radiation at a first distance from the sample;

monitoring a second portion of the modulated return radiation at a second distance from the sample;

processing the first and second portions of the modulated return radiation to determine a modulation characteristic of the sample;

wherein the sample is biological material;

wherein the method further includes determining a physiological property of the tissue using the modulation characteristic; and

wherein the physiological property of the tissue is hypoxia.

38. A spectroscopic method for determining the oxygenation of a biological material, comprising:

irradiating a sample of a biological material with radiation to produce fluorescence from the sample, wherein the fluorescence is modulated by attenuation of the sample;

monitoring a first portion of the modulated fluorescence at a first distance from the sample;

monitoring a second portion of the modulated fluorescence at a second distance from the sample, the second distance being different from the first distance;

comparing the first and second portions of the modulated fluorescence to each other to determine the attenuation of the sample; and

determining oxygenation of the sample using the attenuation of the sample.

39. A spectroscopic method for determining the oxygenation of a biological material, comprising:

irradiating a sample of a biological material with radiation to produce return radiation from the sample, wherein the return radiation is modulated by attenuation of the sample;

monitoring a first portion of the modulated return radiation at a first distance from the sample;

monitoring a second portion of the modulated return radiation at a second distance from the sample;

processing the first and second portions of the modulated return radiation to determine the attenuation of the sample;

determining oxygenation of the sample using the attenuation of the sample;

wherein the oxygenation of the sample is determined by comparing the attenuation of the sample to the attenuation of a sample having a known level of oxygenation.

40. A spectroscopic method for determining the concentration of hemoglobin in a biological material, comprising:

irradiating a sample of biological material with radiation to produce fluorescence from the sample, wherein the fluorescence is modulated by attenuation of the sample;

monitoring a first portion of the modulated fluorescence at a first distance from the sample;

monitoring a second portion of the modulated fluorescence at a second distance from the sample, the second distance being different from the first distance;

comparing the first and second portions of the modulated fluorescence to each other to determine the attenuation of the sample; and

determining the concentration of hemoglobin in the sample using the attenuation of the sample.

41. A spectroscopic method for determining the concentration of hemoglobin in a biological material, comprising:

irradiating a sample of a biological material with radiation to produce return radiation from the sample, wherein the return radiation is modulated by attenuation of the sample;

monitoring a first portion of the modulated return radiation at a first distance from the sample;

monitoring a second portion of the modulated return radiation at a second distance from the sample;

determining the concentration hemoglobin in the sample using the attenuation of the sample;

wherein the concentration of hemoglobin is determined by comparing the attenuation of the sample to the attenuation of a sample having a known concentration of hemoglobin.

42. A method for determining a physiological characteristic of a biological material, comprising:

irradiating a sample of biological material with radiation to produce fluorescence from the sample, wherein the fluorescence is modulated by the sample;

monitoring a first portion of the modulated fluorescence at a first distance from the sample;

monitoring a second portion of the modulated fluorescence at a second distance from the sample, the second distance being different from the first distance; and

comparing the first and second portions of the modulated fluorescence to each other, using a predictive model, to determine a physiological characteristic of the sample.

43. A method for determining a physiological characteristic of a biological material, comprising:

irradiating a sample of a biological material with radiation to produce return radiation from the sample, wherein the return radiation is modulated by the sample;

monitoring a first portion of the modulated return radiation at a first distance from the sample;

monitoring a second portion of the modulated return radiation at a second distance from the sample;

processing the first and second portions of the modulated return radiation, using a predictive model, to determine a physiological characteristic of the sample;

wherein the predictive model is a multivariate linear regression.

44. A method for determining a physiological characteristic of biological material, comprising:

irradiating a sample of biological material with radiation to produce fluorescence from the sample, wherein the fluorescence is modulated by the sample;

monitoring a first portion of the modulated fluorescence at a first distance from the sample;

monitoring a second portion of the modulated fluorescence at a second distance from the sample, the second distance being different from the first distance;

comparing the first and second portions of the modulated fluorescence to each other to determine a modulation characteristic of the sample; and

processing the modulation characteristic, using a predictive model, to determine a physiological characteristic of the sample.

45. A method for determining a physiological characteristic of a biological material, comprising:

irradiating a sample of a biological material with radiation to produce return radiation for the sample, wherein the return radiation is modulated by the sample;

monitoring a first portion of the modulated return radiation at a first distance from the sample;

monitoring a second portion of the modulated return radiation at a second distance from the sample;

processing the first and second portions of the modulated return radiation, using a predictive model, to determine a physiological characteristic of the sample;

wherein the predictive model is a multicriteria associative memory classifier.

46. Apparatus for analyzing a sample, comprising:

a source adapted to emit radiation that is directed at a sample to produce fluorescence from the sample, wherein the fluorescence is modulated by the sample;

a first sensor adapted to monitor the fluorescence at a first distance from the sample and generate a first signal indicative of the intensity of the fluorescence;

a second sensor adapted to monitor the fluorescence at a second distance from the sample and generate a second signal indicative of the intensity of the fluorescence, the second distance being different from the first distance; and

a processor associated with the first sensor and the second sensor and adapted to compare the first and second signals to each other to determine a modulation characteristic of the sample.

47. The apparatus of claim 46, wherein fiber optics transmit the fluorescence to the sensors.

48. Apparatus for analyzing a sample, comprising:

a source adapted to emit radiation that is directed at a sample volume in a sample to produce fluorescence from the sample, such fluorescence including modulated fluorescence resulting from modulation by the sample;

a first sensor adapted to monitor the fluorescence at a first distance from the sample volume and generate a first signal indicative of the intensity of the fluorescence;

a second sensor adapted to monitor the fluorescence at a second distance from the sample volume and generate a second signal indicative of the intensity of the fluorescence, the second distance being different from the first distance; and

a processor associated with the first sensor and the second sensor and adapted to compare the first and second signals to each other to determine a modulation characteristic of the sample.

49. Apparatus for determining a modulation characteristic of a biological material, comprising:

a source adapted to emit excitation light;

a first waveguide disposed at a first distance from the sample adapted to transmit the excitation light from the light source to the biological material to cause the biological material to produce fluorescence, and adapted to collect a first portion of the fluorescence;

a first sensor, associated with the first waveguide, adapted to measure the intensity of the first portion of the fluorescence and generate a first signal indicative of the intensity of the first portion of the fluorescence;

a second waveguide disposed at a second distance from the sample adapted to collect a second portion of the fluorescence, the second distance being different from the first distance;

a second sensor, associated with the second waveguide, adapted to measure the intensity of the second portion of the fluorescence and generate a second signal indicative of the intensity of the second portion of the fluorescence; and

a processor adapted to compare the first and second signals to each other to determine a modulation characteristic of the biological material.

50. Apparatus for analyzing a sample, comprising:

a source adapted to emit radiation that is directed at a sample volume in a sample to produce fluorescence from the sample, such fluorescence including modulated fluorescence resulting from modulation by the sample;

a first sensor, displaced by a first distance from the sample volume adapted to monitor the fluorescence and generate a first signal indicative of the intensity of the fluorescence; and

a second sensor, displaced by a second distance from the sample volume adapted to monitor the fluorescence and generate a second signal indicative of the intensity of fluorescence, the second distance being different from the first distance; and

a processor associated with the first sensor and the second sensor and adapted to compare the first and second signals to each other to determine a physiological property of the sample.

51. Apparatus for determining a physiological property of a biological material, comprising:

a source adapted to emit excitation light;

a first waveguide disposed at a first distance from the sample, and adapted to transmit the excitation light from the light source to the biological material to cause the biological material to produce fluorescence, and further adapted to collect a first portion of the fluorescence;

a first sensor, associated with the first waveguide, for measuring the intensity of the first portion of the fluorescence and generating a first signal representative of the intensity of the first portion;

a second waveguide disposed at a second distance from the sample, and adapted to collect a second portion of the fluorescence, the second distance being different from the first distance;

a second sensor, associated with the first waveguide, for measuring the intensity of the second portion of the fluorescence and generating a second signal representative of the intensity of the second portion; and

a processor adapted to compare the first and second signals to each other to determine a physiological property of the biological material.

52. A spectroscopic method of analyzing a sample, comprising:

irradiating a sample with radiation to produce fluorescence from the sample, wherein the fluorescence is modulated by the sample;

monitoring a first portion of the modulated fluorescence at a first distance from the sample;

monitoring a second portion of the modulated fluorescence at a second distance from the sample, the second distance being different from the first distance;

comparing the first and second portions of the modulated fluorescence to each other to determine a modulation characteristic of the sample;

wherein the sample is a biological material tissue;

wherein the method further includes determining a physiological property of the tissue using the modulation characteristic; and

wherein the physiological property of the tissue is ischemia.

53. A method for determining a physiological characteristic of a biological material, comprising:

irradiating a sample of a biological material with radiation to produce fluorescence from the sample, wherein the fluorescence is modulated by the sample;

monitoring a first portion of the modulated fluorescence at a first distance from the sample;

monitoring a second portion of the modulated fluorescence at a second distance from the sample, the second distance being different from the first distance; and

comparing the first and second portions of the modulated fluorescence to each other, using a predictive model, to determine a physiological characteristic of the sample; ,

wherein the predictive model is multivariate.

54. A spectroscopic method of analyzing a sample, comprising:

irradiating a sample with radiation to produce fluorescence from the sample, wherein the fluorescence is modulated by the sample;

monitoring a first portion of the modulated fluorescence at a first angle from the sample;

monitoring a second portion of the modulated fluorescence at a second angle from the sample; and

comparing the first and second portions of the modulated fluorescence to each other to determine a modulation characteristic of the sample.