United States Patent Application: 0060167348

From: Subject: United States Patent Application: 0060167348 Date: Tue, 15 Jan 2008 14:51:46 -0600 MIME-Version: 1.0 Content-Type: multipart/related; type="text/html"; boundary="----=_NextPart_000_0000_01C85786.26B66160" X-MimeOLE: Produced By Microsoft MimeOLE V6.00.2900.3198 This is a multi-part message in MIME format. ------=_NextPart_000_0000_01C85786.26B66160 Content-Type: text/html; charset="Windows-1252" Content-Transfer-Encoding: quoted-printable Content-Location: http://appft1.uspto.gov/netacgi/nph-Parser?Sect1=PTO1&Sect2=HITOFF&d=PG01&p=1&u=/netahtml/PTO/srchnum.html&r=1&f=G&l=50&s1='20060167348'.PGNR.&OS=DN/20060167348&RS=DN/20060167348 United States Patent Application: 0060167348

(=20 1 of 1 )

United States Patent = Application	*20060167348* =
Kind Code	A1
Arnold; Mark A. ; et = al.	July 27, 2006 =

Method for generating a net analyte signal calibration = model and=20 uses thereof

Abstract

A method for generating a net analyte signal calibration model for = use in=20 detecting and/or quantifying the amount of an analyte in a test subject. = The net=20 analyte signal can be generated by providing a set of in vivo infrared = spectra=20 for a test subject during a period in which an analyte concentration is=20 essentially constant; calculating an optimal subspace of spectra that at = least=20 substantially describes all non-analyte dependent spectral variance in = the in=20 vivo spectra; providing a pure component infrared spectrum for the = analyte; and=20 calculating a net analyte signal spectrum from a data set comprising the = optimal=20 subspace spectra and the pure analyte spectrum. The net analyte signal=20 calibration model can be used, for example, in measuring the = concentration of=20 analyte in a test subject, and/or for evaluating the analytical = significance of=20 an in vivo multivariate calibration model.

Inventors:	Arnold; Mark A.; (Iowa = City, IA)=20 ; Olesberg; Jonathon T.; (Iowa City, IA)
Correspondence Name and = Address:	NEEDLE & ROSENBERG, = P.C. SUITE 1000 999 PEACHTREE STREET ATLANTA GA 30309-3915 US
Serial No.:	042817
Series Code:	11
Filed:	January 24, = 2005

U.S. Current = Class: 600/310; 356/300 =

U.S. Class at = Publication: 600/310; 356/300 =

Intern'l Class: = A61B 5/00 20060101 = A61B005/00; G01J=20 3/00 20060101 G01J003/00

Goverment Interests

[0001] The invention described in the foregoing specification = has been=20 developed in part with finds received from the National Institute of = Diabetes=20 and Digestive and Kidney Diseases of the National Institutes of Health = under=20 grant numbers DK-60657 and DK-02925. The United States Government may = have=20 certain rights under this invention.=20

Claims

1. A method for generating a net analyte signal calibration = model for=20 use in detecting an analyte in a test subject, comprising: a) providing = a set of=20 in vivo infrared spectra for a test subject during a period in which an = analyte=20 concentration is essentially constant; b) calculating an optimal = subspace of=20 spectra that at least substantially describes all non-analyte dependent = spectral=20 variance in the in vivo spectra of step a); c) providing a pure = component=20 infrared spectrum for the analyte; and d) calculating a net analyte = signal=20 spectrum from a data set comprising the optimal subspace spectra of b) = and the=20 pure analyte spectrum of c), wherein the net analyte signal spectrum = identifies=20 one or more in vivo spectral features specific to the analyte. =

2. The=20 method of claim 1, wherein the spectra are absorption spectra. =

3. The=20 method of claim 1, wherein the spectra are reflectance spectra. =

4. The=20 method of claim 1, wherein the spectra are single-beam spectra. =

5. The=20 method of claim 2, wherein the absorption spectra are near infrared = absorption=20 spectra in the range of from approximately 4000 cm.sup.-1 to = approximately 5000=20 cm.sup.-1.

6. The method of claim 2, wherein the absorption = spectra are=20 near infrared absorption spectra in the range of from approximately 5500 = cm.sup.-1 to approximately 6500 cm.sup.-1.

7. The method of = claim 1,=20 wherein the infrared spectra are absorption spectra in the mid infrared = spectral=20 range.

8. The method of claim 7, wherein the absorption spectra = are in=20 the range of from approximately 1200 cm.sup.-1 to approximately 900 = cm.sup.-1.=20

9. The method of claim 1, wherein the analyte is a physiological = chemical.

10. The method of claim 9, wherein the analyte is = glucose,=20 urea, lactate, triglyceride, total protein, cholesterol, or ethanol. =

11.=20 The method of claim 9, wherein the physiological chemical comprises at = least one=20 C--H, N--H, or O--H molecular bond.

12. The method of claim 9, = wherein=20 the analyte is glucose.

13. The method of claim 1, wherein the = test=20 subject is a living organism.

14. The method of claim 13, = wherein the=20 test subject is a plant.

15. The method of claim 13, wherein the = test=20 subject is an animal.

16. The method of claim 15, wherein the = animal is=20 non-mammalian.

17. The method of claim 15, wherein the animal is = mammalian.

18. The method of claim 13, wherein the test subject = is a=20 microbial species.

19. The method of claim 13, wherein the test = subject=20 is a human.

20. The method of claim 1, wherein step b) comprises = a=20 principle component analysis.

21. The method of claim 1, further = comprising reporting the net analyte signal calibration spectrum on a = display=20 device.

22. The method of claim 1, further comprising storing = the net=20 analyte signal spectrum on a recordable medium.

23. A method for = non-invasively measuring the concentration of an analyte in a test = subject,=20 comprising: a) identifying a test subject in need of having an analyte=20 concentration measured; b) providing an in vivo net analyte signal = calibration=20 model for the test subject; c) providing an in vivo infrared spectrum of = the=20 test subject; and d) calculating a predicted concentration of the = analyte in the=20 test subject from a data set comprising the net analyte signal = calibration model=20 and the in vivo infrared spectrum of the test subject.

24. The = method of=20 claim 23, wherein the in vivo net analyte signal calibration model of b) = is=20 generated by the method of claim 1.

25. The method of claim 23, = wherein=20 the analyte is a physiological chemical.

26. The method of claim = 25,=20 wherein the analyte is glucose, urea, lactate, triglyceride, total = protein,=20 cholesterol, or ethanol.

27. The method of claim 26, wherein the = analyte=20 is glucose.

28. The method of claim 23, wherein the test subject = is any=20 living organism.

29. The method of claim 28, wherein the test = subject is=20 a plant.

30. The method of claim 28, wherein the test subject is = a=20 mammal.

31. The method of claim 28, wherein the test subject is = a human.=20

32. The method of claim 23, further comprising reporting the = predicted=20 concentration on a display device.

33. The method of claim 23, = further=20 comprising storing the predicted concentration on a recordable medium=20

34. A method for evaluating the analytical significance of an in = vivo=20 multivariate calibration model, comprising: a) providing an in vivo = multivariate=20 calibration vector for an analyte in a test subject; b) providing an in = vivo net=20 analyte signal calibration vector for the test subject; and c) comparing = the in=20 vivo multivariate calibration vector to the in vivo net analyte signal=20 calibration vector for an analytically significant similarity in at = least one=20 spectral feature.

35. The method of claim 34, wherein the in = vivo net=20 analyte signal calibration vector of b) is generated by the method of = claim 1.=20

36. The method of claim 34, wherein the multivariate calibration = vector=20 is generated by a partial least squares regression method.

37. = The=20 method of claim 34, wherein the multivariate calibration vector is = generated by=20 a principle component regression method.

38. The method of claim = 34,=20 wherein the analyte is a physiological chemical.

39. The method = of claim=20 38, wherein the analyte is glucose, urea, lactate, triglyceride, total = protein,=20 cholesterol, or ethanol.

40. The method of claim 39, wherein the = analyte=20 is glucose.

41. The method of claim 34, wherein the test subject = is any=20 living organism.

42. The method of claim 41, wherein the test = subject is=20 a plant.

43. The method of claim 41, wherein the test subject is = a=20 mammal.

44. The method of claim 41, wherein the test subject is = a human.=20

45. The method of claim 34, further comprising quantifying the=20 analytical significance of the similarity in the least one spectral = feature.=20

46. The method of claim 45, further comprising reporting the = quantified=20 analytical significance on a display device.

47. The method of = claim 45,=20 further comprising storing the quantified analytical significance on a=20 recordable medium.=20

Description

FIELD OF THE INVENTION

[0002] The present invention = relates=20 generally to the field of in vivo spectroscopic analysis of a test = subject and=20 more particularly to a method for generating an in vivo net analyte = signal=20 calibration model and methods for the use thereof.

BACKGROUND OF = THE=20 INVENTION

[0003] Diabetes is one of the fastest growing diseases = today.=20 The World Health Organization estimates that 177 million people = worldwide=20 currently have diabetes and this number is projected to increase to more = than=20 370 million people by the year 2030. The costs associated with diabetes, = including premature death, pain and suffering, and increased financial = burdens.=20 These costs are directly related to the medical complications associated = with=20 chronic hyperglycemia. Early detection and maintaining a tight glycemic = control=20 are paramount to controlling the costs of the diabetic epidemic. =

[0004]=20 The cornerstone of tight glycemic control is frequent glucose = monitoring, where=20 blood glucose concentrations are measured to help administer proper = levels of=20 insulin and maintain euglycemic conditions. To this end, glucose sensing = technology has advanced considerably in recent years to provide tools = for home=20 glucose monitoring and establishing opportunities for tight glycemic = control.=20 The current conventional determination of blood glucose is a routine = invasive=20 procedure typically performed several times a day. In general, this = procedure=20 involves the taking of a small blood sample and evaluating the level of = glucose=20 in the sample. Common instruments used for this use the enzyme glucose = oxidase=20 to convert glucose and oxygen to gluconic acid and hydrogen peroxide. = The level=20 of hydrogen peroxide is then measured by spectroscopic or = electrochemical means=20 which is reflective of the starting glucose concentration. =

[0005] While=20 these daily measurements provide a diabetic patient with the ability to=20 self-monitor and thus better control blood glucose levels, they are not = without=20 drawbacks. In particular, the taking of blood samples several times = daily can be=20 very painful and expose the patient to elevated risks of infection. = Moreover,=20 these methods are not suitable for providing continuous blood glucose=20 measurements. Thus, for example, during the night, a patient must either = be=20 awakened periodically for testing or else run the risk that glucose = levels will=20 drop dangerously low while they sleep.

[0006] Non-invasive = optical=20 sensing of an analyte, such as glucose, has been proposed as an approach = for=20 frequent and painless measurement of glucose in diabetics. However, to = date, all=20 reported attempts to measure glucose non-invasively have involved = collecting=20 spectra from a human and then using a classical statistical multivariate = calibration technique to correlate variations in the spectral = information to=20 blood glucose concentrations. These statistical techniques rely on = regressions=20 to statistically correlate spectral variances to an artificially = assigned=20 glucose concentration. Thus, these measurements are not necessarily = based on=20 actual analyte specific spectral features. Further, these statistical = methods=20 fail to provide direct evidence that the assigned concentration = predictions from=20 the multivariate calibration models are actually based on glucose = specific=20 spectral information. Moreover, in some cases the in vivo spectral = signature for=20 a physiological analyte can be smaller than many weakly or partially = correlated=20 spectral variations, making the use of the conventional statistical = methods very=20 difficult.

[0007] Therefore, in view of the foregoing, there = exists a=20 need for an in vivo calibration method that can identify analyte = specific=20 spectral information. Further, there is also a need for a non-invasive = method of=20 measuring the concentration of an analyte in a test subject. Moreover, = there is=20 also a need for a method for evaluating the analytical significance of = the=20 classical statistical multivariate calibration models.

SUMMARY = OF THE=20 INVENTION

[0008] The present invention is based, in part, upon a = method=20 for generating an in vivo net analyte signal calibration model and = methods for=20 the use thereof.

[0009] In a first aspect, the present invention = provides a method for generating a net analyte signal calibration model = for use=20 in detecting an analyte in a test subject. In accordance with this = aspect, the=20 method comprises providing a set of in vivo infrared spectra for a test = subject=20 during a period in which an analyte concentration is essentially = constant;=20 calculating an optimal subspace of spectra that at least substantially = describes=20 all non-analyte dependent spectral variance in the in vivo spectra; = providing a=20 pure component infrared spectrum for the analyte; and calculating a net = analyte=20 signal spectrum from a data set comprising the optimal subspace spectra = of and=20 the pure analyte spectrum. In one aspect, the net analyte signal = spectrum is=20 capable of identifying one or more in vivo spectral features specific to = the=20 analyte.

[0010] In a second aspect, the present invention = provides a=20 method for non-invasively measuring the concentration of an analyte in a = test=20 subject. In accordance with this aspect, the method comprises = identifying a test=20 subject in need of having an analyte concentration measured; providing = an in=20 vivo net analyte signal calibration model for the test subject; = providing an in=20 vivo infrared spectrum of the test subject; and calculating a predicted=20 concentration of the analyte in the test subject from a data set = comprising the=20 net analyte signal calibration model and the in vivo infrared spectrum = of the=20 test subject.

[0011] In a third aspect, the present invention = provides a=20 method for evaluating the analytical significance of an in vivo = multivariate=20 calibration model. According to this aspect, the method comprises = providing an=20 in vivo multivariate calibration spectrum or vector for an analyte in a = test=20 subject; providing an in vivo net analyte signal calibration vector for = the test=20 subject; and comparing the in vivo multivariate calibration vectors to = the in=20 vivo net analyte signal calibration model for an analytically = significant=20 similarity in at least one spectral feature.

[0012] In still = another=20 aspect, the present invention mentioned provides a net analyte signal=20 calibration model produced by the method described above.

[0013] = Additional aspects of the invention will be set forth, in part, in the = detailed=20 description, Figures and Claims which follow, and in part will be = derived from=20 the detailed description, or may be learned by practice of the = invention. It is=20 to be understood that both the foregoing general description and the = following=20 detailed description are exemplary and explanatory only and are not = restrictive=20 of the invention as disclosed.

BRIEF DESCRIPTION OF THE FIGURES=20

[0014] FIG. 1 shows the time profile of arterial glucose = concentration=20 in the test subject of Example 1.

[0015] FIG. 2 shows the = residual=20 tissue spectra of Example 1 after the removal of baseline factors from = selected=20 points along the time profile depicted in FIG. 1.

[0016] FIG. 3 = shows=20 the average residual tissue spectrum of FIG. 2 in comparison with the = pure=20 component spectrum of glucose and the net analyte signal of glucose as=20 calculated in Example 1.

[0017] FIG. 4 shows the predicted = glucose=20 concentrations as calculated according to Example 2. The open circles = represent=20 predictions during the baseline periods, the solid circles represent = predictions=20 derived from non-baseline spectra and the solid line is the time profile = of the=20 arterial blood glucose concentration of FIG. 1.

[0018] FIG. 5 = shows the=20 standard error of cross-validation vs. the number of factors used to = build the=20 partial least squares calibration model of Example 3.

[0019] = FIG. 6=20 shows a calibration spectrum generated by the partial least squares = model of=20 Example 3 compared to the pure component spectrum of glucose. =

[0020]=20 FIG. 7 shows the predicted glucose concentrations using the partial = least=20 squares calibration model of Example 3.

[0021] FIG. 8 shows a = direct=20 comparison of the synthetic partial least squares calibration model of = Example 4=20 with the in vivo partial least squares calibration model of Example 3.=20

[0022] FIG. 9 shows a comparison of the net analyte signal = calibration=20 vector of Example 1 with the partial least squares calibration vectors = of=20 Examples 3 and 4.

DETAILED DESCRIPTION OF THE INVENTION =

[0023]=20 The present invention may be understood more readily by reference to the = following detailed description, examples and figures and their previous = and=20 following description.

[0024] Before the present compositions, = devices,=20 and/or methods are disclosed and described, it is to be understood that = this=20 invention is not limited to the specific articles, devices, and/or = methods=20 disclosed unless otherwise specified, as such may, of course, vary. It = is also=20 to be understood that the terminology used herein is for the purpose of=20 describing particular aspects only and is not intended to be limiting.=20

[0025] As used herein, the singular forms "a," "an" and "the" = include=20 plural referents unless the context clearly dictates otherwise. Thus, = for=20 example, reference to "a test subject" includes aspects having two or = more such=20 test subjects unless the context clearly indicates otherwise. =

[0026]=20 Ranges may be expressed herein as from "about" one particular value, = and/or to=20 "about" another particular value. When such a range is expressed, = another aspect=20 includes from the one particular value and/or to the other particular = value.=20 Similarly, when values are expressed as approximations, by use of the = antecedent=20 "about," it will be understood that the particular value forms another = aspect.=20 It will be further understood that the endpoints of each of the ranges = are=20 significant both in relation to the other endpoint, and independently of = the=20 other endpoint.

[0027] As used herein, the terms "optional" or=20 "optionally" mean that the subsequently described event or circumstance = may or=20 may not occur, and that the description includes instances where said = event or=20 circumstance occurs and instances where it does not.

[0028] As = used=20 herein, the term or phrase "net analyte signal" refers to the residual = spectrum=20 that is orthogonal to non-analyte dependent sources of spectral = variance,=20 including spectral variations related to chemical components within the = sample=20 matrix and to variations induced by the instrumentation or experimental=20 environment. The net analyte signal can be calculated according to any=20 conventional method such as that described in detail in A. Lober, = Analytical=20 Chemistry 59, 1167 (1986), the entire disclosure of which is hereby = incorporated=20 by reference in its entirety for all purposes.

[0029] As used = herein,=20 the term or phrase "net analyte signal calibration model" refers to a = function=20 used to predict analyte concentrations from subsequent sample spectra. = Such=20 models are derived from the net analyte signal, which is scaled to = produce the=20 proper concentration units when multiplied by values from the unknown = spectrum.=20

[0030] As used herein, the term or phrase "substantially = constant" or=20 "essentially constant" concentration refers to less than a 20% change in = concentration, less than a 15% change in concentration; less than a 10% = change=20 in concentration, less than 5% change in concentration, less than a 2% = change in=20 concentration, less than 1% change in concentration, or even a 0% change = in=20 concentration.

[0031] As used herein, the term or phrase = "background=20 spectra" refers to a subspace of spectra that describes at least a = substantial=20 number of baseline factors attributed to non-analyte dependent spectral=20 variations in in vivo spectra of a test subject. To this end, in one = aspect, a=20 substantial number of baseline factors is the number of baseline factors = that=20 account for substantially or essentially all non-analyte dependent = spectral=20 variances. Thus, as used herein, substantially or essentially all can = refer to=20 at least greater than 80% of the non-analyte dependent spectral = variations, at=20 least greater than 85% of the non-analyte dependent spectral variations, = least=20 greater than 90% of the non-analyte dependent spectral variations, at = least=20 greater than 95% of the non-analyte dependent spectral variations, or = even=20 greater than 99% of the non-analyte dependent spectral variations.=20

[0032] As used herein, the term or phrase "pure component = spectra"=20 refers to the spectra or spectrum of a pure analyte. Accordingly, in one = aspect,=20 pure component spectra can be obtained from a sample comprised of the = analyte of=20 interest dissolved in an aqueous buffer solution.

[0033] As used = herein,=20 the term or phrase "analytical significance" refers to the degree, = likelihood,=20 or probability that one or more spectral features in a first spectrum or = set of=20 spectra results from the same or similar spectral feature described in a = second=20 spectrum or set of spectra. In one aspect, the analytical significance = can be=20 evaluated by comparing the degree of overlap between a first and a = second=20 spectrum or set of spectra. To this end, in one aspect, the analytical=20 significance can be quantified by determining the inner product defined = by the=20 overlap of the two spectra. Alternatively, the analytical significance = can be=20 quantified by determining the relative angle between the two spectra.=20

[0034] As used herein, the term or phrase "spectroscopy system" = refers=20 to, in one aspect, a system comprised of conventional components that = can be=20 used to irradiate a test subject with electromagnetic radiation and = subsequently=20 detect variations in the electromagnetic radiation resulting at least = from an=20 interface with the test subject. For example, in one aspect, a = spectroscopy=20 system can comprise a light source for providing electromagnetic = radiation in a=20 desired band of wavelengths and at a desired level of intensity. The = system can=20 further comprise an interface module for bringing the electromagnetic = radiation=20 to a test site of a test subject. Additionally, the spectroscopy system = can=20 further comprise an interferometer, detector, and suitable data = processing=20 capability for determining the spectrum of the electromagnetic radiation = resulting at least in part from the interface of the electromagnetic = radiation=20 with the test subject. As one of ordinary skill in the art will = appreciate upon=20 practicing the invention disclosed herein, other spectrometer designs = could also=20 be used in place of this described exemplary interferometer based = spectroscopy=20 system. For example, spectroscopy systems according to the present = invention can=20 also include detector array multiplex systems and dispersive systems.=20

[0035] In one aspect, an exemplary spectroscopy system according = to the=20 present invention can be configured to operate in the near infrared=20 electromagnetic region, including radiation in the wavenumber range of = from=20 approximately 4000 cm.sup.-1 to approximately 14500 cm.sup.-1. To this = end, the=20 spectroscopy system can be configured to operate in additional = wavenumbers of=20 5000, 5500, 6000, 6500, 7000, 7500, 8000, 8500, 9000, 9500, 10000, = 10500, 11000,=20 11500, 12000, 12500, 13000, 13500 and 14000 cm.sup.-1 and any range = derived from=20 these values. In still another aspect, and for example when used in an = aqueous=20 environment, like the human body, the spectroscopy system can operate in = the=20 so-called combination spectral range of the near infrared spectrum over = a=20 wavenumber range from approximately 4000 cm.sup.-1 to approximately 5000 = cm.sup.-1. As one of ordinary skill in the art will appreciate, spectral = features in the combination spectral range originate from the = combination of=20 stretching and bending vibrational modes associated with C--H, O--H, and = N--H=20 chemical bonds within the molecules in the sample matrix. In still = another=20 aspect, and again for exemplary aqueous samples, the spectroscopy system = can=20 operate in the so-called first overtone spectral region of the near = infrared=20 spectrum over the wavenumber range from approximately 5500 cm.sup.-1 to=20 approximately 6500 cm.sup.-1. Spectral features in this first overtone = spectral=20 range correspond to the first overtone of C--H chemical bonds within = these=20 sample molecules.

[0036] In an alternative aspect, the = spectroscopy=20 system can be configured to operate in the mid infrared electromagnetic = region,=20 including radiation in the wavenumber range from approximately 300 = cm.sup.-1 to=20 approximately 4000 cm.sup.-1. To this end, the spectroscopy system can = be=20 configured to operate in additional sub-ranges within the wavenumber = bands of=20 500, 1000, 1500, 2000, 2500, 3000, and 3500 cm.sup.-1 and any range = derived from=20 these values. It should also be understood that for both near infrared = and mid=20 infrared spectroscopy measurements, it is not required by the invention = that the=20 wavelength range used be a single contiguous range of wavenumbers. For = example,=20 in still another aspect, a plurality of different segments of shorted = wavenumber=20 ranges can be used.

[0037] As one of ordinary skill in the art = will=20 appreciate, the desired configuration of the spectroscopy system will be = dependent on the particular analyte under investigation. For example, in = one=20 aspect where the analyte is glucose, an exemplary spectroscopy system = can=20 comprise a conventional Fourier transform infrared (FTIR) spectrometer=20 configured to operate in the NIR spectral range from approximately 4000=20 cm.sup.-1 to approximately 5000 cm.sup.-1. The system according to this = aspect=20 can further comprise a 50 W tungsten-halogen bulb with an integrated, = gold=20 coated reflector used as a broad-band light source. After collimation, = incident=20 light from the source can enter a Michaelis interferometer and is = modulated=20 accordingly. A custom fiber optic interface can be provided to bring = this=20 modulated light from the interferometer to the test subject and to = couple the=20 light transmitted through the subject to an external detector. A 1 mm = diameter=20 solid-core low-hydroxy silica fiber terminated with a 4 mm diameter = sapphire=20 ball lens can be used for light delivery to the test subject. = Transmitted light=20 through the test subject can then be collected using a second ball lens = into=20 another solid-core fiber. Collected light can then be coupled using an = aspheric=20 lens pair onto a two-stage thermoelectrically-cooled extended-wavelength = InGaAs=20 detector with a 2.6 micron (3846 cm.sup.-1) cutoff. The current output = from the=20 detector can be converted to a voltage signal using a low-noise, = variable gain,=20 transimpedance amplifier, the output of which can be connected to a set = of=20 electronics to record the detector signal by a computer as a function of = time.=20 The corresponding time domain spectrum can then be converted to a = frequency=20 domain spectrum by the well-known Fourier transformation. This frequency = domain=20 spectrum can then be used for all subsequent data processing, such as = building=20 and testing calibration models as described herein.

[0038] In = addition=20 to the exemplary transmission spectrum spectroscopy system described = above, it=20 should also be understood that in another aspect of the instant = invention,=20 analytical special data can be collected as a diffuse reflectance = spectrum or a=20 transflectance spectrum. In still another aspect, it is also possible = for the=20 methods of the instant invention to be used with total attenuated = reflectance=20 spectra and spectral collected by photoacoustic configuration. To this = end, the=20 present invention can be used in connection with any spectral data, = irrespective=20 of the nature and configuration of the spectroscopy system from which it = was=20 obtained.

[0039] As briefly discussed above, in a first aspect, = the=20 present invention provides a method for generating a net analyte signal=20 calibration model. The calibration model can, in one aspect, be used in=20 detecting an analyte in a test subject. The method comprises obtaining a = set of=20 in vivo spectra for a test subject during a period in which an analyte=20 concentration is substantially constant. Background spectra is then = calculated=20 by determining an optimal subspace of spectra that at least = substantially=20 describes all non-analyte dependent spectral variance in the in vivo = spectra.=20 The net analyte signal spectrum of the pure analyte can then be = calculated from=20 a data set comprising the background spectra and the pure component = spectrum of=20 the analyte. To this end, in one aspect, the net analyte signal spectrum = is=20 capable of identifying at least substantially all unique features of the = analyte=20 spectrum compared to the background spectra. This process results in a = unique=20 and characteristic spectral signature of the analyte in the tested in = vivo=20 environment.

[0040] The in vivo spectra of a test subject can be = obtained by either transmission, diffuse reflectance, transflectance, = attenuated=20 total reflectance, or photoacoustic measurement techniques over the = spectral=20 range of interest that favors the selective measurement of the selected = analyte.=20 To this end, in one aspect, a spectroscopy system as described herein = can be=20 used to obtain the in vivo spectra. In still another aspect, the = spectroscopy=20 system can be configured to operate in the near infrared (NIR) and mid = infrared=20 (MIR) regions as described herein.

[0041] In one aspect, = suitable NIR=20 electromagnetic radiation for use in the present invention coincides = with the=20 absorbance bands of the analyte of interest being measured. For example, = if the=20 analyte is glucose, the appropriate NIR bands are located in the regions = of=20 approximately 5000 cm.sup.-1 to approximately 4000 cm.sup.-1 and/or=20 approximately 6500 cm.sup.-1 to approximately 5500 cm.sup.-1. =

[0042]=20 Since the in vivo spectra obtained and used in the method of the present = invention can be either transmittance spectra, reflectance spectra,=20 transflectance, attenuated total reflectance, or photoacoustic, = considerable=20 latitude is also available in the manner and location in which the NIR = radiation=20 impinges on the test subject. For example, if transmitted NIR radiation = is being=20 measured, the NIR radiation can, in one aspect, impinge on a relatively = thin,=20 fleshy area of a test subject, such as, for example, the fleshy webs = between the=20 fingers or toes or the ear lobe of a human test subject. If reflectance = spectra=20 are to be used, the sampling site can, in another aspect, be = characterized by=20 high blood flow close to the surface, such as, for example, the ventral = surface=20 of the wrist.

[0043] The source of NIR radiation used in the = present=20 invention can be such that it is disposed directly against the surface = of the=20 test subject. For example, a small halogen lamp could be used. = Alternatively,=20 the source can be physically remote from the test subject. In the latter = case,=20 it can be advantageous, although not necessary, to guide the NIR = radiation to=20 the desired irradiation sampling site on the surface of the test = subject, for=20 example, by means of optical fibers.

[0044] The data concerning = the=20 transmitted or reflected NIR radiation is, in one aspect, collected = using a=20 detector. The specific nature of the detector is not critical, provided = it is=20 capable of detecting the pertinent wavelengths of light. An example of a = suitable detector for collection of an absorbance spectrum is a = combination of a=20 dispersive element, e.g., a grating or prism, and an optical = multi-channel=20 analyzer sensitive to NIR radiation. For example, in a case where the = data to be=20 collected is a single beam or reflectance spectrum, an absorbance = spectrum or an=20 interferogram, a suitable detector can comprise a combination of an NIR=20 interferometer and a photon counting detector such as a solid state = indium=20 antimonide detector. To this end, it should be understood that there are = many=20 commercially available detectors that can be used, depending on the = exact=20 wavelengths of electromagnetic radiation being measured. For example, in = many=20 cases multiple detectors are available for any particular wavelength = range and=20 detector can be characterized by its detectivity and, ultimately, the = resulting=20 spectral signal-to-noise ratio.

[0045] The positioning of the = detector=20 relative to the test subject will depend both on the nature and size of = the=20 detector and the environment in which the measurement is being taken. = For some=20 purposes, it will be desirable to have the detector physically separated = from=20 the test subject, both because of detector size and to maximize detector = performance by providing the detector with a stable environment. It will = therefore generally be advantageous to guide the transmitted or = reflected NIR=20 radiation to the detector, for example using optical fibers. =

[0046]=20 Depending on the instrumentation selected, the data concerning the = transmitted=20 or reflected radiation is collected as either a single-beam spectrum, a=20 reflectance spectrum, an absorbance spectrum or an interferogram. In = each case,=20 the collected data with essentially no change in analyte concentration = are then=20 used to calculate a set of background spectra or factors. These = background=20 spectra, as described herein, represent a subspace of spectra that at = least=20 substantially describes all non-analyte dependent spectral variances in = the in=20 vivo spectra. In one aspect, an optimal subspace that incorporates the = primary=20 systematic variations in a set of spectra can be obtained by performing = a=20 conventional principle component analysis.

[0047] As one of = skill in the=20 art will appreciate, the primary systematic variations in a set of = spectra can=20 be captured by one or more baseline factors that depend on the = particular sample=20 or test subject under investigation and/or the noise characteristics of = the=20 spectroscopy system. For example, the spectral variations described by = the=20 principle components in a set of spectra can, in one aspect, be due to = tissue=20 state variations induced by the pressure of the clamp (e.g., changes in = water=20 and protein content), variations in test subject physiology, and/or = instrumental=20 drift.

[0048] The background spectra representing the subspace = at least=20 substantially describing all of the spectral variations can, in one = aspect, be=20 calculated by concatenating the set of in vivo spectra as columns of a = matrix:=20 B=3D[b.sub.1 . . . b.sub.n]

[0049] Where B is the matrix of = background=20 spectra b.sub.1 . . . b.sub.n. These spectra are then mean centered = using the=20 average background spectrum, b.sub.avg: B'=3DB-b.sub.avg

[0050] = A singular=20 value decomposition calculation can then be used to obtain a set of = orthogonal=20 spectra or factors that describe the systematic and nonsystematic = variations in=20 the in vivo spectra obtained during a period when the analyte = concentration is=20 held essentially constant. Additional processing calculations can = utilize the=20 most significant factors that characterize the systematic spectral = variations=20 within this set of background spectra.

[0051] Using conventional = mathematical models known to one of ordinary skill in the art, the net = analyte=20 signal for the particular analyte under investigation can then be = obtained by=20 removing all significant background factors, as described above, from = the pure=20 component spectrum of the analyte. This step produces that component of = the=20 analyte pure component spectrum that is orthogonal, or unique, compared = to the=20 non-analyte in vivo spectral variance accounted for within the = background=20 factors.

[0052] For example, and without limitation, the net = analyte=20 signal can in one aspect be calculated according to the following = mathematical=20 model: U orth =3D U - U V V V .times. V Where U.sub.orth represents the = component=20 of vector U that is orthogonal to vector V. In this example, U = represents the=20 pure component spectrum of the analyte and V represents the factors that = encode=20 at least substantially all non-analytical spectral variance in the in = vivo=20 spectra. In this expression U.sub.orth corresponds to the net analyte = signal. To=20 obtain the net analyte signal of the pure component spectrum, the = corresponding=20 values are substituted into this equation. To this end, as indicated by = the=20 Examples appended hereto, the resulting net analyte signal for a = particular=20 analyte under investigation can be used as a calibration model for use = in=20 detecting and/or quantifying the presence of an analyte in a test = subject from=20 in vivo spectra.

[0053] It should be understood that the method = of the=20 present invention is not limited to any one particular analyte or group = of=20 analytes. To this end, in one aspect the analyte can be any = physiological=20 chemical having a functional group and/or chemical bond capable of = providing an=20 identifiable spectral signature or feature when irradiated with NIR or = MIR=20 radiation. In one aspect, the functional group and/or chemical bond can = be C--H,=20 N--H, O--H, or any combination thereof. Specific and non-limiting = examples of=20 suitable analytes for use in connection with the instant invention = include=20 glucose, urea, lactate, triglyceride, protein, cholesterol, and ethanol. = In one=20 aspect, the analyte is glucose. In still another aspect, the analyte is = urea.=20

[0054] The test subject to which the method of the invention can = be=20 applied is similarly not limited to any one particular test subject. To = this=20 end, the test subject can be any living or deceased living organism = containing a=20 minimum concentration of an analyte as described above. For example, in = one=20 aspect, the test subject can be a plant. According to this aspect, in = one=20 example, the methods of the instant invention could be used to = non-invasively=20 detect and/or determine the sugar level within a fruit in order to = assess the=20 ripeness of the fruit and/or its readiness for harvesting. = Alternatively, in=20 another aspect, the test subject can be an animal. The animal can be = mammalian=20 or non-mammalian. The animal can also be a cold-blooded animal, such as = a fish,=20 a reptile, or an amphibian, or the animal can be a warm-blooded animal, = such as=20 a human, a farm animal, a domestic animal, or a laboratory animal, as = further=20 described herein. Further, the test subject can also be a cellular or = microbial=20 species. To this end, in still another aspect, the test subject can = comprise a=20 cluster of cells or microbial species.

[0055] As one of ordinary = skill=20 in the art will appreciate, it is further possible, although not = required, to=20 pre-process and/or manipulate the spectra before implementing the above=20 described procedure. For example and without limitation, by any known = convention=20 method, spectra can be mean-centered, scaled, windowed, weighted, = filtered, or=20 even linearized without adversely impacting the methods of the instant=20 invention. To this end, if desired, any one or more pre-processing or=20 manipulation step set forth above can be used to, for example, improve=20 performance or enhance identification.

[0056] A scaled = calibration=20 spectrum or calibration vector for the net analyte signal calibration = model can=20 be generated by the following equation: .beta. k =3D NAS k NAS k 2 Where = .beta..sub.k is the calibration vector for analyte k and NAS.sub.k is = the net=20 analyte signal vector for analyte k as described above.

[0057] = In a=20 further aspect, the present invention provides a method for = non-invasively=20 predicting the concentration of an analyte in a test subject. According = to this=20 method, the predicted concentration of the analyte can be calculated = from a data=20 set comprising a net analyte signal calibration model for the analyte in = the=20 test subject and the in vivo spectrum of the test subject. =

[0058] In one=20 aspect, the net analyte signal calibration model can be provided, for = example,=20 from a database of net analyte signal calibration models. Alternatively, = the net=20 analyte signal calibration model can be provided by generating the net = analyte=20 signal calibration model in accordance with the methods set forth = herein.=20

[0059] The in vivo spectrum of the test subject can, in one = aspect, be=20 provided from a database or catalogue of previously obtained spectra for = the=20 test subject. Alternatively, the in vivo spectrum can be provided by any = conventional method, including irradiating the test subject with the = appropriate=20 range of wavelengths or wavenumbers of radiation followed by the = appropriate=20 detection as described herein.

[0060] Once the net analyte = signal=20 calibration model and the in vivo spectrum for the test subject have = been=20 provided, the concentration values for an analyte can be empirically = predicted=20 according to the following expression: C k =3D .beta. k A l + C _ b = Where C.sub.k=20 is the concentration of analyte k, .beta..sub.k is the calibration = vector from=20 the net analyte signal, A is the absorbance spectrum measured for the = sample, l=20 is the optical path length for the sample as measured by any known = conventional=20 method, and C.sub.b is the mean concentration of the analyte in the set = of=20 background spectra used to generate the net analyte signal. This same = basic=20 equation can be used in connection with all other types of spectra, = including=20 single-beam spectra and spectra collected in reflectance or = transflectance=20 optical geometries.

[0061] In still another aspect, the present=20 invention provides a method for evaluating the analytical significance = of an in=20 vivo multivariate calibration model. According to this aspect, the = method=20 comprises providing an in vivo multivariate calibration spectrum for an = analyte=20 in a test subject; providing an in vivo net analyte signal calibration = model for=20 the test subject; and then comparing the in vivo multivariate = calibration=20 spectrum to the in vivo net analyte signal calibration model. A positive = comparison of these calibration spectra indicates that the multivariate=20 calibration model is, at least partially, composed of spectral = information=20 originating from the analyte of interest. In an ideal situation, the two = calibration models will be identical, which indicates that the = multivariate=20 calibration is composed only of analyte specific information and no = spurious=20 concentration correlations are responsible for the calibration function = in the=20 case of the statistically based multivariate calibration process. =

[0062]=20 The method for evaluating the analytical significance of an in vivo = multivariate=20 calibration model can be used in connection with a calibration model = generated=20 for any analyte described herein as well as any test subject described = herein.=20 Furthermore, in one aspect, the multivariate calibration model is a = partial=20 least squares calibration model. In still another aspect, the = multivariate=20 calibration model is a principle component regression model. =

[0063] The=20 analytical significance can be determined by any conventional means of = comparing=20 two or more calibration vectors or spectra. For example, in one aspect, = the=20 analytical significance can be evaluated by comparing the degree of = overlap=20 between a first and a second spectrum or set of spectra. To this end, = the=20 analytical significance can be quantified by determining the inner = product=20 defined by the overlap of the two spectra. Alternatively, the analytical = significance can be quantified by determining the relative angle between = the two=20 spectra.

[0064] A further aspect of the present invention = provides a=20 device for performing any one or more methods of the instant invention. = A device=20 according to this aspect can be used on any test subject in connection = with any=20 analyte, as both are described herein. In one aspect, the device is for=20 generating a net analyte signal calibration model. In an alternative = aspect, the=20 device is for measuring the concentration of an analyte in a test = subject. In=20 still another aspect, the device is for evaluating the analytical = significance=20 of of an in vivo multivariate calibration model. It is further = contemplated that=20 a device according to the instant invention can be hand held and/or = otherwise=20 configured for portable use.

[0065] In one aspect, a device = according to=20 the instant invention can be configured for the non-invasive, = quantitative=20 detection of an analyte in a living or deceased living organism. Such an = apparatus can comprise any one or more components of a spectroscopy = system as=20 described herein. For example, in one aspect, the device comprises a = source of=20 electromagnetic radiation for irradiating a test subject with = electromagnetic=20 radiation such that the radiation is transmitted through or reflected = from the=20 exterior surface of the test subject and is available for detection. The = device=20 can further comprise an interface module for bringing the = electromagnetic=20 radiation to a test site of a test subject. Additionally, the device can = also=20 comprise an interferometer, detector, and suitable data processing = capability=20 for determining the spectrum of the electromagnetic radiation resulting = at least=20 in part from the interface of the electromagnetic radiation with the = test=20 subject. To this end, as one of ordinary skill in the art will = appreciate upon=20 practicing the invention disclosed herein, any configuration of a device = can be=20 utilized provided the device contains the components necessary to detect = and=20 collect spectral data in a transmission, diffuse reflectance, = transflectance,=20 attenuated total reflectance, or photoacoustic configuration. =

[0066] In=20 another aspect, the spectral data collected by the device can be = transferred to=20 a computer readable storage medium from whence it can be read by a = computer=20 processor and processed in accordance with a programmed instruction to = perform a=20 desired function or mathematical expression on the recorded data. A = suitable=20 computer processor for use in this application can include any = commercially=20 available microprocessors.

[0067] The computer program used to = process=20 data obtained and stored in the device can be written in any programming = language capable of, for example, performing a Fourier transform. For = example, a=20 program written in Fortran could be employed to perform the data = analysis=20 described in the Examples herein.

[0068] It is further = envisioned that=20 stored data can be recalled by the computer processor and operated under = a=20 second set of programmed instructions in order to apply a predetermined=20 mathematical model to the data. For example, if a net analyte signal = calibration=20 model is generated and stored in the device, it is envisioned that in = another=20 aspect, that data can be recalled by a processor at some later time and = operated=20 under a set of programmed instructions in order to apply a subsequent = and=20 predetermined mathematical model that will calculate, for example, a = predicted=20 analyte concentration.

[0069] A device according to the instant=20 invention can also comprise one or more means for reporting a result of = a=20 calculation or other mathematical expression, such as the calculation of = a=20 predicted analyte concentration. Non-limiting examples of a reporting = means=20 include a digital display panel, transportable read/write magnetic media = such as=20 computer disks and tapes which can be transported to and read on another = machine, and printers such as thermal, laser or ink-jet printers for the = production of a printed report.

[0070] In still another aspect, = the=20 present invention provides the net analyte signal calibration model = produced by=20 the methods described herein.

Experimental

[0071] The = following=20 examples and experimental data are put forth so as to provide those of = ordinary=20 skill in the art with a complete disclosure and description of how, in = one=20 aspect, a net analyte signal calibration model can be generated, used = and/or=20 evaluated. These examples are intended to be purely exemplary of the = invention=20 and are not intended to limit the scope of what is encompassed within = the spirit=20 and scope of the invention. Efforts have been made to ensure accuracy = with=20 respect to numbers (e.g., amounts, temperature, etc.) However, some = errors and=20 deviations should be accounted for.

[0072] Adult male Sprague = Dawley=20 rats were used as an animal test subject in the following examples. = Retired=20 breeder rats weighing more than 400 g were used. The rats were = anesthetized=20 using sodium pentobarbital for the surgical preparation. Anesthesia was=20 maintained during the course of the procedure by administering = chloralose. A=20 catheter was placed in the right femoral vein for the infusion of = glucose=20 saline, and anesthetic during the experiment. The right femoral artery = was=20 cannulated to provide access to the arterial blood stream. =

[0073] The=20 rats were fasted overnight prior to the clamp experiments. After=20 anesthetization, rectal temperature and pulse-oximetry probes were = inserted to=20 monitor animal temperature, pulse rate, and oxygenation during the = cannulation=20 procedure, which was performed on a heated surgical station. After = surgery, the=20 animal was transferred to the spectroscopy station. Animal temperature = was=20 maintained at 38.0.degree. C. using a closed-loop temperature = controller. Pulse=20 rate and oxygenation were monitored continuously. Supplemental oxygen = was=20 provided to the animal at a rate of approximately 1 L/min. =

[0074] Test=20 spectra were collected using a Nicolet Nexus FTIR spectrometer. Spectra = were=20 collected continuously through a skin fold on the rat's back. A 50 W=20 tungsten-halogen bulb with an integrated, gold coated reflector was used = as a=20 broad-band light source. A custom fiber optic interface was used to = bring light=20 from the spectrometer to the animal and to couple the transmitted light = to an=20 external detector. A 1 mm diameter solid-core low-OH silica fiber = terminated=20 with a 4 mm diameter sapphire ball lens was used for light delivery to = the skin=20 fold. Transmitted light was collected using a second ball lens into = another=20 solid-core fiber. The collected light was then coupled using an aspheric = lens=20 pair onto a two-stage thermoelectrically-cooled extended-wavelength = InGaAs=20 detector with a 2.6 micron cutoff. The current output from the detector = was=20 converted to a voltage signal using a low-noise, variable gain, = transimpedance=20 amplifier, the output of which was connected to the external detector = port of=20 the spectrometer. Spectra were recorded as 60 s averages with 16 = cm.sup.-1 (8=20 nm) spectral resolution.

[0075] After the fiber-optic interface = was=20 attached to the animal, spectra were colleted for approximately 3 hours = while=20 the blood glucose levels were held constant. During this time, isotonic = saline=20 was infused into the venous line at a rate of 2 mL/hr, which is the = average=20 fluid intake rate for a rat of the size used in this experiment. After a = sufficient set of baseline spectra were collected, the saline infusion = was=20 replaced with an infusion of a 50% glucose solution at a rate of 2 mL/hr = for 2=20 hours. After that, the glucose infusion was stopped and saline infusion = resumed.=20 The overall infusion rate was constant over the course of the = experiment.=20

[0076] The fiber optic-interface was located across a skin fold = on the=20 animals back, near the shoulders. After initial placement, the interface = was not=20 moved during the experiment, except for a 45 minute period during which = it was=20 repositioned between each scan. The average thickness of the skin fold = during=20 the course of the experiment was approximately 1.5 mm.

[0077] = Blood=20 samples were collected from the arterial cannula at 5-15 minute = intervals.=20 Arterial glucose readings were measured using HemoCue.RTM. Glucose 201 = Analyzer.=20 When blood glucose values exceeded the functional range of the device = (24 mM),=20 the blood samples were diluted with saline. Calibrations were performed = with a=20 set of diluted and non-diluted blood samples within the functional range = of the=20 device to correct for a proportional error due to the reduction of = hematocrit in=20 the diluted samples. FIG. 1 shows the actual arterial glucose = concentration in=20 the rat over the course of the experiment.

EXAMPLE 1 =

Generation=20 of an in vivo Net Analyte Signal Calibration Model for Glucose in a Lab = Rat as=20 Tested Above

[0078] Using the data collected from the = experimental=20 procedure described above, an in vivo net analyte signal calibration = model for=20 glucose in the lab rat was generated. First, the time profile of = arterial=20 glucose concentration in the lab rat over the course of the experiment = is=20 depicted in FIG. 1. The lower, shaded regions on FIG. 1 indicate the = times=20 during the course of the experiment when the glucose concentration was = held=20 substantially constant and the times from which the obtained spectra = were used=20 to form the baseline background spectra (11:00-13:50 and 17:30-18:00). = The=20 latter set of shaded spectra were included in the background calculation = of the=20 baseline factors in order to correct for long term drift in the = spectroscopy=20 system. During the period from 13:00-13:45, the fiber optic clamp was=20 repositioned on the animal between each scan in order to account for = possible=20 variations due to interface placement and tissue state variations in the = baseline spectra.

[0079] A principle component analysis was then = performed on the baseline spectra and it was determined that a set of = nine=20 baseline factors described the majority of the non-glucose systematic = variations=20 in the baseline spectra. This number of factors was chosen based on a = visual=20 inspection of the residuals after removing an increasing number of = factors.=20 These nine factors defined a subspace that contained the substantial = majority of=20 spectral variance observed in the baseline portion of the experiment.=20

[0080] The subspace describing these nine factors was used to = generate=20 background spectra which were then removed from the in vivo spectra = obtained at=20 selected points throughout the experiment, including times when the=20 concentration of glucose was elevated. The residual tissue spectra after = the=20 removal of the nine baseline factors are illustrated in FIG. 2. When = these=20 factors were removed from the spectra for which the animal's glucose=20 concentration was elevated, significant residual structure remained. The = average=20 residual spectrum after the subtraction of the nine baseline factors is = shown as=20 the solid line in FIG. 3. The dashed line in FIG. 3 represents the pure=20 component absorptivity spectrum of glucose. As illustrated, there was = some=20 similarity between the residual tissue spectra and the pure component = spectra of=20 glucose, both having peaks around 4700, 4400 and 4300 cm.sup.-1. =

[0081]=20 The average residual tissue spectrum was then compared to the net = analyte signal=20 of glucose, which was obtained by similarly removing the nine baseline = factors=20 from the pure component spectrum of glucose. The net analyte signal of = glucose=20 is depicted on FIG. 3 as the dash-dot line. As illustrated, there was an = increased similarity between the residual in vivo tissue spectra = obtained during=20 the time when the glucose concentration was elevated and the net analyte = signal=20 of the glucose. This similarity provides direct evidence of glucose = specific=20 spectral information present within the in vivo spectra.

EXAMPLE = 2=20

Use of the Model to Measure a Predicted Concentration =

[0082] The=20 net analyte signal of glucose generated in Example 1 above was used to = calculate=20 predicted concentrations of glucose at selected periods during the = course of the=20 experimental procedure.

[0083] The glucose concentration was = calculated=20 using the following mathematical model: C glucose =3D .beta. glucose A l = + C _ b=20 Where C.sub.glucose is the concentration of glucose in the animal,=20 .beta..sub.glucose is the glucose calibration vector from the net = analyte=20 signal, A is the absorbance spectrum measured for the rat skin, l is the = optical=20 path length, which is approximately 0.7 mm in this experiment, and = C.sub.b(bar)=20 is the mean concentration of glucose in the set of baseline spectra used = to=20 generate the net analyte signal. The average glucose concentration is = ca., 7 mM=20 for the baseline spectra used in this experiment.

[0084] The = results of=20 this calculation are indicated in FIG. 4, wherein the open circles = represent the=20 spectra used in the calculation of the background baseline factors and = the solid=20 circles are predictions for spectra not in the baseline set. =

[0085] The=20 significant scatter indicated in the period from 13:00-13:45 is due to = added=20 spectral variability caused by the repositioning of the optical fiber = interface=20 between scans. The predicted glucose concentrations closely followed the = arterial glucose measurements indicated by the solid line with an = approximate=20 15-20 minute delay that was to be expected due to the time required for = an=20 increased arterial concentration to result in a corresponding increase = in tissue=20 concentration.

EXAMPLE 3

Generation of in vivo PLS = Multivariate=20 Calibration Model

[0086] The net analyte signal of glucose = generated in=20 Example 1 was also used to evaluate the analytical significance of a = classical=20 partial least squares regression calibration model obtained from the = same=20 spectral data obtained in the experimental procedure.

[0087] A = partial=20 least squares regression calibration model was generated from the data = collected=20 in the foregoing experiment. A set of 54 randomly selected spectra were = removed=20 from the data set and set aside for an independent measure of the = standard error=20 of prediction after the calibration was constructed. All remaining = spectra,=20 (approximately 306) were used to build the partial least squares = calibration=20 model. The number of factors was chosen by performing 50 repetitions of = a=20 randomly selected leave-one-third-out cross validation of the = calibration=20 spectra after mean centering. The average standard error of cross = validation=20 (SECV) from the 50 repetitions is shown in FIG. 5. From the SECV, nine = factors=20 were determined as defining the optimal subspace.

[0088] The = resulting=20 PLS calibration spectrum is depicted on FIG. 5 as the solid line, along = with the=20 pure component absorption spectrum for glucose. As illustrated, there = was little=20 high-frequency noise-like structure in the PLS calibration spectrum. = However,=20 there was only a vague similarity between the PLS calibration spectrum = and the=20 pure component spectrum for glucose. More specifically, the PLS model = was=20 generated using a conventional blind regression of factors in order to = establish=20 a correlation with the concentration of glucose in the lab rat. The PLS=20 calibration is indicated as the dashed line in FIG. 5.

[0089] = The=20 results of applying the PLS calibration model to the in vivo spectra are = shown=20 in FIG. 6. The modeled concentrations for spectra used to build the = calibration=20 model are depicted as open circles while those for the independent = prediction=20 set are shown as solid circles. The solid line shows the glucose = concentrations=20 of blood samples taken from the arterial line.

[0090] The = correlation=20 between the modeled concentration and those determined from the arterial = blood=20 samples shows a good agreement with the exception of the period from = 13:00-13:45=20 during which time the optical fiber interface was repositioned between = in vivo=20 spectra. Once again, during the period of elevated glucose, the modeled = glucose=20 concentrations appear to lag behind the arterial samples by = approximately 15-20=20 minutes, which is to be expected due to the time required for arterial = glucose=20 to propagate to the intracellular and intercellular spaces. =

[0091] In=20 analyzing this PLS calibration model, even though there appears to be an = analytically significant correlation between the arterial glucose and = the=20 predicted concentrations obtained by the PLS regression method, it is = not=20 certain that the correlation is a result of glucose spectral information = and not=20 secondary effects or chance correlations.

EXAMPLE 4 =

Generation=20 of Synthetic PLS Multivariate Calibration Spectrum

[0092] A = synthetic=20 partial least squares calibration model was generated in a similar = manner to the=20 in vivo PLS model. First, the synthetic PLS model was constructed from=20 synthesized spectra based upon the regressions of the tissue spectra in = terms of=20 the major tissue components. Each synthesized spectrum was a linear = combination=20 of the spectra associated with water, glucose, collagen, keratin, fat, a = constant spectrum, and a spectrum representing the temperature = dependence of=20 water. The amount of each component was determined by the regression=20 coefficients from the fitted in vivo spectra. The results of the = synthetic PLS=20 calibration model are illustrated in FIG. 7 along with the PLS = calibration=20 spectra generated from the in vivo spectra. As indicated, the two = calibration=20 models are very similar in their dominant spectral features. =

EXAMPLE 5=20

Use of Net Analyte Signal Calibration Vector to Verify the = Analytical=20 Significance of the Statistical Multivariate Calibration Models Obtained = from=20 Partial Least Squares Analysis of in vivo Spectra

[0093] The net = analyte=20 signal calibration vector obtained from Example 1 was directly compared = to the=20 PLS calibration vectors generated in Examples 3 and 4. The results are = indicated=20 on FIG. 8. As illustrated, the dominant features of all three = calibrations=20 analytically agree in shape and amplitude. Thus, the net analyte signal=20 calibration model, which was derived from glucose specific absorption = features=20 present in in vivo spectra, can be used to provide evidence that the PLS = calibration models do in fact correlate to glucose specific spectral = features.=20

[0094] In view of the foregoing, it will be apparent to those = skilled in=20 the art that various modifications and variations can be made in the = present=20 invention without departing from the scope or spirit thereof. As such, = other=20 aspects of the present invention will become apparent to those skilled = in the=20 art from consideration of the instant specification and practice of the=20 invention disclosed herein.

* * * * *

=20

=20 ------=_NextPart_000_0000_01C85786.26B66160 Content-Type: image/gif Content-Transfer-Encoding: base64 Content-Location: http://appft1.uspto.gov/netaicon1/PTO/AppFThdr.gif R0lGODlhhQI8APcAAAAAAAAAMQAAYwAAnAAxAAAxYwAxnABjYwCcYwCcnACczgDOzgD/MTEAADEA YzEAnDExADExMTExYzExnDFjYzFjnDGcnDGczjHOnDHOzjH/Y2MAAGMxAGMxMWMxY2MxnGNjMWNj Y2NjnGNjzmOcY2OcnGOczmPOzmPO/5wxAJwxMZxjAJxjMZxjY5xjnJycMZycY5ycnJyczpzOzpzO /5z/nJz//84xAM4xY845CM5CEM5jAM5jMc5jY85jzs6cAM6cMc6cY86cnM6czs6c/87OMc7OY87O nM7Ozs7O/87/Y87/zs7//9ZCENZCGNZKGNZKIdZSIdZSKdZaMdZjOd5jOd5jQt5rQt5rSt5zSt5z Ut57Wt6EY+eEY+eEa+eMa+eMc+eUc+eUe+echOeljO+ljO+llO+tlO+tnO+1nO+1pe+9re/GtffG tffGvffOvffOxvfWxvfWzvfe1vfn3v8AY/8xY/9jMf9jY/9jnP+cMf+cnP+czv/OMf/OY//OnP/O zv/O///n3v/n5//v5//v7//37//39///Y///nP//zv////////////////////////////////// //////////////////////////////////////////////////////////////////////////// //////////////////////////////////////////////////////////////////////////// //////////////////////////////////////////////////////////////////////////// //////////////////////////////////////////////////////////////////////////// //////////////////////////////////////////////////////////////////////////// /////////////////////////////////////////////////////ywAAAAAhQI8AAAI/gAXCRxI sKDBgwgTKlzIsKHDhxAjSpxIsaLFixgzatzIsaPHjyBDihxJ0uKhRSdLqlzJsqXLlzBjypxJs6bN mzhz6tzJs6fPn0CDCh1KtKjRo0iTKl3KtKnTp1CjSp1KtarVq1izat3KtavXr2DDih1LtqzZs2jT ql2rdQBEt2zjyp37cIDduwjv4i2oV6/DvoD7rgX8NrBhu3R/Bk7MWKbfhI8HHo4MebLgs5P/Wibc GOdmuJ1Dj6RskPJnxApP750IGvLbiq3zLrSsWXXso7exqpaYW7bL3qKBkua7um9KycOJ2z2uHHVz 5wRPC9y9iHN0wdJT6z15eXbpu8yr/lf/rD35eNrPM58fbj19XvLXSS9+vz3+6vXqs7e3f3969uBF meefc/UdBB19wK2HIILq4XcYcv0NCNd/CB4n4GwRQrhZZQemZ5iH8+Fn4IUkUsheieAxyN+HEj4o 4YguatgggEOhCBqJDNmo4ncbthjjfgo6yCKHrV1YHnC27Qhjjz4OCWR3SyJJXZAyYrjckSuGCB+V Mv64JY1C6diiRWK6V5p9FmoYnmRdxvZYkSlOl+NeBcpZm5RX8pgbkGo+h6ZvDpoZJZ4DrEmlaXES maCIEhp6Y6LRaZgec/fVyReYRJUpJJRYKpmllb0Z2SScHXJ5Z3cZLqiqnoOGmiGq/6l6uKKipZoq 6Yv4Gdqcrvbd2umvjTYkKqZBabopp61OtiZ4RprHnp6Iljpsm3FOa6uvzfF4iLOpRvasrE3St223 5BKIl6XAEomteNlG2m6wwka4KLE6GTtqs6o5ihiyvSYrL7PFvRqrnhaea62ASSrH6LvvopelW0Mq zG/Cx9Y1sK8z3ntbtPNSezG9Odk7EKX8egyYvqiJfGyHb9IpcK0whmdwoYXBnG+7yXFrboxqzjyu tC4D/VnBEae7asYrs7rukSGCLNzA1l7bJ0oF8RpwrDoXfejMe36M63O8rgqtYSinhGzWJm/MNdcx jwmicW/DzB+oj2qNNMcW8+y0T/6aLqoy01p/vatAZis96pJ3Rskb1nH+TPPYTUvtbeN2/gn3iQ7D 6/iy1Q6Lb57rpcQdwYSz26/goIpe+d58Z123okbPmfnpDKv7qeGgLhm22LXnzCTqbkuNc8ss/24q 5qTKHTy0tPOOePPIQc96sWfXWT2kuMf75dLAswk8n9NCGbn2hGY/pa1wd1/55I/LZ+K/B3KsfKAr kjy/qO7nL/3q04c8pW1+81p8VEc439XqP8gTWsc0Fjjnca8/lCtgpd7HJPatK4IowRuPeleyhwVG Zh/DH/wcRqH++W8zIERh7vLmqhEyyEsK7Nr8PEizDmZPeuxr4c4sYz8SVopoyf5TkgGDKChaffCG SKwdA3uINBN65jBhIxv5Zki/2/kIcix6Eqf4BLgUnkqGI7QbYYYWt8tYp25BC+O+rrc2N43PII7D XpcAR8dWpdB4TtRJeHZ3ED6uJGx8NNS2IBJIivgxIofkiK6YU8jSuaskiTSkSiK5kUQKcoF5zKQm N8nJTnryk6AMpShHScpSmvKUqEylKlfJyla68pWwjKUsZ0nLWtrylrjMpS6rQkma3OCXwASmDrIg B6gEkwx7C6YygemSZTqzK8EcA1CceQMdTEEMhTBhL2lCB2r+MgdweMocgFlMp3XTmzdoJjp/2ZVz phMo7nSmDso5vUNsUybMHP9ZGH6JhaesAZhaYF0+C3KGZtJhIANdhDShyc6gJHQOW/glFLLJOnsa JaEo+SZBgpmGgqxzEegUCDCvsFFgqgGkIV3IFG7whV/SU6TLzIEUxnASZ8qUpij9ZUELgoZgCmQO XJDCL3VwBTLMwSAc9WhIe5pPoAr1BlHgwlER+kuSUvWXJz0IRgnC1IamwafB3GlJdRoRjKaUCiaF 6Q2kUMyPJiSpY7WqWm9wUnlSAZlvXSZSsarUX3bUIIMQwxRyUE0tnGEOFE2pVhtKkIjeAK9z/etV qdnXG0h2rsF8AhcoOlbLGoQNWtBBNbGQVcwqMy4YPYRGBzLOX76hss+M52n+F3HOHKxhIO407WwT 8k8pHEK0ASWIbIMZ0OEClLZbVetBx0BYZ0L2p8B8bWefmdODkqG5wcwBZGt7W4HkFiHJvepBlYtc xgo3vArBqHEbSgi03qCg54wCIXTr04S09gbSxe03u1ved673BlTI6w3c+dKcGuS++R3IGrALzBzg tL96Ba95BQKHXwZ4IHKILmwjzFoNn5eacu2wawsyBnl297+oNW8huGBhruZADGv16HgHileMlhau BgbpjBn73IOstKAwvkGBB5rhG+RArRhebXiZ+U8jh+GgdCDDE6RgkJ7CeArMGTCSBVJjdjY5B2ow xCHS0Fz+4hi9AyXEO6/+OlY2L3bND9mqjQdSiCgYeQ2ifcJUdbzlRfSYpy+OcWUlO9CBsgEKv/wz koHJg+OYYatoCDSVCfKGb4phvG8uSGn3CmfCrVYgkYbxpBG64zX3ONSCbrNAitzpRaB61AJp7hlO MocxBLfPAtl0WtCZAzaMTAcB/bGEW53PbsK2y8SeMG+rmU06EPbWfW7tkXO8CGnjus0rvcFCWctf gRhCB1s4hLAz3ddFuPeyXwXwZB977ema1byFVna7GSLnhhq7IHMQ7VDDSe6F/DbY7z12nxPKhl9O Ydg5ZTdts5BaYC9i3AJxr6JVfW8B9/vfDw94v6vmcIhH+5dRKMi3AS7/1kU0Nw2cLTdy50JNHVDB DQSBcRwWUXAdpNzN7u40MMGQaGrjXKU34MJAtODSnP8yuN5EurwN3Fw+ypzmzN44kptOEEN8euc9 RzN1fx7vVs97IfXeLUHY0FyhI5whT695yrHO7q4TRLVGRjgwsZCDg2Kh0hNOe9Rj/cuUtDSYpa6v xas21IHo3eZSN/wNZq52o+uA34pn/N4XQXRhUsEL3d46W9DrXQb/0uwqn+4TOL0Inj827A1psjOh 7cwohCElrX/915kJzN05e5mgV3Wba1/1q7PT9GTQute5ztglLz29ylbm6BHi3l6fXSG3V2bumQl8 t+u+3MDsJhXUQNKE/kY/mKCnesKLn1l6K7vC6qat54OeePXj3ug5QINw1w/6MXOBClHALuiVTxfO L8KxypQDMxd6P6dylYd6DJFt1ERPwjd4pMcDv7RtCAGA2TWA14dk58ZVLbZlB3h8nMd7i6Bm5Dd8 /pd4/vcFOhB8RrZnF3gQFNhgFphPHbhmBGdwcsdOMJYDbWVeL/hNA3gFfMVn1tcQW1V5eNWDRmaB BSgQSCiAOFdkOtBYNqWEBPFlS5hiwycQcXADW1B1onVhLahe7zRQHUh6C/FPOcCClHd0V/h1Pndv zKQG3/RkKAEHX3AFU7WFXTgQ35Z+YehlQ4UGh2AIaKBvZsZYZfh8/wfxVGIwiCwGVT8Hh+ZVcQ5Y gBUXfPKXbVEwCGaYEHrohX5IhoI3UG/wVBK4e+xUCA5HbZ/Ih18oEAVXTWeQTYUghyNYEJR4fRD1 S0+QTa3oba9IgFrIhaAIhkT2aRW2h8CYfmcwBnJAi2fge52Wi2eRfQmBVt22CI9YctX1YZTVjSjh XjeghtaoECuVe6sGTL4GjpyGad4YYdZYYsukA6WFjQWxjX3ljt14XTH1XOV4COKohuKVEOm2THiF YigGdr/kjv/1Tv8UhYuAdwCmj+XIfDeQjfgIjgAJTEf1X1dgKLlljWpQThwZcRd5j2TlZ+tnjQ15 EA2pA1NljwSRkf8DSRAyORD4+F+3dpMCgY8KqEwHqVhooYJLJ49Q8HYHiGldBWcfZYuMpXpwtpQK YXpaRhCslk5O6XVZ+WbPJJWLIAdBZWQ8QAbHYZRIWZFeCWr15VS/JAVSRRBbKRBQeRBEmYU0F1pD lQU35k3jJ3YHIY/wFlL8SIOC52qFSWISdZYLGZeLMJfylAXyN2xcWJh1qVCJOTJJuWpBRVg5EAVa QAb85lYbNlRUMAaGIBBmiZloWZipSThJSU1RYJqoeZmumX2FQAZZ8AQgtwX5JZq79JvAGZzCOZzE WZzGeZzImZzKuZzM2ZzO+ZzQGZ3SOZ3UWZ3WeZ3YmZ3auZ3c2Z0M3vmd4Bme4jmeoBQQADs= ------=_NextPart_000_0000_01C85786.26B66160 Content-Type: image/gif Content-Transfer-Encoding: base64 Content-Location: http://appft1.uspto.gov/netaicon1/PTO/help.gif R0lGODdhPwAYAPIAAAAAADMzmQAA/8zM/////wAAAAAAAAAAACwAAAAAPwAYAAADqEi63P4wSjmq vTjrzXsmA+iNZLkFoqCubOu+cCyvAVrNeK7Ddbr/wFjvFiwahwOjEogMKlRPV3QpaAIVAAFWSshS rb+tOIplHG3JawOAZWu7ZUIRvGvHF2y43vuj68Z5AIKBcXw7fjmAYm+EXUGIOIBvco1tj2hFgiqa AoObgZxMmFQyW2c+pDGhl6ipriyQr6SxsktINbi5uru8vb6/uQA2wMTFxr8ACQA7 ------=_NextPart_000_0000_01C85786.26B66160 Content-Type: image/gif Content-Transfer-Encoding: base64 Content-Location: http://appft1.uspto.gov/netaicon1/PTO/home.gif R0lGODdhPwAYAMQAAAAAADMzmQAA/8zM/////wAAAAAAAAAAAAAAAAAAAKAAAAB/upwAAFgAW7gA W/9/uFkPuGk66gjYojAwKy8AGQBcOgAAAAAAAAAAAAAAAAAXAAAAAAAAAAAAAEROSQBTVywAAAAA PwAYAEAFvCAhjmRpnmiqpsMgvHAsC+Jbz3iuy0FAtDscAPAaFo0Co5KYRAZfvZ/rSa1aYVGgVTTk egmAbxhcYj6z06t6zfMNpK24fE6v2+/4uDvN7l/RV1w0YIMENiYwI397gYRij2SRRIZVgH6XlXs9 m5ydnp+goaKcAD5weaipqnZSmK5Blq+yba2ztlCMW46Et7RaVSZjY4PCNVxhVrFUgpDFksdOsLnA u5PHzYVJmbW9ssrdmN/gflmj5ufoogAhADs= ------=_NextPart_000_0000_01C85786.26B66160 Content-Type: image/gif Content-Transfer-Encoding: base64 Content-Location: http://appft1.uspto.gov/netaicon1/PTO/boolean.gif R0lGODdhPwAYAPIAAAAAADMzmQAA/8zM/////wAAAAAAAAAAACwAAAAAPwAYAAADwEi63P4wSjmq vTjrzXsmA+iNZLkFoqCubOu+cCyvAVrNeK7Ddbr/wFjvFiwahwOjEoiMNXSKpasJUwAAUZxVyqK+ tuCsYKEKE1bkcVpoS1YZV2tUfqbHV9g6IS/zuqx8doJ7DQBqg4Y8bU57anmPiIBxhJSPbD5fjZGV dmOOiJdEmQtXh6WklmSGcJ2KmC9XsXilKrICtLa3sqW0L35cwC2/wcTDxMDGx1LJykpINdDR0tPU 1dbX0QA22Nzd3tcACQA7 ------=_NextPart_000_0000_01C85786.26B66160 Content-Type: image/gif Content-Transfer-Encoding: base64 Content-Location: http://appft1.uspto.gov/netaicon1/PTO/manual.gif R0lGODdhPwAYAPcAAAAAAAAAIQAAUgAAnAAAtQAAvQAAxgAAzgAA1gAA3gAA5wAA7wAA9wAA/wAI xgAIzgAI/wgArQgAxggAzggA1ggA5wgA9wgA/wgIpQgItQgIvQgIxggIzggI3ggI5wgI/xAAxhAI vRAI1hgQpRgYjCEYziEhlCEhxiEpnCEprSEptSEpvSExlCExnCkhrSkphCkprSkpvSkpxikpzikp 1ikxlCkxnCkxpSkxtSkxvSkxxikxzjEpnDEpvTExezExlDExnDExpTExrTExtTExvTExxjEx1jE5 pTE5rTE51jk5ezk5pTk5rTk5tTk51kI550JChEJClL21/729/73G/8a9/8bG78bG/8bO58bO98bO /87G/87O987O/87W987W/9bO/9bW/9be/97e/+fe/+/v7+/v//f3//f////3////5///9/////// //////////////////////////////////////////////////////////////////////////// //////////////////////////////////////////////////////////////////////////// //////////////////////////////////////////////////////////////////////////// //////////////////////////////////////////////////////////////////////////// //////////////////////////////////////////////////////////////////////////// //////////////////////////////////////////////////////////////////////////// //////////////////////////////////////////////////////////////////////////// /////////////////////////////////////////////////////ywAAAAAPwAYAAAI/gDVrDmT 5gyagwjToFGYEOFChwwPRnzYEOLCNBgdHiS4pgyZLleoWMmipUvJKyZTainJkqVJlytVwoSZJWVK ki+7WDkT5owUkFe0XJEytKjRo0RDJjVKZanTKyGVFk06BWoXKVHSXB2RoWvXDSHCig2rIQPYr2ND aBC7lixbthlCxAULIgRYuxEGNCGB5sqWCBMqVEggWLCCBIgVKKiweDFjxoobK4YsuTLjxIULJ1Ag YoMTIWlKZqhgoYHp0wtSN1iw2jTr1qxjw54t+3Xq16cbfGCtYMOJJmiyXBmtOrfx48iTK1+u2wMH HUzSCM+g4MNxjLmxM99uHAJ23KZ7/uuwoXV4heMLMAI4rZ67e9PtZTdQwEHGkfIhzudmjTEN/Pbv cadeceHVB9pV+SHX33oL/qddfw/610CEEQKg2mv00dAEfvrJByGE660GAADYLVhiGgyiaNqIJ66H 2wIVcDDDEFpJsUEHBDr4YYg7qkciiu0FCSKAtvUmw4YghZCAbewB6WSKLD45oZTq+Vilivv1VgRo QimZ45Q/hgkmmEKGaaKKV4boWgMJbKCDEGh0QYWXxqWnYpQ8Donmng32iCVqHrgZBBpdEoZaagyM CJuirI3oaAOMRuqoi5Oup6hpEPCmQRImxDlFCAgsAAFqDJQ6amuwZUrbaqOypqp83OChB8EHbULX VxfEoRdgcrHut9wCvRkBQxpUXBGBBxYwKd+aqNXpK3Kx1tZaphWIcMIQaIQhBagW7Gbabsnuumuy 3ZILY2q9PeGCGVpsMYIGCEyQQAcIcGAvBx1MkO+++vbL77/+BrxvvBNMQHC9HeTLQQZD8CBdFzYM oUMMROQwhMVD9KDDxBfn0EQMFucQ8sgdlyyyyR0XoUMRPRQhcgpFxAADEwJAkQYWPvwQhA1H1GCD DT8EDUQQQQBRww81HK000kwv7XTTUB9BNNE2AAEEEjYEscQLSgQQEAAAOw== ------=_NextPart_000_0000_01C85786.26B66160 Content-Type: image/gif Content-Transfer-Encoding: base64 Content-Location: http://appft1.uspto.gov/netaicon1/PTO/number.gif R0lGODdhPwAYAPIAAAAAADMzmQAA/8zM/////wAAAAAAAAAAACwAAAAAPwAYAAADvUi63P4wSjmq vTjrzXsmA+iNZLkFoqCubOu+cCyvAVrNeK7Ddbr/wFjvFiwahwOjEoiUKVTPXXTZjCkAgquUgKXa ktbnVUzIOqCLVdo8fVVhC8CYO5fTy3MzQK7fC79Od3d2hHWGDIQzby9zg4eFjlx7iTKLLlqPmZFY eFyKgGGcnoiaV6JllH8+MX4CrZNYfrJdkyu1rl2qREu8n6u9wC6WwcTDxMDGx7xINc3Oz9DR0tPU zgA21dna29QACQA7 ------=_NextPart_000_0000_01C85786.26B66160 Content-Type: image/gif Content-Transfer-Encoding: base64 Content-Location: http://appft1.uspto.gov/netaicon1/PTO/ptdl.gif R0lGODdhPwAYALMAAAAAADMzmQAA/8zM/////wAAAAAAAAAAAD8AW0QAAAC/9wMAAHoAAFgAWzUA WwQQACwAAAAAPwAYAEAExZDISau9OOs8hvhg+FUjSQqSqK5iEBAdKwLASoN0/uU6L4cumOdHLBqJ wRixckMRas6aBDAFUQTNX3J47HqRr4GwQy6bz+i0ek0Oc7/wuGBbrFajJenzXkoZ6UtPKnd2e4Il WHwygHKNXVsukZKTlJWWl5iSAC9jbJ6foGlCjqR/bqWoi6eprECrP1NUBDuxsjQpinCMMoVQhIJT fLF+YKOBPnnJwYe0E6bGsMxWh0yyVxRZqtCtrLvcqN7fpEmZ5ebnmAARADs= ------=_NextPart_000_0000_01C85786.26B66160 Content-Type: image/gif Content-Transfer-Encoding: base64 Content-Location: http://appft1.uspto.gov/netaicon1/PTO/hitlist.gif R0lGODdhPwAYAPcAAAAAAAAAIQAAUgAAnAAAtQAAvQAAxgAAzgAA1gAA3gAA5wAA7wAA9wAA/wAI xgAIzgAI/wgArQgAxggAzggA1ggA5wgA9wgA/wgIpQgItQgIvQgIxggIzggI3ggI5wgI/xAAxhAI vRAI1hgQpRgYjCEYziEhlCEhxiEpnCEprSEptSEpvSExlCExnCkhrSkphCkprSkpvSkpxikpzikp 1ikxlCkxnCkxpSkxtSkxvSkxxikxzjEpnDEpvTExezExlDExnDExpTExrTExtTExvTExxjEx1jE5 pTE5rTE51jk5ezk5pTk5rTk5tTk51kI550JChEJClL21/729/73G/8a9/8bG78bG/8bO58bO98bO /87G/87O987O/87W987W/9bO/9bW/9be/97e/+fe/+/v7+/v//f3//f////3////5///9/////// //////////////////////////////////////////////////////////////////////////// //////////////////////////////////////////////////////////////////////////// //////////////////////////////////////////////////////////////////////////// //////////////////////////////////////////////////////////////////////////// //////////////////////////////////////////////////////////////////////////// //////////////////////////////////////////////////////////////////////////// //////////////////////////////////////////////////////////////////////////// /////////////////////////////////////////////////////ywAAAAAPwAYAAAI/gDVrDmT 5gyagwjToFGYEOFChwwPRnzYEOLCNBgdHiS4pgyZLleoWMmipUvJKyZTainJkqVJlytVwoSZJWVK ki+7WDkT5owUkFe0XJEytKjRo0RDJjVKZanTKyGVFk06BWoXKVHSXB2RoWvXDSHCig2rIQPYr2ND aBC7lixbthlCxAULIgRYuxEGNCGB5sqWCBMqVEggWLCCBIgVKKiweDFjxoobK4YsuTLjxIULJ1Ag YoMTIWlKZqhgoYHp0wtSN1iw2jTr1qxjw54t+3Xq16cbfGCtYMOJJmiyXBmtOrfx48iTK1+u2wMH HUzSCM+g4MNx3MyzL4fg+nhvHTa0/g6vcF27ee21TSvgIOOI+BDkc7/GeJp+A/vI8ZvWf122evag XQVfchgBsF8aBjYAAAD24VYgagom2Np9aczmWgUc0NDEe/H5R2GCD9aHIIQhYjfhgQbedtp6Mwyh lRQbdFAcag8u8CBGNjaI4oQ4UlhhjhXKt0BvMwR4RQgJ2JYbRkzeiGCBEn5I45NUfpigbb0VAZpQ SM7oGpQMjlhgiK2RKSV9C5rp2ndCoNEFFV0ax1qIToYZpZQigvijmqt5sIEOQaDBJWGo3VZjnTWW OWJsY6bBAH1mQsCbBkmY4OYUIXDQAHeuSXoogjka2COjTVboJIKiBtnpBwn8yURf1V1QZ+JrC55W a4Sm3WrrgrziGqGEuq65gREwpEHFFRF4YIGSJ57nLG0TSlqBCCcMgUYYUmS6wG6mLWtBfM+et+wH Fni7gGANtPqEC2ZosQUGmSI2QQcTcFBvAhPgS+++8/bL77/+BtwBBwQXXHAH9G6QQQ88SNdFEEPo EMMQORCBQxM49JBDixYT0XHHOHgcMsgil/wxDjHkoEMRK8fQAxEqSAwDEwJAkQYWPvwAhA1MsGDD D0ADLUQQPwdt9NFIJ330EUfcEMTTNtiwRAtHLPGCEgEEBAA7 ------=_NextPart_000_0000_01C85786.26B66160 Content-Type: image/gif Content-Transfer-Encoding: base64 Content-Location: http://appft1.uspto.gov/netaicon1/PTO/bottom.gif R0lGODdhPwAYAPcAAAAAAAAAIQAAUgAAnAAAtQAAvQAAxgAAzgAA1gAA3gAA5wAA7wAA9wAA/wAI xgAIzgAI/wgArQgAxggAzggA1ggA5wgA9wgA/wgIpQgItQgIvQgIxggIzggI3ggI5wgI/xAAxhAI vRAI1hgQpRgYjCEYziEhlCEhxiEpnCEprSEptSEpvSExlCExnCkhrSkphCkprSkpvSkpxikpzikp 1ikxlCkxnCkxpSkxtSkxvSkxxikxzjEpnDEpvTExezExlDExnDExpTExrTExtTExvTExxjEx1jE5 pTE5rTE51jk5ezk5pTk5rTk5tTk51kI550JChEJClL21/729/73G/8a9/8bG78bG/8bO58bO98bO /87G/87O987O/87W987W/9bO/9bW/9be/97e/+fe/+/v7+/v//f3//f////3////5///9/////// //////////////////////////////////////////////////////////////////////////// //////////////////////////////////////////////////////////////////////////// //////////////////////////////////////////////////////////////////////////// //////////////////////////////////////////////////////////////////////////// //////////////////////////////////////////////////////////////////////////// //////////////////////////////////////////////////////////////////////////// //////////////////////////////////////////////////////////////////////////// /////////////////////////////////////////////////////ywAAAAAPwAYAAAI/gDVrDmT 5gyagwjToFGYEOFChwwPRnzYEOLCNBgdHiS4pgyZLleoWMmipUvJKyZTainJkqVJlytVwoSZJWVK ki+7WDkT5owUkFe0XJEytKjRo0RDJjVKZanTKyGVFk06BWoXKVHSXB2RoWvXDSHCig2rIQPYr2ND aBC7lixbthlCxAULIgRYuxEGNCGB5sqWCBMqVEggWLCCBIgVKKiweDFjxoobK4YsuTLjxIULJ1Ag YoMTIWlKZqhgoYHp0wtSN1iw2jTr1qxjw54t+3Xq16cbfGCtYMOJJmiyXBmtOrfx48iTK1+u2wMH HUzSCM+g4MNx3MyzL4fg+nhvHTa0/g6vcL0BxvPa0ydX/VoBBxlHxIcgn/s1RgAAMCrXbz6NetSy meaeDKBdNV9y9/UHAGqoJZhgdxC2Vp+Eq1XAAQ1NyEdfgAp22B9/6KG3wHn+fUiia7edZuEMQ2gl xQYdFNcgifgpmJ9/992IX4763edjifUt0JsMGYIUQgK25cZjiQ6moWOHTer444IQCrlBEaAJdaSM rkUJpZNeRjnlg/V9JwQaXVCxpXGshemkgm4uOCaYb7LpwQY6BIGGloQBmBqJTi444nk1YtRmGocu KWWdpkHAmwZJmIDmFCFw0AB3rjkKAX6cwsapoJsKWmOoNS5Q46lUZvpBAngy0VcXzdRhR+F/bCon 63W9GQFDGlRcEYEHFiQ5W4QR4nYrdsjSdukCFYhwwhBohCFFpQvsZlqwwdJKK7YfYMtsar094YIZ WmyBQaWITdABB+xykEAH6sYL77zy1kvvvfEmMMG+7bLbAbwcZDAED9J1EcQQOsQwRA5E4DDEEDEQ oQMREeNARMMXW4zxxhp3nHHGC+dQhA49JEyECkXgAAMTAkCRBhY+/ACEDUywYMMPOOMsRBA15Ozz z0AH/bMNRwRRdBBB2GDDEi0EscQLSgQQEAAAOw== ------=_NextPart_000_0000_01C85786.26B66160 Content-Type: image/gif Content-Transfer-Encoding: base64 Content-Location: http://appft1.uspto.gov/netaicon1/PTO/cart.gif R0lGODlhfgAYAPcAAAAAAAAAIQAAUgAAnAAAtQAAvQAAxgAAzgAA1gAA3gAA5wAA7wAA9wAA/wAI xgAIzgAI/wgArQgAxggAzggA1ggA5wgA9wgA/wgIpQgItQgIvQgIxggIzggI3ggI5wgI/xAAxhAI vRAI1hgQpRgYjCEYziEhlCEhxiEpnCEprSEptSEpvSExlCExnCkhrSkphCkprSkpvSkpxikpzikp 1ikxlCkxnCkxpSkxtSkxvSkxxikxzjEpnDEpvTExezExlDExnDExpTExrTExtTExvTExxjEx1jE5 pTE5rTE51jk5ezk5pTk5rTk5tTk51kI550JChEJClL21/729/73G/8a9/8bG78bG/8bO58bO98bO /87G/87O987O/87W987W/9bO/9bW/9be/97e/+fe/+/v7+/v//f3//f////3////5///9/////// //////////////////////////////////////////////////////////////////////////// //////////////////////////////////////////////////////////////////////////// //////////////////////////////////////////////////////////////////////////// //////////////////////////////////////////////////////////////////////////// //////////////////////////////////////////////////////////////////////////// //////////////////////////////////////////////////////////////////////////// //////////////////////////////////////////////////////////////////////////// /////////////////////////////////////////////////////ywAAAAAfgAYAEAI/gDVqFmz 5syZNGkMolnIsOFChA/TOGQIcSKaihMxOtTYkCNFiRYjRkRIEmTIhQZLduly5UqDlzBjypxJs6bN mzhz6tzJc6EWLS1tsok5tEFRozCPHk1K9KVSp0yNFp06FSpSq1RnPsVa1arMhS23bOFJtqzZs2h5 IlwpRQrNrV2Rbr3qlWhXuHfvOtUrVebSplzrwlwYJkzbtIgTK14s02Bhg21Xtmz5s2Vby1IwT75y eTPnzJ47t6RCRfRk06RNaw4NejQV151jt74yZcpmtm4bLMySJSjj38CD21zb5bDf44iXxhXOcy7O vy8RVnb5li5VNtihDt1+M653r9+v/sKNSl4r9vBf0bRcyfw3dOg74TNGyJt9+/v4yQoQ8JJgmTJk kCHZFaRZYQVvP62UIEtXKKiFgw4uKOFPD3Yx4YQRVnihhhxiaKEW9X24EoRdINhhhQYSZ1x+LLZ4 k3QVuijjjDDR1xt1NOaYH3G+AcaUfPGdV15iQOYkZHw08Yijj95lZ115ymmnnJNyPUVlVVT29deW SV0pVZbRpTFgdViBp51c5nlZ5priRRWec3QJZeZMC61EWk3csfkjmj62yeabbs4JJ5dPFpreSrXJ Sd6Rc/rF6FyP9iVloODl6eiWeSoFHUKk9aijkVqRdV6Riplhxk9ifaoqcwhhgYUPDD6sKmtiSigR QAABAQA7 ------=_NextPart_000_0000_01C85786.26B66160 Content-Type: image/gif Content-Transfer-Encoding: base64 Content-Location: http://appft1.uspto.gov/netaicon1/PTO/order.gif R0lGODlhfgAYAPcAAAAAAAAAIQAAUgAAnAAAtQAAvQAAxgAAzgAA1gAA3gAA5wAA7wAA9wAA/wAI xgAIzgAI/wgArQgAxggAzggA1ggA5wgA9wgA/wgIpQgItQgIvQgIxggIzggI3ggI5wgI/xAAxhAI vRAI1hgQpRgYjCEYziEhlCEhxiEpnCEprSEptSEpvSExlCExnCkhrSkphCkprSkpvSkpxikpzikp 1ikxlCkxnCkxpSkxtSkxvSkxxikxzjEpnDEpvTExezExlDExnDExpTExrTExtTExvTExxjEx1jE5 pTE5rTE51jk5ezk5pTk5rTk5tTk51kI550JChEJClL21/729/73G/8a9/8bG78bG/8bO58bO98bO /87G/87O987O/87W987W/9bO/9bW/9be/97e/+fe/+/v7+/v//f3//f////3////5///9/////// //////////////////////////////////////////////////////////////////////////// //////////////////////////////////////////////////////////////////////////// //////////////////////////////////////////////////////////////////////////// //////////////////////////////////////////////////////////////////////////// //////////////////////////////////////////////////////////////////////////// //////////////////////////////////////////////////////////////////////////// //////////////////////////////////////////////////////////////////////////// /////////////////////////////////////////////////////ywAAAAAfgAYAEAI/gDVqFmz 5syZNGkMolnIsOFChA/TOGQIcSKaihMxOtTYkCNFiRYjRkRIEmTIhQZLduly5UqGDBUqNJhJs6bN mzhz6tzJs6dPmwuCLpgZtKYCBRw4yJBx5MhCLVpahgjBJkHMmWwaZMXKtSvNrVu12gwrtmzZsGDB YlV79mtXNmS9ooUAAS5WBRo0sEliwsTCllu2RIgwYULMmAmsHl58NPHPx5BpxjxKWcHkypYrNE6c 4Ojiz4cTHxUhYsMGJ06ECEG4UooU0x06sFkw+y1buzXZmvWa+7Zbrbd9A3+7dmzOrbSDwjXNRkaT JgvDhHE9NWnQDx8ia9/OvXuDoDGF/i44avrJExcuDEo36Hply5ZQW7qWL4X++yvz7+Ovrz9/SypU +PeegAAKaF9//P1HhYL5NZjgFVNMcV9rUkQRxUJZZNHSS+EJ5d2HIIaYE3YeeJCUDjowwQRrXVAX QkxZxcXbjDLeRJZuIm6HVlFALRBTUmzQ8BxC8V0xVQJwJSnXkrkxeRyTaEH5m25UGndcksopwJwO qv11xUovHcVjjjnKWKNPZ/IUFHkbGGEEDDAglOFKQQQxxBAoxhADmXz2qRMRRKigQhFF4IADnCoK IAAUUBBURhlkkOHeFQBaYUWGUK2UKUtfdrHpp1BpoamonoY66qigokoqqKmW2ipU/nOWutKpXWC6 qqldWMqiayOM8NKvwGZg2lTEFkssmXkZG0KyxjJbrLPE5vXSVNOGMGwIIIAw1bXbbjDYAAM8RwIJ RJLKYQUWWNDTmH62S5R4Q/GEHY9snnDCkGlkuGEGR2Hn7r/b0VUTuzmxiaINNrAo1YsynWmmlTwl GaWIafak5Ls3HQWkDKopbCTDMX43lF1ohfwVbhc3OeVwJ0eJm1gvq/XycHHVLHLKbFQg8Z0eHznb yFLS6GTLE7dF3NFGT1z0jBhjlSVzRXSJxkoA+lxUlWbZnDXERneNtdJB7yY2lcp5sGWdC60U4VRs cDATXWnFnHJxwNUo8W9Zx8wyZ8xgy2XyWDMDVxd2VTHHhooIAdjSYCWmSzDAPz382N0+8UgXeBWQ Zu+dZpgBVWAYYFAdB45Bbrp3ScUWW2FJvXQnDzwghAUWPvjwww9AAIGwiqf3vh3CSyzRQgt1Bv/C C0ooEUAAAQEAOw== ------=_NextPart_000_0000_01C85786.26B66160 Content-Type: image/gif Content-Transfer-Encoding: base64 Content-Location: http://appft1.uspto.gov/netaicon1/PTO/image.gif R0lGODdhPwAYAPcAAAAAAAAAIQAAUgAAnAAAtQAAvQAAxgAAzgAA1gAA3gAA5wAA7wAA9wAA/wAI xgAIzgAI/wgArQgAxggAzggA1ggA5wgA9wgA/wgIpQgItQgIvQgIxggIzggI3ggI5wgI/xAAxhAI vRAI1hgQpRgYjCEYziEhlCEhxiEpnCEprSEptSEpvSExlCExnCkhrSkphCkprSkpvSkpxikpzikp 1ikxlCkxnCkxpSkxtSkxvSkxxikxzjEpnDEpvTExezExlDExnDExpTExrTExtTExvTExxjEx1jE5 pTE5rTE51jk5ezk5pTk5rTk5tTk51kI550JChEJClL21/729/73G/8a9/8bG78bG/8bO58bO98bO /87G/87O987O/87W987W/9bO/9bW/9be/97e/+fe/+/v7+/v//f3//f////3////5///9/////// //////////////////////////////////////////////////////////////////////////// //////////////////////////////////////////////////////////////////////////// //////////////////////////////////////////////////////////////////////////// //////////////////////////////////////////////////////////////////////////// //////////////////////////////////////////////////////////////////////////// //////////////////////////////////////////////////////////////////////////// //////////////////////////////////////////////////////////////////////////// /////////////////////////////////////////////////////ywAAAAAPwAYAAAI/gDVrDmT 5gyagwjToFGYEOFChwwPRnzYEOLCNBgdHiS4pgyZLleoWMmipUvJKyZTainJkqVJlytVwoSZJWVK ki+7WDkT5owUkFe0XJEytKjRo0RDJjVKZanTKyGVFk06BWoXKVHSXB2RoWvXDSHCim1AtqxZsxrE pg27NkRbDRlCxAULIgRYuxEGNCGB5sqWCBMqVEggWLCCBGcTK05cQYHjxo8PJ1BQuPBkERucCElT MkMFC4kXLB5NWvGC04k/iFaw4UQTNFmueEZdurbt2x88cNDBJE3sDAo+3B5+GwJZ0WdZ67ChVXYF 4tCHo0augIOMI81DPD+LMbr3s8gb/lSXsfmq9sTdv38Pv6ACBxpNsm8Pn1699/DuZwzRKmVDB9pk pYfRgPURWBaBaQQ44IH1gbcAazLEB1IICYSn4IXdNdhAhglumKCAH4YI3nGsFbGZUBQC6CGGInq4 IIguLrjiaA9uoIMQaHRBRYrcdcjhijAG2WGPNHpgYxBooEgYkUCKKCSLUGrYAASraZCECTlOEQIH U5r1JIgyIqhggVI2sAAEHyRgIxN9dQGchfYxqF6NRsCQBhVXROCBBXB6J2Z0olFZgQgnDIFGGFJs uYBwcTba3mmsPeGCGVpsgcGWiDUaJwcddDABBxkMwYNvXQQxhA4xaBonESoUgQMMG0wIAEUaWPjw AxA2MKGqfTYs0UIQS7ygRAABAQA7 ------=_NextPart_000_0000_01C85786.26B66160 Content-Type: image/gif Content-Transfer-Encoding: base64 Content-Location: http://appft1.uspto.gov/netaicon1/PTO/top.gif R0lGODlhPwAYAPcAAAAAAAAAIQAAUgAAnAAAtQAAvQAAxgAAzgAA1gAA3gAA5wAA7wAA9wAA/wAI xgAIzgAI/wgArQgAxggAzggA1ggA5wgA9wgA/wgIpQgItQgIvQgIxggIzggI3ggI5wgI/xAAxhAI vRAI1hgQpRgYjCEYziEhlCEhxiEpnCEprSEptSEpvSExlCExnCkhrSkphCkprSkpvSkpxikpzikp 1ikxlCkxnCkxpSkxtSkxvSkxxikxzjEpnDEpvTExezExlDExnDExpTExrTExtTExvTExxjEx1jE5 pTE5rTE51jk5ezk5pTk5rTk5tTk51kI550JChEJClL21/729/73G/8a9/8bG78bG/8bO58bO98bO /87G/87O987O/87W987W/9bO/9bW/9be/97e/+fe/+/v7+/v//f3//f////3////5///9/////// //////////////////////////////////////////////////////////////////////////// //////////////////////////////////////////////////////////////////////////// //////////////////////////////////////////////////////////////////////////// //////////////////////////////////////////////////////////////////////////// //////////////////////////////////////////////////////////////////////////// //////////////////////////////////////////////////////////////////////////// //////////////////////////////////////////////////////////////////////////// /////////////////////////////////////////////////////ywAAAAAPwAYAAAI/gDVrDmT 5gyagwjToFGYEOFChwwPRnzYEOLCNBgdHiS4pgyZLleoWMmipUvJKyZTainJkqVJlytVwoSZJWVK ki+7WDkT5owUkFe0XJEytKjRo0RDJjVKZanTKyGVFk06BWoXKVHSXB2RoWvXDSHCig2rIQPYr2ND aBC7lixbthlCxAULIgRYuxEGNCGB5sqWCBMqVEggWLCCBIgVKKiweDFjxoobK4YsuTLjxIULJ1Ag YoMTIWlKZqhgoYHp0wtSN1iw2jTr1qxjw54t+3Xq16cbfGCtYMOJJmiyXBmtOrfx48iTK1+u2wMH HUzSCM+g4MNx3MyzL4fg+nhvHTa0/g6vcD05xvNptKtX/VoBBxlHxIcgnxv7cQAYAehXz5z9afcy gHbVfNrhll99qCXY3YII0rdaBRzQ0IR89MnGIG4LHNgaeqahl96FCFoI4QxDaCXFBh0UpyCCDWjY Yhr6YfRijOlhCCJqvckwIUghJGAbi9hlCOOGQ+Z3oIs/tlbfAr0VAZpQPaqYoH0znnYkjFcCcB2V rn0nBBpdUBGlcUn+6GKWWSrIpWsebKBDEGhASRhqt5W52oGxoUfjefv9yCUEvGmQhAlgThECBw1w 5xqgis62Wp+rQaDffvlN6uiaC0DwQQJuMtFXF9TZlyR/rrlIZn+9GQFDGlRcEYEHuxbYiZyNK5q2 H4MhzgZoBSKcMAQaYUhx6AK7mRarBQ6Syt+xHxy7QAWp9faEC2ZosQUGhyI2QQccdMtBAtwmsO24 HZBrbrnonjuut+x220G5HGQwBA/SdRHEEDrEMEQOROAwxBAxEKFDD/0SUfDBOBicMMIKN3ywDjoU EXEMAatQBA4wMCEAFGlg4cMPQNjABAs2/GCyyUIEUUMNP6zcMssuxwzzzC/XvHIQOOMMxBE2LNFC EEu8oEQAAQEAOw== ------=_NextPart_000_0000_01C85786.26B66160--

U.S. Current = Class:	600/310; 356/300 =
U.S. Class at = Publication:	600/310; 356/300 =
Intern'l Class: =	A61B 5/00 20060101 = A61B005/00; G01J=20 3/00 20060101 G01J003/00