CA2183085A1 - Method and apparatus for non-invasive detection of physiological chemicals, particularly glucose - Google Patents
Method and apparatus for non-invasive detection of physiological chemicals, particularly glucoseInfo
- Publication number
- CA2183085A1 CA2183085A1 CA002183085A CA2183085A CA2183085A1 CA 2183085 A1 CA2183085 A1 CA 2183085A1 CA 002183085 A CA002183085 A CA 002183085A CA 2183085 A CA2183085 A CA 2183085A CA 2183085 A1 CA2183085 A1 CA 2183085A1
- Authority
- CA
- Canada
- Prior art keywords
- spectrum
- glucose
- interferogram
- data
- test subject
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Abandoned
Links
- WQZGKKKJIJFFOK-GASJEMHNSA-N Glucose Natural products OC[C@H]1OC(O)[C@H](O)[C@@H](O)[C@@H]1O WQZGKKKJIJFFOK-GASJEMHNSA-N 0.000 title claims abstract description 101
- 239000008103 glucose Substances 0.000 title claims abstract description 101
- 238000000034 method Methods 0.000 title claims abstract description 72
- 239000000126 substance Substances 0.000 title claims abstract description 44
- 238000001514 detection method Methods 0.000 title claims description 5
- 238000001228 spectrum Methods 0.000 claims abstract description 60
- 238000012360 testing method Methods 0.000 claims abstract description 43
- 230000005855 radiation Effects 0.000 claims abstract description 38
- 238000001914 filtration Methods 0.000 claims abstract description 31
- 238000002835 absorbance Methods 0.000 claims abstract description 20
- 238000000862 absorption spectrum Methods 0.000 claims abstract description 18
- 238000005259 measurement Methods 0.000 claims abstract description 18
- 238000013178 mathematical model Methods 0.000 claims abstract description 10
- 230000001678 irradiating effect Effects 0.000 claims abstract description 5
- 230000008569 process Effects 0.000 claims description 11
- 230000009466 transformation Effects 0.000 claims description 6
- 238000010238 partial least squares regression Methods 0.000 claims description 5
- 238000012545 processing Methods 0.000 abstract description 6
- 238000010276 construction Methods 0.000 abstract description 4
- 229960001031 glucose Drugs 0.000 description 93
- WQZGKKKJIJFFOK-VFUOTHLCSA-N beta-D-glucose Chemical compound OC[C@H]1O[C@@H](O)[C@H](O)[C@@H](O)[C@@H]1O WQZGKKKJIJFFOK-VFUOTHLCSA-N 0.000 description 92
- 235000001727 glucose Nutrition 0.000 description 79
- 230000006870 function Effects 0.000 description 25
- 239000008280 blood Substances 0.000 description 16
- 210000004369 blood Anatomy 0.000 description 16
- 230000003595 spectral effect Effects 0.000 description 14
- URAYPUMNDPQOKB-UHFFFAOYSA-N triacetin Chemical compound CC(=O)OCC(OC(C)=O)COC(C)=O URAYPUMNDPQOKB-UHFFFAOYSA-N 0.000 description 14
- 238000004458 analytical method Methods 0.000 description 11
- 108091003079 Bovine Serum Albumin Proteins 0.000 description 10
- 229940098773 bovine serum albumin Drugs 0.000 description 10
- 230000003287 optical effect Effects 0.000 description 9
- 238000010521 absorption reaction Methods 0.000 description 7
- 239000012491 analyte Substances 0.000 description 7
- 239000001087 glyceryl triacetate Substances 0.000 description 7
- 235000013773 glyceryl triacetate Nutrition 0.000 description 7
- 229960002622 triacetin Drugs 0.000 description 7
- UHOVQNZJYSORNB-UHFFFAOYSA-N Benzene Chemical compound C1=CC=CC=C1 UHOVQNZJYSORNB-UHFFFAOYSA-N 0.000 description 6
- 230000004044 response Effects 0.000 description 6
- 230000000875 corresponding effect Effects 0.000 description 5
- 239000011159 matrix material Substances 0.000 description 5
- 239000008363 phosphate buffer Substances 0.000 description 5
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 description 5
- 229910052736 halogen Inorganic materials 0.000 description 4
- 238000012544 monitoring process Methods 0.000 description 4
- 238000002474 experimental method Methods 0.000 description 3
- WPYVAWXEWQSOGY-UHFFFAOYSA-N indium antimonide Chemical compound [Sb]#[In] WPYVAWXEWQSOGY-UHFFFAOYSA-N 0.000 description 3
- 238000005457 optimization Methods 0.000 description 3
- 241000894007 species Species 0.000 description 3
- RGHNJXZEOKUKBD-SQOUGZDYSA-N D-gluconic acid Chemical compound OC[C@@H](O)[C@@H](O)[C@H](O)[C@@H](O)C(O)=O RGHNJXZEOKUKBD-SQOUGZDYSA-N 0.000 description 2
- MHAJPDPJQMAIIY-UHFFFAOYSA-N Hydrogen peroxide Chemical compound OO MHAJPDPJQMAIIY-UHFFFAOYSA-N 0.000 description 2
- 229910019142 PO4 Inorganic materials 0.000 description 2
- XUIMIQQOPSSXEZ-UHFFFAOYSA-N Silicon Chemical compound [Si] XUIMIQQOPSSXEZ-UHFFFAOYSA-N 0.000 description 2
- 230000008901 benefit Effects 0.000 description 2
- 238000004364 calculation method Methods 0.000 description 2
- HVYWMOMLDIMFJA-DPAQBDIFSA-N cholesterol Chemical compound C1C=C2C[C@@H](O)CC[C@]2(C)[C@@H]2[C@@H]1[C@@H]1CC[C@H]([C@H](C)CCCC(C)C)[C@@]1(C)CC2 HVYWMOMLDIMFJA-DPAQBDIFSA-N 0.000 description 2
- 150000001875 compounds Chemical class 0.000 description 2
- 238000013016 damping Methods 0.000 description 2
- 206010012601 diabetes mellitus Diseases 0.000 description 2
- 235000013399 edible fruits Nutrition 0.000 description 2
- 238000012417 linear regression Methods 0.000 description 2
- 230000033001 locomotion Effects 0.000 description 2
- 239000013307 optical fiber Substances 0.000 description 2
- NBIIXXVUZAFLBC-UHFFFAOYSA-K phosphate Chemical compound [O-]P([O-])([O-])=O NBIIXXVUZAFLBC-UHFFFAOYSA-K 0.000 description 2
- 239000010452 phosphate Substances 0.000 description 2
- 108090000623 proteins and genes Proteins 0.000 description 2
- 102000004169 proteins and genes Human genes 0.000 description 2
- 238000000985 reflectance spectrum Methods 0.000 description 2
- 229910052710 silicon Inorganic materials 0.000 description 2
- 239000010703 silicon Substances 0.000 description 2
- 239000007787 solid Substances 0.000 description 2
- 238000002834 transmittance Methods 0.000 description 2
- 101150034533 ATIC gene Proteins 0.000 description 1
- BSYNRYMUTXBXSQ-UHFFFAOYSA-N Aspirin Chemical compound CC(=O)OC1=CC=CC=C1C(O)=O BSYNRYMUTXBXSQ-UHFFFAOYSA-N 0.000 description 1
- 238000012935 Averaging Methods 0.000 description 1
- RGHNJXZEOKUKBD-UHFFFAOYSA-N D-gluconic acid Natural products OCC(O)C(O)C(O)C(O)C(O)=O RGHNJXZEOKUKBD-UHFFFAOYSA-N 0.000 description 1
- 238000005033 Fourier transform infrared spectroscopy Methods 0.000 description 1
- 108010015776 Glucose oxidase Proteins 0.000 description 1
- 235000000177 Indigofera tinctoria Nutrition 0.000 description 1
- 101150081532 KLK8 gene Proteins 0.000 description 1
- JVTAAEKCZFNVCJ-UHFFFAOYSA-M Lactate Chemical compound CC(O)C([O-])=O JVTAAEKCZFNVCJ-UHFFFAOYSA-M 0.000 description 1
- 241001465754 Metazoa Species 0.000 description 1
- 238000011497 Univariate linear regression Methods 0.000 description 1
- XSQUKJJJFZCRTK-UHFFFAOYSA-N Urea Chemical compound NC(N)=O XSQUKJJJFZCRTK-UHFFFAOYSA-N 0.000 description 1
- ZVQOOHYFBIDMTQ-UHFFFAOYSA-N [methyl(oxido){1-[6-(trifluoromethyl)pyridin-3-yl]ethyl}-lambda(6)-sulfanylidene]cyanamide Chemical compound N#CN=S(C)(=O)C(C)C1=CC=C(C(F)(F)F)N=C1 ZVQOOHYFBIDMTQ-UHFFFAOYSA-N 0.000 description 1
- 239000000654 additive Substances 0.000 description 1
- 230000000996 additive effect Effects 0.000 description 1
- 238000013459 approach Methods 0.000 description 1
- QVGXLLKOCUKJST-UHFFFAOYSA-N atomic oxygen Chemical compound [O] QVGXLLKOCUKJST-UHFFFAOYSA-N 0.000 description 1
- 238000005452 bending Methods 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 230000017531 blood circulation Effects 0.000 description 1
- 239000000872 buffer Substances 0.000 description 1
- 239000004202 carbamide Substances 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 235000012000 cholesterol Nutrition 0.000 description 1
- 239000000306 component Substances 0.000 description 1
- 238000004590 computer program Methods 0.000 description 1
- 239000000470 constituent Substances 0.000 description 1
- 238000007405 data analysis Methods 0.000 description 1
- 238000009795 derivation Methods 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 230000001066 destructive effect Effects 0.000 description 1
- 210000000624 ear auricle Anatomy 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000011156 evaluation Methods 0.000 description 1
- 239000003925 fat Substances 0.000 description 1
- 239000000835 fiber Substances 0.000 description 1
- 210000005224 forefinger Anatomy 0.000 description 1
- 235000012208 gluconic acid Nutrition 0.000 description 1
- 239000000174 gluconic acid Substances 0.000 description 1
- 238000007446 glucose tolerance test Methods 0.000 description 1
- 150000002367 halogens Chemical class 0.000 description 1
- 238000000338 in vitro Methods 0.000 description 1
- 238000001727 in vivo Methods 0.000 description 1
- 229940097275 indigo Drugs 0.000 description 1
- COHYTHOBJLSHDF-UHFFFAOYSA-N indigo powder Natural products N1C2=CC=CC=C2C(=O)C1=C1C(=O)C2=CC=CC=C2N1 COHYTHOBJLSHDF-UHFFFAOYSA-N 0.000 description 1
- 229910052738 indium Inorganic materials 0.000 description 1
- APFVFJFRJDLVQX-UHFFFAOYSA-N indium atom Chemical compound [In] APFVFJFRJDLVQX-UHFFFAOYSA-N 0.000 description 1
- 208000015181 infectious disease Diseases 0.000 description 1
- 238000007689 inspection Methods 0.000 description 1
- 238000011545 laboratory measurement Methods 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 239000000203 mixture Substances 0.000 description 1
- 238000000491 multivariate analysis Methods 0.000 description 1
- 238000012314 multivariate regression analysis Methods 0.000 description 1
- 239000001301 oxygen Substances 0.000 description 1
- 229910052760 oxygen Inorganic materials 0.000 description 1
- 150000002978 peroxides Chemical class 0.000 description 1
- 239000010453 quartz Substances 0.000 description 1
- 230000003252 repetitive effect Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000005070 sampling Methods 0.000 description 1
- 230000035945 sensitivity Effects 0.000 description 1
- VYPSYNLAJGMNEJ-UHFFFAOYSA-N silicon dioxide Inorganic materials O=[Si]=O VYPSYNLAJGMNEJ-UHFFFAOYSA-N 0.000 description 1
- 238000003860 storage Methods 0.000 description 1
- 235000000346 sugar Nutrition 0.000 description 1
- 210000003813 thumb Anatomy 0.000 description 1
- 210000003371 toe Anatomy 0.000 description 1
- 230000001131 transforming effect Effects 0.000 description 1
- 238000000411 transmission spectrum Methods 0.000 description 1
- 150000003626 triacylglycerols Chemical class 0.000 description 1
- 229910052721 tungsten Inorganic materials 0.000 description 1
- 239000010937 tungsten Substances 0.000 description 1
- -1 tungsten halogen Chemical class 0.000 description 1
- 210000000707 wrist Anatomy 0.000 description 1
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01J—MEASUREMENT OF INTENSITY, VELOCITY, SPECTRAL CONTENT, POLARISATION, PHASE OR PULSE CHARACTERISTICS OF INFRARED, VISIBLE OR ULTRAVIOLET LIGHT; COLORIMETRY; RADIATION PYROMETRY
- G01J3/00—Spectrometry; Spectrophotometry; Monochromators; Measuring colours
- G01J3/28—Investigating the spectrum
- G01J3/45—Interferometric spectrometry
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/145—Measuring characteristics of blood in vivo, e.g. gas concentration, pH value; Measuring characteristics of body fluids or tissues, e.g. interstitial fluid, cerebral tissue
- A61B5/14532—Measuring characteristics of blood in vivo, e.g. gas concentration, pH value; Measuring characteristics of body fluids or tissues, e.g. interstitial fluid, cerebral tissue for measuring glucose, e.g. by tissue impedance measurement
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/145—Measuring characteristics of blood in vivo, e.g. gas concentration, pH value; Measuring characteristics of body fluids or tissues, e.g. interstitial fluid, cerebral tissue
- A61B5/1455—Measuring characteristics of blood in vivo, e.g. gas concentration, pH value; Measuring characteristics of body fluids or tissues, e.g. interstitial fluid, cerebral tissue using optical sensors, e.g. spectral photometrical oximeters
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/72—Signal processing specially adapted for physiological signals or for diagnostic purposes
- A61B5/7235—Details of waveform analysis
- A61B5/7253—Details of waveform analysis characterised by using transforms
- A61B5/7257—Details of waveform analysis characterised by using transforms using Fourier transforms
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/17—Systems in which incident light is modified in accordance with the properties of the material investigated
- G01N21/25—Colour; Spectral properties, i.e. comparison of effect of material on the light at two or more different wavelengths or wavelength bands
- G01N21/31—Investigating relative effect of material at wavelengths characteristic of specific elements or molecules, e.g. atomic absorption spectrometry
- G01N21/35—Investigating relative effect of material at wavelengths characteristic of specific elements or molecules, e.g. atomic absorption spectrometry using infrared light
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/17—Systems in which incident light is modified in accordance with the properties of the material investigated
- G01N21/25—Colour; Spectral properties, i.e. comparison of effect of material on the light at two or more different wavelengths or wavelength bands
- G01N21/31—Investigating relative effect of material at wavelengths characteristic of specific elements or molecules, e.g. atomic absorption spectrometry
- G01N21/35—Investigating relative effect of material at wavelengths characteristic of specific elements or molecules, e.g. atomic absorption spectrometry using infrared light
- G01N21/359—Investigating relative effect of material at wavelengths characteristic of specific elements or molecules, e.g. atomic absorption spectrometry using infrared light using near infrared light
Landscapes
- Physics & Mathematics (AREA)
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Spectroscopy & Molecular Physics (AREA)
- Medical Informatics (AREA)
- Animal Behavior & Ethology (AREA)
- Pathology (AREA)
- Engineering & Computer Science (AREA)
- Biomedical Technology (AREA)
- Heart & Thoracic Surgery (AREA)
- Optics & Photonics (AREA)
- Molecular Biology (AREA)
- Surgery (AREA)
- Biophysics (AREA)
- General Health & Medical Sciences (AREA)
- Public Health (AREA)
- Veterinary Medicine (AREA)
- General Physics & Mathematics (AREA)
- Emergency Medicine (AREA)
- Measurement Of The Respiration, Hearing Ability, Form, And Blood Characteristics Of Living Organisms (AREA)
- Investigating Or Analysing Materials By Optical Means (AREA)
- Measuring Or Testing Involving Enzymes Or Micro-Organisms (AREA)
- Investigating Or Analysing Biological Materials (AREA)
Abstract
Non-invasive measurements of physiological chemicals such as glucose are made using infrared radiation and a signal processing system that permits the construction of a device suited for home use. The level of a selected physiological chemical in a test subject is determined in a non-invasive and quantitative manner by a method comprising the steps of: (a) irradiating a portion of the test subject with near-infrared radiation such that the radiation is transmitted through or reflected from the test subject; (b) collecting data concerning the transmitted or reflected infrared radiation using a detector, (c) digitally filtering the collected data to isolate a portion of the data indicative of the physiological chemical; and (d) determining the amount of physiological chemical in the test subject by applying a defined mathematical model to the digitally filtered data. The data collected may be in the form of an absorbance spectrum, i.e., a representation of the variation in absorbance as a function of wavelength; a single beam spectrum or in the form of an interferogram, i.e., a representation of infrared light intensity as a function of the retardation of the moving mirror in the interferometer, and thus of time.
Description
WO 95/2~046 ~ i 8 3 0 8 ~ F ~ ~ 1556 METliOD AND APPARATUS FOR NON - INVASIVE DETECTION OF
PHYSIO~.OGICAL rT~FMTt~T q, pARTTCITT ~RrY GLUCOSE
The invention described in this application has been developed in part with funds received from the US National Institutes of Xealth under grant num.ber DR 45126. The United States GL~V~ ~ t may have certain rights under this invention.
BACRGROUND OF T~E lNV~ ~liJN
This application relates to a method and apparatus for the non-invasive, quantitative meaaurement o~
physiological rhPmirll~, particularly glucose, in a patient.
Determination of blood glucose is a routine proce-dure performed several times a day by many diabetics. In general, this procedure involves the taking of a small blood sample and evaluating the level of glucose in the sample. Common instL, t~ used for this purpose use the enzyme glucose oxidase to convert glucose and oxygen to gluconic acid and hydrogen peroxide, and then measure the level of peroxide by either spectroscopic or electrochemical means.
While these daily measurements provide the diabetic patient with the ability to self-monitor and thus better control blood glucose levels, they are not without draw-backs. In particular, the taking of blood sam.ples several time6 a day can be painful and exposes the patient to a risk of infection. Moreover, using this existing technology it is impossible to obtain a nntin-louS blood glucose mea~urement. Thus, during the night, a patient must either be awakened periodically for testing or run the risk that glucose levels will drop to dangerous levels 3 0 as they s leep .
Wo 95/22046 , - ~ PCT/usss~0lss6 ~
2183~85 In view of the foregoing, there exists a real need for a non-invasive method of measuring blood glucose in a patient. It has been suggested that this could be accom-plished using near-infrared (NIR) radiation. Thus, U.S, 5 Patent No. 5,086,229 of Rosenthal, which is incorporated herein by reference, de3cribes a system using a plurality of in~rAr~-l light emitting diodes and a detector to measure blood glucose. The infrared measurement of glucose in the body is immensely complicated, however, 10 becau6e of the suhstantial ab~orhAnnes of fats and proteins in the near-infrared. Thus, while the Rosenthal patent presents an interesting theory, there is no pl-hl; ch~d documentation that demonstrates the ability of the Rosenthal system to measure glucose noninvasively.
15 Furth~ ~~e, the wavelengths used in the Rosent_al syatem have never been shown to possess useful glucose inf ormation .
It i8 an obj ect of t_e present invention to provide a practically useful non-invasive near-infrared detector 20 for blood glucose and other physiological chemicale that is adaptable for continuous use.
It i8 a further object of the present invention to provide a method for the non-invasive quantitative measurement of physiological chemicals and particularly 25 glucose in a patient that can be used either nn~;nn~usly or intermittently.
It is still a further object of the invention to provide an apparatus for cnn~;nl~nus or intermittent non-invasive monitoring of hlood glucose and other 30 physiological chemicals in a patient.
~UMMARY OF THE LNV~ ~N
In accordance with the invention, non-invasive measurements of physiological chemicals such as glucose ~, WO 95/22046 2 1 8 3 0 ~ 5 PCTIU595~0155C
are made using infrared radiation and a signal processing system that permitæ the co~struction of a device suited for home use. Thus, the level of a selected physiolo~;cAl chemical in a test subject is detPrmlnp~ in a non-invasive 5 and quantitative manner by a method comprising the ateps of:
(a) irradiating a portion of the test subject with near-infrared radiation such that the radiation is transmitted through or reflected from the te6t subject;
(b) collectinq data conf Prnin~ the transmitted or reflected infrared radiation using a detector;
(c) digitally filtering the collected data to isolate a portion of the data indicative of the physiological chemical; and (d) detPrm;ning the amount of physiological chemical in the test subject by applying a defined mathe-matical model to the digitally filtered data. I'he data collected may be in the form of an Ahsorh~nme spectrum, i.e., a representation of the variation in absorbance as a function of wavelength; or in the form of an interfero-grAm, i.e., a representation of infrared light intensity as a function of the retardation of the moving mirror in the interf erometer, and thus of time .
In a further aspect of the invention, there i6 provided a device for measurement of a chemical in a sample comprising:
(a) means for collecting data con,Prnin~ near-inf rared radiation transmitted through or ref lected f rom the sample;
(b) means for digitally filtering the collected data to isolate a portion of data indicative of the chemi ~al;
W0 95/22046 PcrJuS9S/01556 218308~ --~, 4 (c) means for app1ying a definèd mathematical model to the digitally filtered data, whereby the amount of chem.ical in tbe test subject is deT-P~min~d; and (d) means for reporting the amount of chem.ical in the sample. This device can be incorporated into a complete apparatus for the non-invasive mea~ul, t of blood glucose. Such an apparatus would also include a source of infrared radiation and a t~hAn~ pm for directing the infrared radiation from the source to the test subject and from the test subject to the means for collecting the spectrum.
BRIEF DESCRIPTION OF THE nRPT~r~N~.q Fig. 1 is a flow chart depicting the method of the invention;
Fig. 2 is a flow chart depicting a process for digitally filtering an absorbance spectrum;
Fig. 3 is a schem,atic representation of an interf erometer;
Fig. 4 shows a sam.ple interferogram;
Fig. 5 shows a device in accordance with the present invention for measurement of glucose in a sam.ple;
Fig. 6 shows an apparatus in accordance with the present invention for non-invasive monitoring of physiological chemicals in a test subject;
Fig . 7 shows unf iltered interf erogram segments taken of phosphate buffer and three concentrations of glucose;
Fig. 8 shows digitally filtered interferogram segments taken of phosphate buffer and three concentrations of glucose;
Fig. 9 shows the relat~nQhl~ between filtered interferogram segment magnitude and actual glucose concentration;
Wo 95l22046 PCTIUS9S~OISS6 2183085 -= -Fig. 10 shows the relationship between glucose concentration de~rmin~d from digitally filtered - absorbance spectra of hu~n test subjects and actual glucose levels as de~ n~d by blood testing;
5 Fig. 11 shows the relationship between glucose concentration detPrm;n~d from unfiltered absorbance spectra of human test subjects and actual glucose levels as det~ined by blood testing;
Fig. 12 displays the data obtained from applying three distinct filters to an interferogram obtained from a glucose/BSA/triacetin sample;
Fig. 13 plots the standard error of calibration (solid line) and prediction ~dashed line) versus the number of PLS f actors used to f orm the calibration model of a filtered interferogram in a two-filter analysis based on interferogram points 340-640 for the glucose filter and points 200-301 for the BSA filter;
Fig. 14 is a correlation plot of glucose concentrations estimated using a 14 term calibration model versus the corresponding actual concentrations;
Fig. 15 is a correlation plot of glucose concentrations estimated using a 15 term calibration model for an unfiltered interferogram versus the corresponding actual concentrations;
nT~ T.Fn DESCRIPTION OF THE ll`JV~ 'lU~!I
Fig . 1 is a f low chart depicting the method of the invention, as applied to the non-invasive determination of blood glucose in a human patient. As shown, the first step of the method is the irradiation of the test subject, in this case the patient, with near-infrared radiation.
Suitable NIR radiation for use in the present invention ct i nrit~P5 with the absorbance bands of glucos~
or other physiological chemical being measured. For Wo95/220~6 r~ 1556~
21~308S
glucose, these bands are located in the regions of 5000-4000 cm l and 6500-5800 cm l. The intensity of the light at the selected wavelen~th should be on the order of 700 lux or greater. Such radiation can be produced by a 300 W
5 tungsten-halogen lamp.
9ince the spectrum obtained and used in the method of the present invention may be either a transmittance spectrum or a ref lectance spectrum, considerable latitude is available in the manner and location in which the NIR
lO radiation impinges on the test subject. For example, if transmitted NIR radiation is being measured, the NIR
radiation should impinge on a relatively thin, fleshy area of the patient such as the f leshy webs between the f ingers or toes or the ear lobe . If ref lectance spectra are to be 15 used, the sampling site should be characterized by high blood flow close to the surface, such as the ventral surface of the wrist.
The source of NIR radiation used in the present invention may be such that it is disposed directly against 20 the surface of the test subject. For example, a small halogen lamp could be used. Alternatively, the source may be physically remote from the test subject. In the latter case, it is advantageous, although not n~cP~s~ry, to guide the NIR radiation to the desired irradiation site on the 25 surface of the test subject, for example by means of optical f ibers .
In the second step of the method, the data concern-ing the transmitted or reflected NIR radiation is collected using a detector. The specific nature of the 30 detector is not critical, provided it is capable of detecting the pertinent wavelengths of light and respond-ing rapidly enough to be com~oatible with the other compon-ents of the device. An example of a suitable detector for collection of an absorbance spectrum is a combination of a 0 95J~2046 ~ ~ 8 ~ F~ ~ 556 dispersive element, e.g., a grating or prism, and an optical multi-channel analyzer sensitive to NIR radiation.
In the case where the data is to be collected as a single beam spectrum, an absorbance spectrum or an interf erogram, 5 a suitable detector is a combination of a NIR interf ero -meters and a photon counting detector such as a solid state indium an~ detector.
The positioning of the detector relative to the test subj ect will depend both on the nature and size of the 10 detector and the environment in which the mea~uL~ ~ is being taken. For most purposes, it will be desirable to have the detector physically separated from the test subject, both because of dPtector size and to m~~;m;7e detector perf ormance by providing the detector with a 15 stable environment. It will therefore generally be advantageous to guide the transmitted or reflected NIR
r~ ;on to the detector, for example using optical f ibers .
Depending on the instrumentation selected, the data 20 corl~ern;n~ the transmitted or reflected radiation is collected as either an absorbance spectrum or an interferogram. In either case, the next step as shown in Fig. 1 is to digitally filter the data to isolate the portion of the data which is indicative of the chemical of 25 interest. The specific manner in which this step ig perf ormed depends on the f orm of the collected data .
.
Digital F; 1 ter;n~ of an Absorbance ~pectrum When the data collected is in the f orm of an absorbance spectrum, the process for filtering the data is 30 shown in Fig. 2. As shown, the absorbance spectrum, or a portion thereof containing information about the chemical being tested for, is transformed usinq a Fourier transform into a Fourier domain spectrum Wo gs/22046 ~ 1 8 3 0 8 ~ PCrlUS95/OlSS6 ,~
The transformed spectrum is then filtered by multi-plying it by a Gaussian function that weights spectral information associated with the chemical being analyzed.
In the case of glucose, a suitable Gaussian function i8 5 aefined by a mean position of 0.037 f (digital frequency units) and a standard deviation width of 0.011 f. This function emphasizes a glucose-associated ~h50rh~nf~e band appearing in the spectral region between 6500 and 5800 cm 1. Other peaks, and other Gaussian functions for the same 10 peak may also be employed. For example, the glucose-zlssociated absorbance band at 4400 cm l can be emphasized.
For physiological chemicals other than glucose, appropriate ~P~Asi~n functions may be i(qPntifi~ For absorbance spectra processing, the proper Gaussian 15 function can be obtained by analyzing the shapes of the ~hsorh;~nce bands for the compound of interest.
After multiplication of the spectrum with the Gaussian function, the resulting filtered spectrum is converted back to an absorbance spectrum using an inverse Fourier 2 0 transf orm.
While the use of a Fourier filtering process such as that described above is a preferred method of performing the method of the invention, other digital filtering methods may also be used. Such methods include finite 25 impulse response and inf inite impulse response digital f ilters .
Digital Filtering of a Single Beam Spectrum or an Interf erogram - =
An important aspect of non-invasive measureme~t of 30 glucose and other l~hysiological chemicals using interfero-grams is the ability to analyze this data in the absence of a contemporaneously collected background or ref erence spectra. This is the case because no convenient method ~ WO 95/220~16 2 1 ~ 3 Qg S PCT/US9SJ01556 ~ g e~cists to collect a background spectrum such as the spectrum of a particular human subject without glucose.
When the data c~)nClorn; n~ the transmitted or ref lected inf rared radiatio~ is collected in the f orm of 5 an interferogram, it may be processed in either of two general ways. First, the interferogram may be converted to a conventional single beam spectrum using a Fourier transform and then filtered by multiplication with an ayylUyliate Gaussian function. Fûr example, in the case of a non-invasive determination of glucose, a Gaussian function having half-power points of 4445.4 and 4354.6 cm-can be used to emphasize the glucose absorbance band at 4400 cm~l. This corresponds to a ~All~SiAn furlction having a mean position of 0.139 f and a standard deviation of 0.00122 f. Other Gaussian functions for digital filtering to obtain information from other glucose absr~rhAnc~ or for absorption bands associated with other chemicals of interest can be obtained by matching functions to the width and location of the absorption bands.
This filterins differs from the process described above, in that ratioing to eliminate the background signal is not performed. secause of this, after multiplication of the transformed :iueU~Ll the resulting filtered spectrum can be subjected to an inverse Fourier transformation to yield a filtered interferogram. Such an interferogram is much simpler in Arp~ArAnce than the original interferogram, because much of the noise and absorbances not associated with the chemical of interest have been removed. Further, there is a clear correlation between the magnitude of the interferogram signal and the amount of the chemical being measured in the test subject.
The magnitude of the interferogram peaks is advan-tageously measured by the simple procedure of integrating the interf erogram to determine the total power of the Wo gsl22046 , ~ p ~ . lS56 ~
signal. ThiG yields a single value related to the glucoae concentration. Alternatively, the individual points in the filtered interferogram could be submitted to a multivariate calibration method such as a partial least 5 squares regression. This would yield a multivariate del relating the individual interferoqram point intensities to glucose concentration.
Alternatively to transforming the interferogram into a single beam spectrum, the interferogram itself may be 10 subjected to a digital filtering process, and the magnitude of the filtered interferogram peaks analyzed using a prP~iPfinP~l mathematical model. We have defined a model which predicts an approximately linear relatinn~hip between analyte co~centration and f iltered interf erogram 15 intensity as given by the formula C NP
where C is the concentration, P is the magnitude of the filtered seqment, and N is a sensitivity term defined as N=
Z [k] Z
~ k=nl where Z [k] is an interferogram domain function. The approximate linear rela~;c.n~h;rl between the analyte 20 concentration and magnitude of the filtered interferogram segment is based on the assumDtion that a NacLaurin series approximation used in the derivation of the relati~n~hir is valid. This approximation only holds for small absorbance values. For example, Sparks et al., Anal.
25 Chem. 54: 1922-1926 (1982) have estimated that this ~ Wo 9sl22~)46 218 3 0 8 S PCrmss5/DI556 approximation has less than 596 error if the absorbance value is less than 0.12. For glucose analysis using the present invention, peak absorbance of the glucose band is on the order of 10 4 or less. Thus, the use of this 5 approximation appears valid.
The approximate linear rela~inn~h;r also depends on the value of N being substantially constant across the calibration samples. This re~auires that the sample pathlength, molar absorptivity and the intensity of the filtered reference spectrum be constant with respect to concentration .
Variation in sample pathlength is not a problem in laboratory measurements due to the use of a fixed pathlength sample cell. It could present more of a problem in actual noninvasive tests performed on body tissue. Moreover, since the entire spectral band is isolated by the bAn~rA~s filter rather than a single vav~ ' cr, some change in molar absorptivity might be expected across the concentration range. These two factors could lead to limitations on the dynamic range over which the linear relationship would hold, potentially making several calibration factors necessary. They should not, however, prevent the use of the linear approximation for making noninvasive glucose determinations.
The assumption of constant intensity of the filtered reference spectrum i8 problematical, as some variation in infrared intensity is common in a Fourier transform inf rared PYrPr; t due to changes over time in instl, ~_A1 characteristics such as interferometer alignment and detector temperature. These factors can be uv~ through the use of a nn~^l; 7ation step.
Finally, the linear relatinncll;~ set forth above assumes that only a single analyte band has been isolated by the bandpass filter. If other analyte bands or bands wo 95/22046 PCT/USg5~01556 ~g3~5: ~
from other constituents in the sample matrix fall within the filter bandpass, their interferogram signatures will add to the signature of the targeted analyte band. The filtered interferogram will then possess a more complex 5 structure and the simple univariate linear relat;on~h;p will not apply. The selection of an appropriate bandpass filter is therefore quite important to the succes6 of the direct measurement of interf erogram inte~sity as a method for noninvasive determination of physiological ,h.omi 10 If suitable filters cannot be identified for a given analyte and sample matrix, however, a multivariate analysis procedure can be used on place of the linear approximation as discussed in 13xamples 4 and 5 herein.
Application of digital filtering directly to the 15 interferogram depends on the fact that each wavelength of light is represented in the interferogram by a modulated sine wave whose frequency (in hertz) depends on the wave-length of the light and the interferometer mirror velocity in accordance with the equation frequency ~vd v, ' ^r mirror ofsignal 2XofradiationXvelocity 20 where the ~c-v. ' cr has units of cm l, the mirror velocity is expressed in units of cm/sec. By selecting an electronic f ilter that selects out only the portion of the interf erogram having the f requency def ined by this equation f or the selected absorbance of the chemical being 25 meagured, a simplified interferogram is obtained.
The construction of electronic filters to select for specified wavelengths is ~ ce in the electronic arts. Briefly, the filter would be implemented as a digital filter. The software r~n~rQll ;n~ the device would 30 implement the filter as part of the data processing. The ~ Wo 95~22046 2 1 8 3 0 8 $ PCTrUSs5/~ 56 digital filter used implements an approximation of the convolution sum y[n] = ~ x[k] h [n-k]
k=-~
' ;' where y[n] is the filtered interferogram at point n, x[k]
is the original unfiltered interferogram at point k, and h [n-k] are points on the impulse response of the digital filter, i.e., the interferogram-domain representation of the frequency response filter.
While the digital filtering procedure for interfero-grams has been described above in terms of a Fourier filtering process, other digital f;ltl~r;n~ methods may also be used. In particular, methods such as f inite impulse response and inf inite impulse response f ilters can be used. These filter design methods approximate the infinite summation in the convolution sum above. ~hese methods of f er the advantage of reducing the size of the interferogram required, which can lower the performance requirements of the interf erometer .
While the method of the invention can be practiced on complete interferograms, it has been found that short segments of an interferogram can be used as well. This has great significance to the application of this terhn~l ogy in relatively rigorous environments, because it facilitates the construction of a smaller interferometer with less stringent mechanical tolerances.
An interferometer such as that shown in Fig. 3 includes a 50: 50 beam splitter 30, two mirrors 31 and 32, and optical track 33 and a detector 34. Infrared radiation is passed through the beam splitter 30 and wo gsl22046 ~ . Pcr/usss/olss6 21~3~8~ --directed toward the two mirrors 31 and 32, oriented at 90 degrees to one another, which reflect it back towards the detector 34. At the same time, the mirror 31 is moved back and forth on the optical track 33. The movement of the mirror 31 results in sUccessive occurrences of cons tructive and des tructive interf erence . Thus, the signal received at the detector is a harmonic signal whose f requency is determined by the f reguency of incident light and the mirror velocity. A sample of the detected signal, called an interferogram, is shown in Fig. 4.
InterfeLuSIL can be converted by a mathematical transform into conventional spectra of very high quality.
~lowever, the mechanical ~ r; ty of the interferometer, and particularly the need for precise control of the move-ment of the mirror over a substantial length of travel and for precise optical ~ t, limits the ap~?licability of interferometers in rigorous environments. The mechanical complexity of the instrument could be substantially reduced if the length of the interferogram were shortened, because the length of mirror travel could be shorter. It is generally recognized, however, that the use of short interft,Lu~Lcu~s significantly reduces the quality of the spectrum which can be obtained upon transformation of the interf erogram to a conventional absorbance spectrum.
The reason for this loss of quality has to do with the mathematical approximation normally used to perform the transformation, the fast Fourier transform or FFT.
The FFT is an approximation of the infinite Fourier integral in which it is assumed that the time signal is zero outside the region actually sampled. Supplying a very short interf erogram to the FFT has the ef f ect of convolving a boxcar function with the true (i.e.
infinitely sampled) interferogram. The Fourier transform of the boxcar function is a (sin x) /x function. When this ~ W0 95/22046 2 1 ~ 3 ~ 8 ~ ss6 function i5 superimposed on the true 6pectrum, a lowering of resolution results.
In conventional interferometers in use today, the total travel of the mirror is around 2 cm, with spectral 5 collection actually oCcllrri n~ over approximately 1 cm of this length. It has surprisingly been found, however, that short interferoqram segments, i.e., interft~ Lcu.~
collected over distance of mirror travel of about 0 . 003 to 0.10 cm can be used effectively in the method of the 10 present invention. This opens the door to the cons truction of interf erometers that are smaller and inherently more rugged than their laboratory predecessors, and which thus are suited for use, e.g., by an individual diabetic in the home.
A further advantage of the method as of the present invention is ability to dispense with separate background spectral measurements through the use of interf ~:ru-JLeu..s .
Such measurements are necessary in most systems to correct for instrumental variations and the absorbance of com-20 pounds other than the ~ , ulld of interest. In the case of non-invasive monitoring of physiolo~;cAl chemicals, however, this background is highly complex. Fur~h~ e, it is not readily apparent how one would obtain a spectrum of the test subject, minus the chemical being analyzed.
25 When using interfeL~,yL~.~, however, no background is nPcpsE;lry. This is the case, because in an interferogram, the portion of the signal which can be attributed to the overall instrument function is a wide feature which is represented in the interf erogram as a rapidly damping 30 sinusoidal signal. In contrast, absorption bands of compounds are narrower and thus their interf erogram repre-sentation does not damp as rapidly. Thus, by taking the analyzed data from an interferogram segment removed from the centerburst, the instrumentation flPrPnflPnt signal has Wo 95/22046 -~ 3 0 8 ~ Pcrluss5/01ss6 essentially died out, leaving only the signals associated wi th actual inf rared absorbances .
Regardless of the form of the collected data, the final step in the method as ~hown in Fig. l is the 5 determination of glucose or other physiological chemical concentration by application of a previously detprm; nPd mathematical (calibration) model. In either case, the input into the model is the filtered data, and the output is the estimated concentration of glucose or other lO physiological chemical of interest. The model is aeveloped based upon a multivariate regression analysis, such as partial least squares (PLS) regression. In PLS, the in'lPpPn~lPnt VAr;AhlP~ are the filtered data and the dPrPn~lPnt variable is the glucose concentration. Other 15 multivariate regression methods such as principal _ onPnts reqression can also be used.
The mathematical del resulting from application of any of these techniques to a set of data in which the actual glucose cr~nrPn~r~tions are known can be expressed 20 as a series of term3 the value of which correspond to measured properties of the data for a sample, and factors or regression coefficients. For example, a model found to be useful in dete~ninin~ glucose conrPntrations in vivo using data in absorbance spect~um form has the form c j =bo +blx~, l + - +bhX~, h 25 where c1 is the predicted glucose concentration corres-ponding to spectrum i, the x1 values are the partial least-3quares factor scores computed from the absorbance data, and the b term3 are the coefficients detPrm;nPd from a multiple linear regression analysis of the measured 30 glucose concentrations corresponding to the calibration .
~ W0 95/220~6 ~ 1 8 3 Q 8 5 1~ 015~6 absorbance data and the h sets of partial least-sguares scores computed from the calibration absorbance data.
To analyze an unknown sample, values measured for the sample are substituted into this equation to produce - 5 an estimate of the glucose concentration in the sample.
As a general rule, device limitations will probably reguire that this model be established for each chemical being measured using instrumentation separate from the device actually being used by a test subject to perform day-to-day nitoring. Nevertheless, some calibration function may be included in an individual device to adapt it to a specific test subject.
While the method of the invention has been described prinr;r~lly with respect to the monitoring of glucose in a human subject, the scope of the invention is not so limited. Physiological chemicals other than glu-cose, such as urea, lactate, triglycerides, total protein and cholesterol may also be guantitatively detected using the method of the invention through the selection of a~Lu~Liate digital filtering p~ tr-rs. Fur~hr ~L~, because the filtering is all post-collection processing performed on stored data, a single spectrum can be anal-yzed numerous times to provide analytical results on re than one physiological chemical.
Test subjects to which the method of the invention may be applied include not only human subjects, but also animal subjects, e.g., during veterinary procedures, and even living plants . For example, non- invasive determina-tion of sugar levels within fruits could be made to assess the ripeness of the fruit and its reP~in~s for harveat-ing . Thus, as used in the specif ication and claims here -of, the term test subject refers to any living organism contalning physiological chemicals.
WO 95/22046 2 1 8 3 0 8 5 PCI/US95/01556 ~
A further aspec't of the E~resent invention is a device for measurement of glu~cosb~'in a sample as shown in Fig. 5. This device, whic~ can be used on any type of sample, not just a living organism, includes means for 5 collecting a transmitted or reflected infrared radiation such as interferometer 50 or a combination of a dispersive element and an optical multichannel analyzer. The collected spectral data is transferred to a computer read-able storage medium 51 from whence it is read by computer 10 processor 52 and processed in accordance with a ~L~yL -~instruction to digitally filter the collected ~e- LL~-- to isolate a portion of the spectrum indicative of glucose.
9uitable computer ~rocessors for use in this application include Intel 486 and PentiumlD microprocessors or MIPS
15 R3000 and R4000 processors of the type used in Silicon Graphic8 workstations.
The computer program used to perform the digital filtering steps can be written in any pLUyL ;nq language capable of performing a Fourier transform and multiplying 20 a data set by a Gaussian function. For example, a program written in Fortran 77 was employed to perform the data analysis described in the example herein.
Once the stored data have been digitally filtered, the computer processor 52 is operated under a second set 25 of LJlOyL -d instructions in order to apply a predeter-mined mathematical model to the data. The result of this application i8 a numeric estimate for the cr~nc~ntr~tion of glucose or other chemical of interest in the test subject.
The device of the invention also includes one or 30 more means for reporting the amount of glucose in the sample . Examples of means f or reporting the amount of glucose include digital display panels 53, transportable read/write magnetic media such as computer disks and tapes which can be transported to and read on another machine, ~ Wo 95/22046 and printers such as thermal, laser or ink- jet printers f or the production of a printed report .
A further aspect of the present invention is an apparatus for the non-invasive, quantitative detection of 5 physiological chemicals in living organisms. Such an apparatus may be a combination of a device for glucose measurement as described above, or a comparable device having the digital filter parameters and standard values targeted for a different physiological chemical with a 10 means for irradiating the surface of the living organism with near-infrared radiation in such a way that the spectral data can be collected in a transmittance, diffuse reflectance or ~r~n~flectance configuration.
Fig . 6A shows an apparatus of this type f or collect -15 ing spectral information. The apparatus has a housingwhich encloses a light source 61, a detector 62, and a microprocessor 63. Infrared radiation from light source 61 is transmitted through optical fiber 64 to the test 6ubj ect 65 . Transmitted radiation i8 collected through 20 optical fiber 66 and transmitted to detector 62. The 6ignal from the detector is then transmitted to micropro-cessor 63. Microprocessor 63 digitally filters the signal from the detector, and applies the pre~lef;~Pd mathematical model to the filtered data to determine a value for glu-25 cose concentration which is displayed on display 67.Display 67 may be an LED or LCD display.
Other conf igurations of the apparatus can be util -ized in addition to the specif ic conf iguration shown in Fig. 6A. For example, the apparatus can be configured to 30 collect infrared radiation in a diffuse reflectance mode, rather than in a transmission mode. In the diffuse reflectance mode, a portion of the incident radiation penetrates a short distance into the sample before being scattered out of the sample. A fraction of this back-Wo 95~2046 2 1 8 ~~ PCr/uss5/ol556 ~
scattered radiation is collected and measured by the detector .
The apparatus may also be operated in a transf lec -tance mode. In a transflectance configuration, optical 5 fibers 64 and 66 are positioned apart by a distance D as shown in Fig . 6B . The second f iber collects light that has entered the sample and traversed a short distance within the sample medium.
In vitro measurements of glucose concentration were made on a series of thirteen glucose samples having nt~ations varying from 1.25 to 19.7 ~I glucose in pH
7 . 4 phosphate buf f er ( 0 .1 M) . The samples were placed in 1 mm pathlength cuvettes constructed from infrared quartz 15 and transmittance interfelu,~L of each were obtained using a Nicolet 740 spectrometer in conjunction with a tungsten-halogen light source and a cryogenically cooled InSb near-infrared detector. Double sided interfeLo.~L~,~
were collected by averaging across 256 co-added scans.
20 Either two or three replicate interfeLu~L~.~ were made for each sample, producing a total of 38 interferograms across the 13 samples.
To evaluate glucose concentration, the glucose absorption located at 4400 cm l was used and i601ated by 25 digital filtering. The filtering process involved application of a fast Fourier transform to the interferogram to produce a single beam ~ec~L,;
multiplication of the transformed interferogram with a Gaussian function having half -power points at 4445 . 6 and 30 4354.6 cm l; and application of the inverse fast Fourier transform to obtain the filtered interferogram.
Fig 7 shows four interf~LoyLcu.s segments (points 600-800 relative to the interferogram peak maximum or ~W095l22046 2~ dg~'`'`'`' PcTll~S9SlOlSS6 centerburst) obtained for phosphate buffer, 3 .22 ~
glucose, 7 . 95 }[~q glucose and 19 . 7 mM glucose . No obvious information relating to glucose concentration can be observed. Fig. 8 shows these same interfe:LuyL~Ls after - 5 the digital filtering procedure described above. The waveforms have been greatly simplified due to the removal of all frecluencies outside the bandpass of the Gaussian function. The interferogram segment for the buffer de-creases in intensity to a minimum point or node, followed by a slowly damping signal due to the water absorption feature. The interferograms of samples cont~;n;n~ glucose exhibit a less obvious node, but increased magnitudes are observed with increasing glucose co~c~ntr~tions.
Fig. 9 shows the results obtained when the magni-tudes of the filtered interfelu~L~,~ are plotted as a function of actual glucose concentration. An excellent correlation between the magnitude and the concentration is observed (Univariate linear regression: R s~uared=99.0%;
standard error of calibration = O . 62 mM) .
Ill vivo experiments on the non-invasive detection of glucose were performed using a Nicolet 740 research grade spectrometer configured with a 250 W tungsten halogen lamp, CaF~ beam splitter and InSb detector . Non- invasive measurements in accordance with the invention were taken at the same time as conventional invasive measurements on three individuals of varied physiognomy during the course of glucose tolerance tests. The beam of light from the spectrometer having a spectral range from 7000 to 5000 cm was passed through a portion of the webbing between the thumb and forefinger~ This particular ~v~v~ ' cr region was isolated using a standard astronomical H-band optical f ilter .
WO 95/22046 2 1 8 ~ 8~ PCT/US95/01556 ~
~ .i ,..
The spectra were collected in interferogram form using the interf erometer and InSb detector of the spectro -meter, and then converted to single beam spectrum ~intensity versus ~ V~ ' ~r) using the Ar- _-nying 5 Nicolet computer. All single beam spectra` were subsequently transf erred to a Silicon Graphics Indigo computer for data processing.
Blood samples taken during the test were analyzed for qlucose using a Yellow Springs Instruments (YSI) model 10 2300 glucoGe analyzer.
To process the spectra, the single beam spectra were f irst converted to absorbance units by dividing each spec-trum by a water re~erence spectrum and then computing the neqative logarithm. Only the portion of the spectrum 15 between 6500 and 5800 cm l was used in all subsequent calculations .
Spectra were then normalized using the signal between 5955 and 5951 cm~l, a ~vc.vl ` ?r region that contains no glucose inormation. The llnr~ l; 7ed spectra 20 were then digitally Fourier filtered to ~l iminAte noise and other spectral features that distract from glucose inf ormation .
The first step in the Fourier ~iltering process was to perform a Fourier transorm on the absorbance spectrum.
25 This resulted in a Fourier domain spectrum, which was then multiplied by a Gaussian function defined by a mean position of 0 . 037 f and a standard deviation width of O . 011 f. Thesc parameters were selected to isolate the glucose absorption. The filtered transformed spectrum 30 was then converted back to an absorbance spectrum by applying an inverse Fourier transform.
The spectra obtained in this way were analyzed using the standard PLS regression procedure. This procedure essentially goes into the data set and weights the 095/22046 ~ 8308$ pcrluss5/ol556 different spectral frequencies by correlating spectral variation with analyte concentration. The result is a - series of factors and regression coefficients which can be applied to individual spectra to obtain a predicted value 5 of glucose concentration.
Fig. 10 shows the correlation between glucose concentration as detPrm;nPd from the spectra and the glucose level as detPrm;nPc~ from the blood 6amples taken.
The open symbols in the graph represent the data that were 10 used to generate the calibration model. The solid symbols represent results from an independent set of spectra used only for prediction purposes.
A11 of the results cluster around the unity line, and the prP~1; ct; on points are within the general spread 15 and scatter of the calibration points. Thus, the ability of the method of the invention to predict blood glucose levels is conf; '. Moreover, since there is no clustering of data according to individual, a global calibration rather than individual calibrations may be 20 possible.
To evaluate the importance of the diqital filtering s tep in Example 2, the data taken in that experiment were reevaluated with the digital filtering step omitted and a 25 separate set of PLS factors were developed. The resulting plot of spectral glucose versus blood glucose measurements is shown in Fig. 11. As is a~parent, the data in this case has a much poorer corlelation. Thus, the importance of an appropriate digital f iltering technique is clear.
Wo 95/22046 PcrNS9Sl015S6 ~
~ 1 8 3 1~ 8 ~r 2 4 ~E~IPLE 4 To test the validity of the derived linear relation-ship between concentration and interferogram peak magnitude, a series of pr~l ;minAry experiments were 5 performed using solution sa~mples of benzene in CS2. CS2 has essentially zero absorbance in the region of the C-H
out-of-plane bending band of benzene at 1036 cm l that was used as the targeted band for the analysis. The linear approximation described above and a sim~le univariate 10 model were successful in relating the magnitude of the filtered interferogram segments to concentration.
A second series of tests were run using samples of glucose in phosphate buffer. The glucose C-H combination bAnd centered at 4400 cm~1 was used as the targeted band 15 for the analysis. This band is located on top of a broad background ~hssrhAnce of water spanning the 4000-5000 cm region, and thus presents a greater challenge since it violates the assumption made in deriving the linear rela~ic n~hir that only a single absorbing species would be 20 detected within the filter bandpass.
Calibration results made using the univariate linear approximation exhibited ~ubstantial errors. Since the information in the interfeL~L~w representing glucose and water should still be additive, however, it was 25 hypothesized that a multivariate calibration model could be used to correct for the interference due to water.
Partial least squares regression was used to build models of the f orm C=bo+blX1+- +b"x~
where C is glucose concentration and the n+1 bi values are 30 regression coefficients detPrm;n~d from a multiple linear regression analysis of the calibration data set. The independent variables, xl, are factor scores obtained from WO 95/22046 PcTn~ss5/0~556 the PLS analysis of the filtered interferogram segment.
The original independent variables used to obtain these latent variables are the intensities of the individual points in the filtered interferogram segment.
This approach was highly successful, yielding excellent glucose calibration models. Models based on three PLS factors were typically found to be optimal.
Standard errors of calibration and prediction for these models after optimization of the bandpass filter wave-length and width and the region of the interferogram being analyzed ranged f rom 0 . 4 to 0 . 5 mM .
Direct analysis was performed on interferogram data for mixtures of glucose, bovine serum albumin (BSA) and triacetin in pH 7 . 4, 0 .1 M phosphate buf f er. This syn-thetic matrix provides many of the ~hAllpnqes of an actual biological matrix, while offering control and reproduci-bility. In particular, this matrix is challenging because both BSA and triacetin have siqn; f i c~nt absorption in the 4000-5000 cm l range.
As in the analysis of glucose in phosphate buffer described in Example 4, multivariate calibration models were nPcP~s~ry to account for the presence of multiple absorbing species. In addition, utilization of multiple filters centered on different parts of the spectral region of interest was found to be effective.
Specifically, filters of various widths which were centered on the glucose band at 4400 cm-l, the BSA band at 4600 cm l, and the triacetin band at 4440 cm l were studied.
Each filter was applied to the raw interferogram, produc-ing one, two or three filtered interf~ro~L~a lPpPn~;n~ on the number of f ilters used. Segments of these f iltered interferograms were then used together in PLS analysis W0 95/22046 PCr/usss/olss6 218308~ ~
Fig. 12 displays the data obtained from applying three distinct filters to an interferogram obtained from a glucose/BSA/triacetin sample. Segments of 1000 points from each of the three filtered interf~Lu~Lcu;.a are 5 plotted, with the segments obtained by use of the glucose, BSA and triacetin filters being plotted as points 1-1000, 1001-2000 and 2001-3000, respectively. Shorter segments of each of these regions (denoted by the arrows in Fig.
12) were then ~nvestigated.
For cases using two and three f ilters, optimization of eight and twelve experimental variables is required, respectively. For each filter, the optimal bandpass position and width must be found, as well as the optimal interf erogram segment location and length. This process 15 involves repetitive calculations on standard data sets and evaluation of which factors provide the best ultimate calibration set.
Fig . 13 plots the 8 tandard error of calibration (solid line) and prediction (dashed line) versus the 20 number of PLS factors used to form the calibration model in a two-filter analysis based on interferogram points 340-640 for the glucose filter and points 200-301 for the BSA filter. The optimal model size in this case is 14 terms. This model exhibited an R2 value of 98.4596, and 25 standard errors of calibration and prediction of 0 . 704 and 0. 841 n~, respectively.
Fig. 14 is a correlation plot of glucose concentra-tions estimated using this 14 term calibration model versus the corresponding actual c~nc~on~ations. Calibra-30 tion samples are indicated by closed circles, whileprediction samples are indicated by open triangles.
Inspection of this data reveals that the prediction data falls within the span of variation of the calib-ation data o gs/22046 ~ 1 ~} 3 0 8 ~ PCrlUS95101556 and that a strong correlation exists between estimated and actual concentrations.
Examination of residuals from the model reveals no relationship between residuals and either BSA or 5 triacetin concentration. There is some evidence of non-linearity in the plot of residuals versus estimated concentrations, however. This suggests that the presence of signals from multiple absorbing species within the filter bandpass has violated the assumptions of the linear approximation equation set forth above.
The use of multivariate calibration models provides the ability for the model itself to extract signals from overlapping information in the interferogram. This raises the question as to whether the digital filtering step is necessary. To address this question, an optimization was conducted to identify the best segment in the unfiltered spectrum for use in building a multivariate calibration model. The best model found was based on interferogram points 100-450 and included 15 PLS factors. Fiqure 15 is a correlation plot analogous to Figure 14 for this model.
As can be seen, the standard errors of calibration and prediction have both increased by a factor of approxi-mately three, and significantly greater scatter exists in the data points. Thus, it appears that the filtering step is essential to obtaining good calibration models from the i~terf erog~ t~ .
PHYSIO~.OGICAL rT~FMTt~T q, pARTTCITT ~RrY GLUCOSE
The invention described in this application has been developed in part with funds received from the US National Institutes of Xealth under grant num.ber DR 45126. The United States GL~V~ ~ t may have certain rights under this invention.
BACRGROUND OF T~E lNV~ ~liJN
This application relates to a method and apparatus for the non-invasive, quantitative meaaurement o~
physiological rhPmirll~, particularly glucose, in a patient.
Determination of blood glucose is a routine proce-dure performed several times a day by many diabetics. In general, this procedure involves the taking of a small blood sample and evaluating the level of glucose in the sample. Common instL, t~ used for this purpose use the enzyme glucose oxidase to convert glucose and oxygen to gluconic acid and hydrogen peroxide, and then measure the level of peroxide by either spectroscopic or electrochemical means.
While these daily measurements provide the diabetic patient with the ability to self-monitor and thus better control blood glucose levels, they are not without draw-backs. In particular, the taking of blood sam.ples several time6 a day can be painful and exposes the patient to a risk of infection. Moreover, using this existing technology it is impossible to obtain a nntin-louS blood glucose mea~urement. Thus, during the night, a patient must either be awakened periodically for testing or run the risk that glucose levels will drop to dangerous levels 3 0 as they s leep .
Wo 95/22046 , - ~ PCT/usss~0lss6 ~
2183~85 In view of the foregoing, there exists a real need for a non-invasive method of measuring blood glucose in a patient. It has been suggested that this could be accom-plished using near-infrared (NIR) radiation. Thus, U.S, 5 Patent No. 5,086,229 of Rosenthal, which is incorporated herein by reference, de3cribes a system using a plurality of in~rAr~-l light emitting diodes and a detector to measure blood glucose. The infrared measurement of glucose in the body is immensely complicated, however, 10 becau6e of the suhstantial ab~orhAnnes of fats and proteins in the near-infrared. Thus, while the Rosenthal patent presents an interesting theory, there is no pl-hl; ch~d documentation that demonstrates the ability of the Rosenthal system to measure glucose noninvasively.
15 Furth~ ~~e, the wavelengths used in the Rosent_al syatem have never been shown to possess useful glucose inf ormation .
It i8 an obj ect of t_e present invention to provide a practically useful non-invasive near-infrared detector 20 for blood glucose and other physiological chemicale that is adaptable for continuous use.
It i8 a further object of the present invention to provide a method for the non-invasive quantitative measurement of physiological chemicals and particularly 25 glucose in a patient that can be used either nn~;nn~usly or intermittently.
It is still a further object of the invention to provide an apparatus for cnn~;nl~nus or intermittent non-invasive monitoring of hlood glucose and other 30 physiological chemicals in a patient.
~UMMARY OF THE LNV~ ~N
In accordance with the invention, non-invasive measurements of physiological chemicals such as glucose ~, WO 95/22046 2 1 8 3 0 ~ 5 PCTIU595~0155C
are made using infrared radiation and a signal processing system that permitæ the co~struction of a device suited for home use. Thus, the level of a selected physiolo~;cAl chemical in a test subject is detPrmlnp~ in a non-invasive 5 and quantitative manner by a method comprising the ateps of:
(a) irradiating a portion of the test subject with near-infrared radiation such that the radiation is transmitted through or reflected from the te6t subject;
(b) collectinq data conf Prnin~ the transmitted or reflected infrared radiation using a detector;
(c) digitally filtering the collected data to isolate a portion of the data indicative of the physiological chemical; and (d) detPrm;ning the amount of physiological chemical in the test subject by applying a defined mathe-matical model to the digitally filtered data. I'he data collected may be in the form of an Ahsorh~nme spectrum, i.e., a representation of the variation in absorbance as a function of wavelength; or in the form of an interfero-grAm, i.e., a representation of infrared light intensity as a function of the retardation of the moving mirror in the interf erometer, and thus of time .
In a further aspect of the invention, there i6 provided a device for measurement of a chemical in a sample comprising:
(a) means for collecting data con,Prnin~ near-inf rared radiation transmitted through or ref lected f rom the sample;
(b) means for digitally filtering the collected data to isolate a portion of data indicative of the chemi ~al;
W0 95/22046 PcrJuS9S/01556 218308~ --~, 4 (c) means for app1ying a definèd mathematical model to the digitally filtered data, whereby the amount of chem.ical in tbe test subject is deT-P~min~d; and (d) means for reporting the amount of chem.ical in the sample. This device can be incorporated into a complete apparatus for the non-invasive mea~ul, t of blood glucose. Such an apparatus would also include a source of infrared radiation and a t~hAn~ pm for directing the infrared radiation from the source to the test subject and from the test subject to the means for collecting the spectrum.
BRIEF DESCRIPTION OF THE nRPT~r~N~.q Fig. 1 is a flow chart depicting the method of the invention;
Fig. 2 is a flow chart depicting a process for digitally filtering an absorbance spectrum;
Fig. 3 is a schem,atic representation of an interf erometer;
Fig. 4 shows a sam.ple interferogram;
Fig. 5 shows a device in accordance with the present invention for measurement of glucose in a sam.ple;
Fig. 6 shows an apparatus in accordance with the present invention for non-invasive monitoring of physiological chemicals in a test subject;
Fig . 7 shows unf iltered interf erogram segments taken of phosphate buffer and three concentrations of glucose;
Fig. 8 shows digitally filtered interferogram segments taken of phosphate buffer and three concentrations of glucose;
Fig. 9 shows the relat~nQhl~ between filtered interferogram segment magnitude and actual glucose concentration;
Wo 95l22046 PCTIUS9S~OISS6 2183085 -= -Fig. 10 shows the relationship between glucose concentration de~rmin~d from digitally filtered - absorbance spectra of hu~n test subjects and actual glucose levels as de~ n~d by blood testing;
5 Fig. 11 shows the relationship between glucose concentration detPrm;n~d from unfiltered absorbance spectra of human test subjects and actual glucose levels as det~ined by blood testing;
Fig. 12 displays the data obtained from applying three distinct filters to an interferogram obtained from a glucose/BSA/triacetin sample;
Fig. 13 plots the standard error of calibration (solid line) and prediction ~dashed line) versus the number of PLS f actors used to f orm the calibration model of a filtered interferogram in a two-filter analysis based on interferogram points 340-640 for the glucose filter and points 200-301 for the BSA filter;
Fig. 14 is a correlation plot of glucose concentrations estimated using a 14 term calibration model versus the corresponding actual concentrations;
Fig. 15 is a correlation plot of glucose concentrations estimated using a 15 term calibration model for an unfiltered interferogram versus the corresponding actual concentrations;
nT~ T.Fn DESCRIPTION OF THE ll`JV~ 'lU~!I
Fig . 1 is a f low chart depicting the method of the invention, as applied to the non-invasive determination of blood glucose in a human patient. As shown, the first step of the method is the irradiation of the test subject, in this case the patient, with near-infrared radiation.
Suitable NIR radiation for use in the present invention ct i nrit~P5 with the absorbance bands of glucos~
or other physiological chemical being measured. For Wo95/220~6 r~ 1556~
21~308S
glucose, these bands are located in the regions of 5000-4000 cm l and 6500-5800 cm l. The intensity of the light at the selected wavelen~th should be on the order of 700 lux or greater. Such radiation can be produced by a 300 W
5 tungsten-halogen lamp.
9ince the spectrum obtained and used in the method of the present invention may be either a transmittance spectrum or a ref lectance spectrum, considerable latitude is available in the manner and location in which the NIR
lO radiation impinges on the test subject. For example, if transmitted NIR radiation is being measured, the NIR
radiation should impinge on a relatively thin, fleshy area of the patient such as the f leshy webs between the f ingers or toes or the ear lobe . If ref lectance spectra are to be 15 used, the sampling site should be characterized by high blood flow close to the surface, such as the ventral surface of the wrist.
The source of NIR radiation used in the present invention may be such that it is disposed directly against 20 the surface of the test subject. For example, a small halogen lamp could be used. Alternatively, the source may be physically remote from the test subject. In the latter case, it is advantageous, although not n~cP~s~ry, to guide the NIR radiation to the desired irradiation site on the 25 surface of the test subject, for example by means of optical f ibers .
In the second step of the method, the data concern-ing the transmitted or reflected NIR radiation is collected using a detector. The specific nature of the 30 detector is not critical, provided it is capable of detecting the pertinent wavelengths of light and respond-ing rapidly enough to be com~oatible with the other compon-ents of the device. An example of a suitable detector for collection of an absorbance spectrum is a combination of a 0 95J~2046 ~ ~ 8 ~ F~ ~ 556 dispersive element, e.g., a grating or prism, and an optical multi-channel analyzer sensitive to NIR radiation.
In the case where the data is to be collected as a single beam spectrum, an absorbance spectrum or an interf erogram, 5 a suitable detector is a combination of a NIR interf ero -meters and a photon counting detector such as a solid state indium an~ detector.
The positioning of the detector relative to the test subj ect will depend both on the nature and size of the 10 detector and the environment in which the mea~uL~ ~ is being taken. For most purposes, it will be desirable to have the detector physically separated from the test subject, both because of dPtector size and to m~~;m;7e detector perf ormance by providing the detector with a 15 stable environment. It will therefore generally be advantageous to guide the transmitted or reflected NIR
r~ ;on to the detector, for example using optical f ibers .
Depending on the instrumentation selected, the data 20 corl~ern;n~ the transmitted or reflected radiation is collected as either an absorbance spectrum or an interferogram. In either case, the next step as shown in Fig. 1 is to digitally filter the data to isolate the portion of the data which is indicative of the chemical of 25 interest. The specific manner in which this step ig perf ormed depends on the f orm of the collected data .
.
Digital F; 1 ter;n~ of an Absorbance ~pectrum When the data collected is in the f orm of an absorbance spectrum, the process for filtering the data is 30 shown in Fig. 2. As shown, the absorbance spectrum, or a portion thereof containing information about the chemical being tested for, is transformed usinq a Fourier transform into a Fourier domain spectrum Wo gs/22046 ~ 1 8 3 0 8 ~ PCrlUS95/OlSS6 ,~
The transformed spectrum is then filtered by multi-plying it by a Gaussian function that weights spectral information associated with the chemical being analyzed.
In the case of glucose, a suitable Gaussian function i8 5 aefined by a mean position of 0.037 f (digital frequency units) and a standard deviation width of 0.011 f. This function emphasizes a glucose-associated ~h50rh~nf~e band appearing in the spectral region between 6500 and 5800 cm 1. Other peaks, and other Gaussian functions for the same 10 peak may also be employed. For example, the glucose-zlssociated absorbance band at 4400 cm l can be emphasized.
For physiological chemicals other than glucose, appropriate ~P~Asi~n functions may be i(qPntifi~ For absorbance spectra processing, the proper Gaussian 15 function can be obtained by analyzing the shapes of the ~hsorh;~nce bands for the compound of interest.
After multiplication of the spectrum with the Gaussian function, the resulting filtered spectrum is converted back to an absorbance spectrum using an inverse Fourier 2 0 transf orm.
While the use of a Fourier filtering process such as that described above is a preferred method of performing the method of the invention, other digital filtering methods may also be used. Such methods include finite 25 impulse response and inf inite impulse response digital f ilters .
Digital Filtering of a Single Beam Spectrum or an Interf erogram - =
An important aspect of non-invasive measureme~t of 30 glucose and other l~hysiological chemicals using interfero-grams is the ability to analyze this data in the absence of a contemporaneously collected background or ref erence spectra. This is the case because no convenient method ~ WO 95/220~16 2 1 ~ 3 Qg S PCT/US9SJ01556 ~ g e~cists to collect a background spectrum such as the spectrum of a particular human subject without glucose.
When the data c~)nClorn; n~ the transmitted or ref lected inf rared radiatio~ is collected in the f orm of 5 an interferogram, it may be processed in either of two general ways. First, the interferogram may be converted to a conventional single beam spectrum using a Fourier transform and then filtered by multiplication with an ayylUyliate Gaussian function. Fûr example, in the case of a non-invasive determination of glucose, a Gaussian function having half-power points of 4445.4 and 4354.6 cm-can be used to emphasize the glucose absorbance band at 4400 cm~l. This corresponds to a ~All~SiAn furlction having a mean position of 0.139 f and a standard deviation of 0.00122 f. Other Gaussian functions for digital filtering to obtain information from other glucose absr~rhAnc~ or for absorption bands associated with other chemicals of interest can be obtained by matching functions to the width and location of the absorption bands.
This filterins differs from the process described above, in that ratioing to eliminate the background signal is not performed. secause of this, after multiplication of the transformed :iueU~Ll the resulting filtered spectrum can be subjected to an inverse Fourier transformation to yield a filtered interferogram. Such an interferogram is much simpler in Arp~ArAnce than the original interferogram, because much of the noise and absorbances not associated with the chemical of interest have been removed. Further, there is a clear correlation between the magnitude of the interferogram signal and the amount of the chemical being measured in the test subject.
The magnitude of the interferogram peaks is advan-tageously measured by the simple procedure of integrating the interf erogram to determine the total power of the Wo gsl22046 , ~ p ~ . lS56 ~
signal. ThiG yields a single value related to the glucoae concentration. Alternatively, the individual points in the filtered interferogram could be submitted to a multivariate calibration method such as a partial least 5 squares regression. This would yield a multivariate del relating the individual interferoqram point intensities to glucose concentration.
Alternatively to transforming the interferogram into a single beam spectrum, the interferogram itself may be 10 subjected to a digital filtering process, and the magnitude of the filtered interferogram peaks analyzed using a prP~iPfinP~l mathematical model. We have defined a model which predicts an approximately linear relatinn~hip between analyte co~centration and f iltered interf erogram 15 intensity as given by the formula C NP
where C is the concentration, P is the magnitude of the filtered seqment, and N is a sensitivity term defined as N=
Z [k] Z
~ k=nl where Z [k] is an interferogram domain function. The approximate linear rela~;c.n~h;rl between the analyte 20 concentration and magnitude of the filtered interferogram segment is based on the assumDtion that a NacLaurin series approximation used in the derivation of the relati~n~hir is valid. This approximation only holds for small absorbance values. For example, Sparks et al., Anal.
25 Chem. 54: 1922-1926 (1982) have estimated that this ~ Wo 9sl22~)46 218 3 0 8 S PCrmss5/DI556 approximation has less than 596 error if the absorbance value is less than 0.12. For glucose analysis using the present invention, peak absorbance of the glucose band is on the order of 10 4 or less. Thus, the use of this 5 approximation appears valid.
The approximate linear rela~inn~h;r also depends on the value of N being substantially constant across the calibration samples. This re~auires that the sample pathlength, molar absorptivity and the intensity of the filtered reference spectrum be constant with respect to concentration .
Variation in sample pathlength is not a problem in laboratory measurements due to the use of a fixed pathlength sample cell. It could present more of a problem in actual noninvasive tests performed on body tissue. Moreover, since the entire spectral band is isolated by the bAn~rA~s filter rather than a single vav~ ' cr, some change in molar absorptivity might be expected across the concentration range. These two factors could lead to limitations on the dynamic range over which the linear relationship would hold, potentially making several calibration factors necessary. They should not, however, prevent the use of the linear approximation for making noninvasive glucose determinations.
The assumption of constant intensity of the filtered reference spectrum i8 problematical, as some variation in infrared intensity is common in a Fourier transform inf rared PYrPr; t due to changes over time in instl, ~_A1 characteristics such as interferometer alignment and detector temperature. These factors can be uv~ through the use of a nn~^l; 7ation step.
Finally, the linear relatinncll;~ set forth above assumes that only a single analyte band has been isolated by the bandpass filter. If other analyte bands or bands wo 95/22046 PCT/USg5~01556 ~g3~5: ~
from other constituents in the sample matrix fall within the filter bandpass, their interferogram signatures will add to the signature of the targeted analyte band. The filtered interferogram will then possess a more complex 5 structure and the simple univariate linear relat;on~h;p will not apply. The selection of an appropriate bandpass filter is therefore quite important to the succes6 of the direct measurement of interf erogram inte~sity as a method for noninvasive determination of physiological ,h.omi 10 If suitable filters cannot be identified for a given analyte and sample matrix, however, a multivariate analysis procedure can be used on place of the linear approximation as discussed in 13xamples 4 and 5 herein.
Application of digital filtering directly to the 15 interferogram depends on the fact that each wavelength of light is represented in the interferogram by a modulated sine wave whose frequency (in hertz) depends on the wave-length of the light and the interferometer mirror velocity in accordance with the equation frequency ~vd v, ' ^r mirror ofsignal 2XofradiationXvelocity 20 where the ~c-v. ' cr has units of cm l, the mirror velocity is expressed in units of cm/sec. By selecting an electronic f ilter that selects out only the portion of the interf erogram having the f requency def ined by this equation f or the selected absorbance of the chemical being 25 meagured, a simplified interferogram is obtained.
The construction of electronic filters to select for specified wavelengths is ~ ce in the electronic arts. Briefly, the filter would be implemented as a digital filter. The software r~n~rQll ;n~ the device would 30 implement the filter as part of the data processing. The ~ Wo 95~22046 2 1 8 3 0 8 $ PCTrUSs5/~ 56 digital filter used implements an approximation of the convolution sum y[n] = ~ x[k] h [n-k]
k=-~
' ;' where y[n] is the filtered interferogram at point n, x[k]
is the original unfiltered interferogram at point k, and h [n-k] are points on the impulse response of the digital filter, i.e., the interferogram-domain representation of the frequency response filter.
While the digital filtering procedure for interfero-grams has been described above in terms of a Fourier filtering process, other digital f;ltl~r;n~ methods may also be used. In particular, methods such as f inite impulse response and inf inite impulse response f ilters can be used. These filter design methods approximate the infinite summation in the convolution sum above. ~hese methods of f er the advantage of reducing the size of the interferogram required, which can lower the performance requirements of the interf erometer .
While the method of the invention can be practiced on complete interferograms, it has been found that short segments of an interferogram can be used as well. This has great significance to the application of this terhn~l ogy in relatively rigorous environments, because it facilitates the construction of a smaller interferometer with less stringent mechanical tolerances.
An interferometer such as that shown in Fig. 3 includes a 50: 50 beam splitter 30, two mirrors 31 and 32, and optical track 33 and a detector 34. Infrared radiation is passed through the beam splitter 30 and wo gsl22046 ~ . Pcr/usss/olss6 21~3~8~ --directed toward the two mirrors 31 and 32, oriented at 90 degrees to one another, which reflect it back towards the detector 34. At the same time, the mirror 31 is moved back and forth on the optical track 33. The movement of the mirror 31 results in sUccessive occurrences of cons tructive and des tructive interf erence . Thus, the signal received at the detector is a harmonic signal whose f requency is determined by the f reguency of incident light and the mirror velocity. A sample of the detected signal, called an interferogram, is shown in Fig. 4.
InterfeLuSIL can be converted by a mathematical transform into conventional spectra of very high quality.
~lowever, the mechanical ~ r; ty of the interferometer, and particularly the need for precise control of the move-ment of the mirror over a substantial length of travel and for precise optical ~ t, limits the ap~?licability of interferometers in rigorous environments. The mechanical complexity of the instrument could be substantially reduced if the length of the interferogram were shortened, because the length of mirror travel could be shorter. It is generally recognized, however, that the use of short interft,Lu~Lcu~s significantly reduces the quality of the spectrum which can be obtained upon transformation of the interf erogram to a conventional absorbance spectrum.
The reason for this loss of quality has to do with the mathematical approximation normally used to perform the transformation, the fast Fourier transform or FFT.
The FFT is an approximation of the infinite Fourier integral in which it is assumed that the time signal is zero outside the region actually sampled. Supplying a very short interf erogram to the FFT has the ef f ect of convolving a boxcar function with the true (i.e.
infinitely sampled) interferogram. The Fourier transform of the boxcar function is a (sin x) /x function. When this ~ W0 95/22046 2 1 ~ 3 ~ 8 ~ ss6 function i5 superimposed on the true 6pectrum, a lowering of resolution results.
In conventional interferometers in use today, the total travel of the mirror is around 2 cm, with spectral 5 collection actually oCcllrri n~ over approximately 1 cm of this length. It has surprisingly been found, however, that short interferoqram segments, i.e., interft~ Lcu.~
collected over distance of mirror travel of about 0 . 003 to 0.10 cm can be used effectively in the method of the 10 present invention. This opens the door to the cons truction of interf erometers that are smaller and inherently more rugged than their laboratory predecessors, and which thus are suited for use, e.g., by an individual diabetic in the home.
A further advantage of the method as of the present invention is ability to dispense with separate background spectral measurements through the use of interf ~:ru-JLeu..s .
Such measurements are necessary in most systems to correct for instrumental variations and the absorbance of com-20 pounds other than the ~ , ulld of interest. In the case of non-invasive monitoring of physiolo~;cAl chemicals, however, this background is highly complex. Fur~h~ e, it is not readily apparent how one would obtain a spectrum of the test subject, minus the chemical being analyzed.
25 When using interfeL~,yL~.~, however, no background is nPcpsE;lry. This is the case, because in an interferogram, the portion of the signal which can be attributed to the overall instrument function is a wide feature which is represented in the interf erogram as a rapidly damping 30 sinusoidal signal. In contrast, absorption bands of compounds are narrower and thus their interf erogram repre-sentation does not damp as rapidly. Thus, by taking the analyzed data from an interferogram segment removed from the centerburst, the instrumentation flPrPnflPnt signal has Wo 95/22046 -~ 3 0 8 ~ Pcrluss5/01ss6 essentially died out, leaving only the signals associated wi th actual inf rared absorbances .
Regardless of the form of the collected data, the final step in the method as ~hown in Fig. l is the 5 determination of glucose or other physiological chemical concentration by application of a previously detprm; nPd mathematical (calibration) model. In either case, the input into the model is the filtered data, and the output is the estimated concentration of glucose or other lO physiological chemical of interest. The model is aeveloped based upon a multivariate regression analysis, such as partial least squares (PLS) regression. In PLS, the in'lPpPn~lPnt VAr;AhlP~ are the filtered data and the dPrPn~lPnt variable is the glucose concentration. Other 15 multivariate regression methods such as principal _ onPnts reqression can also be used.
The mathematical del resulting from application of any of these techniques to a set of data in which the actual glucose cr~nrPn~r~tions are known can be expressed 20 as a series of term3 the value of which correspond to measured properties of the data for a sample, and factors or regression coefficients. For example, a model found to be useful in dete~ninin~ glucose conrPntrations in vivo using data in absorbance spect~um form has the form c j =bo +blx~, l + - +bhX~, h 25 where c1 is the predicted glucose concentration corres-ponding to spectrum i, the x1 values are the partial least-3quares factor scores computed from the absorbance data, and the b term3 are the coefficients detPrm;nPd from a multiple linear regression analysis of the measured 30 glucose concentrations corresponding to the calibration .
~ W0 95/220~6 ~ 1 8 3 Q 8 5 1~ 015~6 absorbance data and the h sets of partial least-sguares scores computed from the calibration absorbance data.
To analyze an unknown sample, values measured for the sample are substituted into this equation to produce - 5 an estimate of the glucose concentration in the sample.
As a general rule, device limitations will probably reguire that this model be established for each chemical being measured using instrumentation separate from the device actually being used by a test subject to perform day-to-day nitoring. Nevertheless, some calibration function may be included in an individual device to adapt it to a specific test subject.
While the method of the invention has been described prinr;r~lly with respect to the monitoring of glucose in a human subject, the scope of the invention is not so limited. Physiological chemicals other than glu-cose, such as urea, lactate, triglycerides, total protein and cholesterol may also be guantitatively detected using the method of the invention through the selection of a~Lu~Liate digital filtering p~ tr-rs. Fur~hr ~L~, because the filtering is all post-collection processing performed on stored data, a single spectrum can be anal-yzed numerous times to provide analytical results on re than one physiological chemical.
Test subjects to which the method of the invention may be applied include not only human subjects, but also animal subjects, e.g., during veterinary procedures, and even living plants . For example, non- invasive determina-tion of sugar levels within fruits could be made to assess the ripeness of the fruit and its reP~in~s for harveat-ing . Thus, as used in the specif ication and claims here -of, the term test subject refers to any living organism contalning physiological chemicals.
WO 95/22046 2 1 8 3 0 8 5 PCI/US95/01556 ~
A further aspec't of the E~resent invention is a device for measurement of glu~cosb~'in a sample as shown in Fig. 5. This device, whic~ can be used on any type of sample, not just a living organism, includes means for 5 collecting a transmitted or reflected infrared radiation such as interferometer 50 or a combination of a dispersive element and an optical multichannel analyzer. The collected spectral data is transferred to a computer read-able storage medium 51 from whence it is read by computer 10 processor 52 and processed in accordance with a ~L~yL -~instruction to digitally filter the collected ~e- LL~-- to isolate a portion of the spectrum indicative of glucose.
9uitable computer ~rocessors for use in this application include Intel 486 and PentiumlD microprocessors or MIPS
15 R3000 and R4000 processors of the type used in Silicon Graphic8 workstations.
The computer program used to perform the digital filtering steps can be written in any pLUyL ;nq language capable of performing a Fourier transform and multiplying 20 a data set by a Gaussian function. For example, a program written in Fortran 77 was employed to perform the data analysis described in the example herein.
Once the stored data have been digitally filtered, the computer processor 52 is operated under a second set 25 of LJlOyL -d instructions in order to apply a predeter-mined mathematical model to the data. The result of this application i8 a numeric estimate for the cr~nc~ntr~tion of glucose or other chemical of interest in the test subject.
The device of the invention also includes one or 30 more means for reporting the amount of glucose in the sample . Examples of means f or reporting the amount of glucose include digital display panels 53, transportable read/write magnetic media such as computer disks and tapes which can be transported to and read on another machine, ~ Wo 95/22046 and printers such as thermal, laser or ink- jet printers f or the production of a printed report .
A further aspect of the present invention is an apparatus for the non-invasive, quantitative detection of 5 physiological chemicals in living organisms. Such an apparatus may be a combination of a device for glucose measurement as described above, or a comparable device having the digital filter parameters and standard values targeted for a different physiological chemical with a 10 means for irradiating the surface of the living organism with near-infrared radiation in such a way that the spectral data can be collected in a transmittance, diffuse reflectance or ~r~n~flectance configuration.
Fig . 6A shows an apparatus of this type f or collect -15 ing spectral information. The apparatus has a housingwhich encloses a light source 61, a detector 62, and a microprocessor 63. Infrared radiation from light source 61 is transmitted through optical fiber 64 to the test 6ubj ect 65 . Transmitted radiation i8 collected through 20 optical fiber 66 and transmitted to detector 62. The 6ignal from the detector is then transmitted to micropro-cessor 63. Microprocessor 63 digitally filters the signal from the detector, and applies the pre~lef;~Pd mathematical model to the filtered data to determine a value for glu-25 cose concentration which is displayed on display 67.Display 67 may be an LED or LCD display.
Other conf igurations of the apparatus can be util -ized in addition to the specif ic conf iguration shown in Fig. 6A. For example, the apparatus can be configured to 30 collect infrared radiation in a diffuse reflectance mode, rather than in a transmission mode. In the diffuse reflectance mode, a portion of the incident radiation penetrates a short distance into the sample before being scattered out of the sample. A fraction of this back-Wo 95~2046 2 1 8 ~~ PCr/uss5/ol556 ~
scattered radiation is collected and measured by the detector .
The apparatus may also be operated in a transf lec -tance mode. In a transflectance configuration, optical 5 fibers 64 and 66 are positioned apart by a distance D as shown in Fig . 6B . The second f iber collects light that has entered the sample and traversed a short distance within the sample medium.
In vitro measurements of glucose concentration were made on a series of thirteen glucose samples having nt~ations varying from 1.25 to 19.7 ~I glucose in pH
7 . 4 phosphate buf f er ( 0 .1 M) . The samples were placed in 1 mm pathlength cuvettes constructed from infrared quartz 15 and transmittance interfelu,~L of each were obtained using a Nicolet 740 spectrometer in conjunction with a tungsten-halogen light source and a cryogenically cooled InSb near-infrared detector. Double sided interfeLo.~L~,~
were collected by averaging across 256 co-added scans.
20 Either two or three replicate interfeLu~L~.~ were made for each sample, producing a total of 38 interferograms across the 13 samples.
To evaluate glucose concentration, the glucose absorption located at 4400 cm l was used and i601ated by 25 digital filtering. The filtering process involved application of a fast Fourier transform to the interferogram to produce a single beam ~ec~L,;
multiplication of the transformed interferogram with a Gaussian function having half -power points at 4445 . 6 and 30 4354.6 cm l; and application of the inverse fast Fourier transform to obtain the filtered interferogram.
Fig 7 shows four interf~LoyLcu.s segments (points 600-800 relative to the interferogram peak maximum or ~W095l22046 2~ dg~'`'`'`' PcTll~S9SlOlSS6 centerburst) obtained for phosphate buffer, 3 .22 ~
glucose, 7 . 95 }[~q glucose and 19 . 7 mM glucose . No obvious information relating to glucose concentration can be observed. Fig. 8 shows these same interfe:LuyL~Ls after - 5 the digital filtering procedure described above. The waveforms have been greatly simplified due to the removal of all frecluencies outside the bandpass of the Gaussian function. The interferogram segment for the buffer de-creases in intensity to a minimum point or node, followed by a slowly damping signal due to the water absorption feature. The interferograms of samples cont~;n;n~ glucose exhibit a less obvious node, but increased magnitudes are observed with increasing glucose co~c~ntr~tions.
Fig. 9 shows the results obtained when the magni-tudes of the filtered interfelu~L~,~ are plotted as a function of actual glucose concentration. An excellent correlation between the magnitude and the concentration is observed (Univariate linear regression: R s~uared=99.0%;
standard error of calibration = O . 62 mM) .
Ill vivo experiments on the non-invasive detection of glucose were performed using a Nicolet 740 research grade spectrometer configured with a 250 W tungsten halogen lamp, CaF~ beam splitter and InSb detector . Non- invasive measurements in accordance with the invention were taken at the same time as conventional invasive measurements on three individuals of varied physiognomy during the course of glucose tolerance tests. The beam of light from the spectrometer having a spectral range from 7000 to 5000 cm was passed through a portion of the webbing between the thumb and forefinger~ This particular ~v~v~ ' cr region was isolated using a standard astronomical H-band optical f ilter .
WO 95/22046 2 1 8 ~ 8~ PCT/US95/01556 ~
~ .i ,..
The spectra were collected in interferogram form using the interf erometer and InSb detector of the spectro -meter, and then converted to single beam spectrum ~intensity versus ~ V~ ' ~r) using the Ar- _-nying 5 Nicolet computer. All single beam spectra` were subsequently transf erred to a Silicon Graphics Indigo computer for data processing.
Blood samples taken during the test were analyzed for qlucose using a Yellow Springs Instruments (YSI) model 10 2300 glucoGe analyzer.
To process the spectra, the single beam spectra were f irst converted to absorbance units by dividing each spec-trum by a water re~erence spectrum and then computing the neqative logarithm. Only the portion of the spectrum 15 between 6500 and 5800 cm l was used in all subsequent calculations .
Spectra were then normalized using the signal between 5955 and 5951 cm~l, a ~vc.vl ` ?r region that contains no glucose inormation. The llnr~ l; 7ed spectra 20 were then digitally Fourier filtered to ~l iminAte noise and other spectral features that distract from glucose inf ormation .
The first step in the Fourier ~iltering process was to perform a Fourier transorm on the absorbance spectrum.
25 This resulted in a Fourier domain spectrum, which was then multiplied by a Gaussian function defined by a mean position of 0 . 037 f and a standard deviation width of O . 011 f. Thesc parameters were selected to isolate the glucose absorption. The filtered transformed spectrum 30 was then converted back to an absorbance spectrum by applying an inverse Fourier transform.
The spectra obtained in this way were analyzed using the standard PLS regression procedure. This procedure essentially goes into the data set and weights the 095/22046 ~ 8308$ pcrluss5/ol556 different spectral frequencies by correlating spectral variation with analyte concentration. The result is a - series of factors and regression coefficients which can be applied to individual spectra to obtain a predicted value 5 of glucose concentration.
Fig. 10 shows the correlation between glucose concentration as detPrm;nPd from the spectra and the glucose level as detPrm;nPc~ from the blood 6amples taken.
The open symbols in the graph represent the data that were 10 used to generate the calibration model. The solid symbols represent results from an independent set of spectra used only for prediction purposes.
A11 of the results cluster around the unity line, and the prP~1; ct; on points are within the general spread 15 and scatter of the calibration points. Thus, the ability of the method of the invention to predict blood glucose levels is conf; '. Moreover, since there is no clustering of data according to individual, a global calibration rather than individual calibrations may be 20 possible.
To evaluate the importance of the diqital filtering s tep in Example 2, the data taken in that experiment were reevaluated with the digital filtering step omitted and a 25 separate set of PLS factors were developed. The resulting plot of spectral glucose versus blood glucose measurements is shown in Fig. 11. As is a~parent, the data in this case has a much poorer corlelation. Thus, the importance of an appropriate digital f iltering technique is clear.
Wo 95/22046 PcrNS9Sl015S6 ~
~ 1 8 3 1~ 8 ~r 2 4 ~E~IPLE 4 To test the validity of the derived linear relation-ship between concentration and interferogram peak magnitude, a series of pr~l ;minAry experiments were 5 performed using solution sa~mples of benzene in CS2. CS2 has essentially zero absorbance in the region of the C-H
out-of-plane bending band of benzene at 1036 cm l that was used as the targeted band for the analysis. The linear approximation described above and a sim~le univariate 10 model were successful in relating the magnitude of the filtered interferogram segments to concentration.
A second series of tests were run using samples of glucose in phosphate buffer. The glucose C-H combination bAnd centered at 4400 cm~1 was used as the targeted band 15 for the analysis. This band is located on top of a broad background ~hssrhAnce of water spanning the 4000-5000 cm region, and thus presents a greater challenge since it violates the assumption made in deriving the linear rela~ic n~hir that only a single absorbing species would be 20 detected within the filter bandpass.
Calibration results made using the univariate linear approximation exhibited ~ubstantial errors. Since the information in the interfeL~L~w representing glucose and water should still be additive, however, it was 25 hypothesized that a multivariate calibration model could be used to correct for the interference due to water.
Partial least squares regression was used to build models of the f orm C=bo+blX1+- +b"x~
where C is glucose concentration and the n+1 bi values are 30 regression coefficients detPrm;n~d from a multiple linear regression analysis of the calibration data set. The independent variables, xl, are factor scores obtained from WO 95/22046 PcTn~ss5/0~556 the PLS analysis of the filtered interferogram segment.
The original independent variables used to obtain these latent variables are the intensities of the individual points in the filtered interferogram segment.
This approach was highly successful, yielding excellent glucose calibration models. Models based on three PLS factors were typically found to be optimal.
Standard errors of calibration and prediction for these models after optimization of the bandpass filter wave-length and width and the region of the interferogram being analyzed ranged f rom 0 . 4 to 0 . 5 mM .
Direct analysis was performed on interferogram data for mixtures of glucose, bovine serum albumin (BSA) and triacetin in pH 7 . 4, 0 .1 M phosphate buf f er. This syn-thetic matrix provides many of the ~hAllpnqes of an actual biological matrix, while offering control and reproduci-bility. In particular, this matrix is challenging because both BSA and triacetin have siqn; f i c~nt absorption in the 4000-5000 cm l range.
As in the analysis of glucose in phosphate buffer described in Example 4, multivariate calibration models were nPcP~s~ry to account for the presence of multiple absorbing species. In addition, utilization of multiple filters centered on different parts of the spectral region of interest was found to be effective.
Specifically, filters of various widths which were centered on the glucose band at 4400 cm-l, the BSA band at 4600 cm l, and the triacetin band at 4440 cm l were studied.
Each filter was applied to the raw interferogram, produc-ing one, two or three filtered interf~ro~L~a lPpPn~;n~ on the number of f ilters used. Segments of these f iltered interferograms were then used together in PLS analysis W0 95/22046 PCr/usss/olss6 218308~ ~
Fig. 12 displays the data obtained from applying three distinct filters to an interferogram obtained from a glucose/BSA/triacetin sample. Segments of 1000 points from each of the three filtered interf~Lu~Lcu;.a are 5 plotted, with the segments obtained by use of the glucose, BSA and triacetin filters being plotted as points 1-1000, 1001-2000 and 2001-3000, respectively. Shorter segments of each of these regions (denoted by the arrows in Fig.
12) were then ~nvestigated.
For cases using two and three f ilters, optimization of eight and twelve experimental variables is required, respectively. For each filter, the optimal bandpass position and width must be found, as well as the optimal interf erogram segment location and length. This process 15 involves repetitive calculations on standard data sets and evaluation of which factors provide the best ultimate calibration set.
Fig . 13 plots the 8 tandard error of calibration (solid line) and prediction (dashed line) versus the 20 number of PLS factors used to form the calibration model in a two-filter analysis based on interferogram points 340-640 for the glucose filter and points 200-301 for the BSA filter. The optimal model size in this case is 14 terms. This model exhibited an R2 value of 98.4596, and 25 standard errors of calibration and prediction of 0 . 704 and 0. 841 n~, respectively.
Fig. 14 is a correlation plot of glucose concentra-tions estimated using this 14 term calibration model versus the corresponding actual c~nc~on~ations. Calibra-30 tion samples are indicated by closed circles, whileprediction samples are indicated by open triangles.
Inspection of this data reveals that the prediction data falls within the span of variation of the calib-ation data o gs/22046 ~ 1 ~} 3 0 8 ~ PCrlUS95101556 and that a strong correlation exists between estimated and actual concentrations.
Examination of residuals from the model reveals no relationship between residuals and either BSA or 5 triacetin concentration. There is some evidence of non-linearity in the plot of residuals versus estimated concentrations, however. This suggests that the presence of signals from multiple absorbing species within the filter bandpass has violated the assumptions of the linear approximation equation set forth above.
The use of multivariate calibration models provides the ability for the model itself to extract signals from overlapping information in the interferogram. This raises the question as to whether the digital filtering step is necessary. To address this question, an optimization was conducted to identify the best segment in the unfiltered spectrum for use in building a multivariate calibration model. The best model found was based on interferogram points 100-450 and included 15 PLS factors. Fiqure 15 is a correlation plot analogous to Figure 14 for this model.
As can be seen, the standard errors of calibration and prediction have both increased by a factor of approxi-mately three, and significantly greater scatter exists in the data points. Thus, it appears that the filtering step is essential to obtaining good calibration models from the i~terf erog~ t~ .
Claims (19)
1. A method for non-invasive quantitative detection of a physiological chemical in a test subject comprising the steps of:
(a) irradiating a portion of the test subject with near-infrared radiation such that the radiation is transmitted through or reflected from the test subject;
(b) collecting data concerning the transmitted or reflected infrared radiation using a detector;
(c) digitally filtering the collected data to isolate a portion of the data indicative of the physiological chemical; and (d) determining the amount of physiological chemical in the test subject by applying a defined mathematical model to the digitally filtered data.
(a) irradiating a portion of the test subject with near-infrared radiation such that the radiation is transmitted through or reflected from the test subject;
(b) collecting data concerning the transmitted or reflected infrared radiation using a detector;
(c) digitally filtering the collected data to isolate a portion of the data indicative of the physiological chemical; and (d) determining the amount of physiological chemical in the test subject by applying a defined mathematical model to the digitally filtered data.
2. A method according to claim 1, wherein the digital filtering is performed by a Fourier filtering process.
3. A method according to claim 1, wherein the collected data is in the form of an absorbance spectrum.
4. A method according to claim 3, wherein the spectrum is digitally filtered by performing a Fourier transform on the absorbance spectrum; multiplying the transformed spectrum by a Gaussian function; and applying an inverse Fourier transformation to the multiplied transformed spectrum.
5. A method according to claim 4, wherein the mathematical model applied to the digitally filtered spectrum applies coefficients developed using a partial least squares regression multivariate calibration procedure.
6. A method according to claim 5, wherein the spectrum is digitally filtered by performing a Fourier transform on the absorbance spectrum; multiplying the transformed spectrum by a Gaussian function; and applying an inverse Fourier transformation to the multiplied transformed spectrum.
7. A method according to claim 6, wherein the spectrum is digitally filtered by performing a Fourier transform on the absorbance spectrum; multiplying the transformed spectrum by a Gaussian function; and applying an inverse Fourier transformation to the multiplied transformed spectrum.
8. A method according to claim 7, wherein the spectrum that is digitally filtered spans the region from 6500 to 5800 cm-1.
9. A method according to claim 8, wherein the magnitude of the digitally filtered spectrum is compared to the standard value using a series of regression coefficients developed using a partial least squares regression multivariate calibration procedure.
10. A method according to claim 1, wherein the collected data is in the form of a single beam spectrum.
11. A method according to claim 1, wherein the collected data is in the form of an interferogram.
12. A method according to claim 11, wherein only a short segment of an entire interferogram is analyzed.
13. A method according to claims 11 or 12, wherein the digital filtering step comprises the steps of performing a fast Fourier transform on the interferogram to obtain a transformed spectrum; multiplying the transformed spectrum with a Gaussian function to select for a peak diagnostic for the physiological chemical; and applying an inverse Fourier transform to the multiplied, transformed spectrum.
14. A method according to claim 13, wherein the digital filtering step comprises the steps of performing a fast Fourier transform on the interferogram to obtain a transformed spectrum; multiplying the transformed spectrum with a Gaussian function to select for a peak diagnostic for the physiological chemical; and applying an inverse Fourier transform to the multiplied, transformed spectrum.
15. A method according to claim 14, wherein the Gaussian function is selected to isolate the glucose-associated absorbance at 4400 cm-1.
16. A method according to claim 1, wherein the collected spectrum is an interferogram and the digital filtering is performed directly on the interferogram to select out waveforms within the interferogram having a frequency equal to where the wavenumber has units of cm-1, mirror velocity is the velocity of mirror travel in the interferometer used to generate the interferogram expressed in units of cm/sec.
17. A method according to any of claims 1-16, wherein the chemical is glucose.
18. A device for measurement of glucose concentration in a sample comprising:
(a) means for collecting data concerning near-infrared radiation transmitted through or reflected from the sample;
(b) means for digitally filtering the collected data to isolate a portion of data indicative of glucose;
(c) means for applying a defined mathematical model to the digitally filtered data, whereby the amount of glucose in the test subject is determined; and (d) means for reporting the amount of glucose in the sample.
(a) means for collecting data concerning near-infrared radiation transmitted through or reflected from the sample;
(b) means for digitally filtering the collected data to isolate a portion of data indicative of glucose;
(c) means for applying a defined mathematical model to the digitally filtered data, whereby the amount of glucose in the test subject is determined; and (d) means for reporting the amount of glucose in the sample.
19. An apparatus for the non-invasive measurement of one or more physiological chemicals in a test subject, comprising, (a) means for irradiating a portion of the surface of the test subject with near-infrared radiation such that transmitted or reflected near-infrared radiation is available for collection;
(b) means for collecting data concerning the transmitted or reflected near-infrared radiation;
(c) means for digitally filtering the collected data to isolate a portion of data indicative of the physiological chemical;
(d) means for applying a defined mathematical model to the digitally filtered data, whereby the amount of chemical in the test subject is determined; and (e) means for reporting the amount of chemical in the sample.
(b) means for collecting data concerning the transmitted or reflected near-infrared radiation;
(c) means for digitally filtering the collected data to isolate a portion of data indicative of the physiological chemical;
(d) means for applying a defined mathematical model to the digitally filtered data, whereby the amount of chemical in the test subject is determined; and (e) means for reporting the amount of chemical in the sample.
Applications Claiming Priority (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
US08/195,709 | 1994-02-14 | ||
US08/195,709 US5459317A (en) | 1994-02-14 | 1994-02-14 | Method and apparatus for non-invasive detection of physiological chemicals, particularly glucose |
Publications (1)
Publication Number | Publication Date |
---|---|
CA2183085A1 true CA2183085A1 (en) | 1995-08-17 |
Family
ID=22722457
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CA002183085A Abandoned CA2183085A1 (en) | 1994-02-14 | 1995-02-07 | Method and apparatus for non-invasive detection of physiological chemicals, particularly glucose |
Country Status (6)
Country | Link |
---|---|
US (2) | US5459317A (en) |
EP (1) | EP0745216A1 (en) |
JP (1) | JPH09509739A (en) |
AU (1) | AU703370B2 (en) |
CA (1) | CA2183085A1 (en) |
WO (1) | WO1995022046A1 (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US7711411B2 (en) * | 2002-10-25 | 2010-05-04 | Nir Technologies Inc. | Method of in-vivo measurement of fat content of a body and apparatus therefor |
Families Citing this family (136)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US6222189B1 (en) | 1992-07-15 | 2001-04-24 | Optix, Lp | Methods of enhancing optical signals by mechanical manipulation in non-invasive testing |
US5459317A (en) * | 1994-02-14 | 1995-10-17 | Ohio University | Method and apparatus for non-invasive detection of physiological chemicals, particularly glucose |
US5796476A (en) * | 1995-06-28 | 1998-08-18 | Kyoto Dai-Ichi Kagaku Co., Ltd. | Method of optically measuring component in solution |
US6240306B1 (en) | 1995-08-09 | 2001-05-29 | Rio Grande Medical Technologies, Inc. | Method and apparatus for non-invasive blood analyte measurement with fluid compartment equilibration |
EP0762109A3 (en) * | 1995-08-30 | 1997-09-17 | Kyoto Daiichi Kagaku Kk | Method of and apparatus for measuring lactic acid in organism |
US6040578A (en) | 1996-02-02 | 2000-03-21 | Instrumentation Metrics, Inc. | Method and apparatus for multi-spectral analysis of organic blood analytes in noninvasive infrared spectroscopy |
US5747806A (en) * | 1996-02-02 | 1998-05-05 | Instrumentation Metrics, Inc | Method and apparatus for multi-spectral analysis in noninvasive nir spectroscopy |
US5666956A (en) * | 1996-05-20 | 1997-09-16 | Buchert; Janusz Michal | Instrument and method for non-invasive monitoring of human tissue analyte by measuring the body's infrared radiation |
KR100376646B1 (en) * | 1996-07-01 | 2003-03-15 | 파마시아 에이비 | Delivery device and method for its operation |
US6544193B2 (en) * | 1996-09-04 | 2003-04-08 | Marcio Marc Abreu | Noninvasive measurement of chemical substances |
US5871442A (en) | 1996-09-10 | 1999-02-16 | International Diagnostics Technologies, Inc. | Photonic molecular probe |
US6594510B2 (en) | 1996-09-10 | 2003-07-15 | Xoetronics Llc | Photonic molecular probe |
US6340978B1 (en) * | 1997-01-31 | 2002-01-22 | Making Everlasting Memories, Ltd. | Method and apparatus for recording and presenting life stories |
TW352335B (en) * | 1997-03-25 | 1999-02-11 | Matsushita Electric Works Ltd | Method of determining a glucose concentration in a target by using near-infrared spectroscopy |
US5961451A (en) * | 1997-04-07 | 1999-10-05 | Motorola, Inc. | Noninvasive apparatus having a retaining member to retain a removable biosensor |
US6628809B1 (en) | 1999-10-08 | 2003-09-30 | Lumidigm, Inc. | Apparatus and method for identification of individuals by near-infrared spectrum |
US7890158B2 (en) | 2001-06-05 | 2011-02-15 | Lumidigm, Inc. | Apparatus and method of biometric determination using specialized optical spectroscopy systems |
GB2328279B (en) | 1997-08-12 | 2001-10-10 | Abbott Lab | Optical glucose detector |
US6265151B1 (en) | 1998-03-27 | 2001-07-24 | Seroptix, Inc. | Apparatus and method for infectious disease detection |
US6097975A (en) * | 1998-05-13 | 2000-08-01 | Biosensor, Inc. | Apparatus and method for noninvasive glucose measurement |
US6700613B1 (en) * | 1998-06-16 | 2004-03-02 | Eastman Kodak Company | Data-reading image capture apparatus, camera, and method of use |
US7098037B2 (en) | 1998-10-13 | 2006-08-29 | Inlight Solutions, Inc. | Accommodating subject and instrument variations in spectroscopic determinations |
US6441388B1 (en) | 1998-10-13 | 2002-08-27 | Rio Grande Medical Technologies, Inc. | Methods and apparatus for spectroscopic calibration model transfer |
US6157041A (en) * | 1998-10-13 | 2000-12-05 | Rio Grande Medical Technologies, Inc. | Methods and apparatus for tailoring spectroscopic calibration models |
US6519486B1 (en) | 1998-10-15 | 2003-02-11 | Ntc Technology Inc. | Method, apparatus and system for removing motion artifacts from measurements of bodily parameters |
US7991448B2 (en) * | 1998-10-15 | 2011-08-02 | Philips Electronics North America Corporation | Method, apparatus, and system for removing motion artifacts from measurements of bodily parameters |
US6329488B1 (en) | 1998-11-10 | 2001-12-11 | C. R. Bard, Inc. | Silane copolymer coatings |
US6596401B1 (en) | 1998-11-10 | 2003-07-22 | C. R. Bard Inc. | Silane copolymer compositions containing active agents |
US7436511B2 (en) * | 1999-01-22 | 2008-10-14 | Sensys Medical, Inc. | Analyte filter method and apparatus |
US6501982B1 (en) * | 1999-01-22 | 2002-12-31 | Sensys Medical, Inc. | System for the noninvasive estimation of relative age |
US6816605B2 (en) | 1999-10-08 | 2004-11-09 | Lumidigm, Inc. | Methods and systems for biometric identification of individuals using linear optical spectroscopy |
US6361960B1 (en) | 1999-11-09 | 2002-03-26 | Environmentally Sensitive Solutions, Inc. | Method and test kit for measuring concentration of a cleaning agent in a wash liquor |
US6687521B2 (en) | 2000-02-03 | 2004-02-03 | Hamamatsu Photonics K.K. | Noninvasion biological optical measuring instrument, measured portion holding device, and method for manufacturing the same |
EP1256315A4 (en) * | 2000-02-03 | 2004-07-28 | Hamamatsu Photonics Kk | Noninvasion biological optical measuring instrument, measured portion holding device, and method for manufacturing the same |
US6549861B1 (en) | 2000-08-10 | 2003-04-15 | Euro-Celtique, S.A. | Automated system and method for spectroscopic analysis |
WO2002016905A2 (en) | 2000-08-21 | 2002-02-28 | Euro-Celtique, S.A. | Near infrared blood glucose monitoring system |
WO2002024065A1 (en) | 2000-09-22 | 2002-03-28 | Knobbe, Lim & Buckingham | Method and apparatus for real-time estimation and control of pysiological parameters |
US6574490B2 (en) | 2001-04-11 | 2003-06-03 | Rio Grande Medical Technologies, Inc. | System for non-invasive measurement of glucose in humans |
US6983176B2 (en) | 2001-04-11 | 2006-01-03 | Rio Grande Medical Technologies, Inc. | Optically similar reference samples and related methods for multivariate calibration models used in optical spectroscopy |
US6862091B2 (en) | 2001-04-11 | 2005-03-01 | Inlight Solutions, Inc. | Illumination device and method for spectroscopic analysis |
US6865408B1 (en) | 2001-04-11 | 2005-03-08 | Inlight Solutions, Inc. | System for non-invasive measurement of glucose in humans |
US7126682B2 (en) | 2001-04-11 | 2006-10-24 | Rio Grande Medical Technologies, Inc. | Encoded variable filter spectrometer |
US7043288B2 (en) | 2002-04-04 | 2006-05-09 | Inlight Solutions, Inc. | Apparatus and method for spectroscopic analysis of tissue to detect diabetes in an individual |
JP4498636B2 (en) | 2001-04-27 | 2010-07-07 | 日本サーモスタット株式会社 | Thermostat device |
KR100389906B1 (en) | 2001-05-09 | 2003-07-04 | 삼성전자주식회사 | Apparatus and method for measuring concentration of component in target material |
US20020197600A1 (en) * | 2001-05-15 | 2002-12-26 | Maione Theodore E. | Method for determining the presence of infection in an individual |
US6544212B2 (en) | 2001-07-31 | 2003-04-08 | Roche Diagnostics Corporation | Diabetes management system |
US6668277B1 (en) * | 2001-09-14 | 2003-12-23 | The Regents Of The University Of California | Web-based multi-channel analyzer |
WO2003048396A1 (en) * | 2001-12-03 | 2003-06-12 | Seroptix, Inc. | A method for identifying markers |
AU2003200359A1 (en) * | 2002-02-11 | 2003-08-28 | Bayer Healthcare, Llc | Non-invasive System for the Determination of Analytes in Body Fluids |
US6654125B2 (en) | 2002-04-04 | 2003-11-25 | Inlight Solutions, Inc | Method and apparatus for optical spectroscopy incorporating a vertical cavity surface emitting laser (VCSEL) as an interferometer reference |
US7027848B2 (en) | 2002-04-04 | 2006-04-11 | Inlight Solutions, Inc. | Apparatus and method for non-invasive spectroscopic measurement of analytes in tissue using a matched reference analyte |
US20040010185A1 (en) * | 2002-07-11 | 2004-01-15 | Optical Sensors, Inc. | Method for measuring a physiologic parameter using a preferred site |
US7620212B1 (en) | 2002-08-13 | 2009-11-17 | Lumidigm, Inc. | Electro-optical sensor |
US7233817B2 (en) | 2002-11-01 | 2007-06-19 | Brian Yen | Apparatus and method for pattern delivery of radiation and biological characteristic analysis |
US20050033127A1 (en) * | 2003-01-30 | 2005-02-10 | Euro-Celtique, S.A. | Wireless blood glucose monitoring system |
KR20060002923A (en) | 2003-04-04 | 2006-01-09 | 루미다임 인크. | Multispectral biometric sensor |
US7394919B2 (en) | 2004-06-01 | 2008-07-01 | Lumidigm, Inc. | Multispectral biometric imaging |
US7668350B2 (en) | 2003-04-04 | 2010-02-23 | Lumidigm, Inc. | Comparative texture analysis of tissue for biometric spoof detection |
US7627151B2 (en) | 2003-04-04 | 2009-12-01 | Lumidigm, Inc. | Systems and methods for improved biometric feature definition |
US7751594B2 (en) | 2003-04-04 | 2010-07-06 | Lumidigm, Inc. | White-light spectral biometric sensors |
US7460696B2 (en) | 2004-06-01 | 2008-12-02 | Lumidigm, Inc. | Multispectral imaging biometrics |
US7545963B2 (en) | 2003-04-04 | 2009-06-09 | Lumidigm, Inc. | Texture-biometrics sensor |
US7539330B2 (en) | 2004-06-01 | 2009-05-26 | Lumidigm, Inc. | Multispectral liveness determination |
US7347365B2 (en) | 2003-04-04 | 2008-03-25 | Lumidigm, Inc. | Combined total-internal-reflectance and tissue imaging systems and methods |
US7271912B2 (en) * | 2003-04-15 | 2007-09-18 | Optiscan Biomedical Corporation | Method of determining analyte concentration in a sample using infrared transmission data |
WO2004097365A2 (en) * | 2003-04-30 | 2004-11-11 | Mcgill University | Method and system for measuring lactate levels in vivo |
US6975892B2 (en) * | 2003-10-21 | 2005-12-13 | Oculir, Inc. | Methods for non-invasive analyte measurement from the conjunctiva |
US6968222B2 (en) | 2003-05-02 | 2005-11-22 | Oculir, Inc. | Methods and device for non-invasive analyte measurement |
US6958039B2 (en) * | 2003-05-02 | 2005-10-25 | Oculir, Inc. | Method and instruments for non-invasive analyte measurement |
US7039448B2 (en) * | 2003-05-02 | 2006-05-02 | Diramed, Llc | Zero corrected optical blood analyte detector |
US20050007582A1 (en) * | 2003-07-07 | 2005-01-13 | Lumidigm, Inc. | Methods and apparatus for collection of optical reference measurements for monolithic sensors |
US20050073690A1 (en) * | 2003-10-03 | 2005-04-07 | Abbink Russell E. | Optical spectroscopy incorporating a vertical cavity surface emitting laser (VCSEL) |
US20060224057A1 (en) * | 2003-10-21 | 2006-10-05 | Oculir, Inc. | Methods for non-invasive analyte measurement |
US20050090723A1 (en) * | 2003-10-23 | 2005-04-28 | Nassar Saeed | Method and apparatus for non-invasive measuring of physiological glucose concentration in bodies of humans or animals |
EP1683058A2 (en) * | 2003-10-29 | 2006-07-26 | Novo Nordisk A/S | Medical advisory system |
US7263213B2 (en) | 2003-12-11 | 2007-08-28 | Lumidigm, Inc. | Methods and systems for estimation of personal characteristics from biometric measurements |
US20080009688A1 (en) * | 2004-04-14 | 2008-01-10 | Oculir, Inc. | Methods for non-invasive analyte measurement |
US20060258919A1 (en) * | 2004-04-14 | 2006-11-16 | Oculir, Inc. | Non-Invasive Analyte Measurement Device for Measuring Tears and Other Ocular Elements Using Electromagnetic Radiation and Method of Using the Same |
KR101330887B1 (en) * | 2004-05-14 | 2013-11-18 | 세모메테크 에이/에스 | A method and a system for the assessment of samples |
US20110163163A1 (en) * | 2004-06-01 | 2011-07-07 | Lumidigm, Inc. | Multispectral barcode imaging |
CA2998199A1 (en) | 2004-06-01 | 2005-12-15 | Kwalata Trading Limited | Methods for use with stem cells involving culturing on a surface with antibodies |
US8229185B2 (en) | 2004-06-01 | 2012-07-24 | Lumidigm, Inc. | Hygienic biometric sensors |
US7508965B2 (en) | 2004-06-01 | 2009-03-24 | Lumidigm, Inc. | System and method for robust fingerprint acquisition |
EP1786834B1 (en) | 2004-07-14 | 2016-02-17 | Glusense Ltd. | Implantable power sources and sensors |
WO2006020292A2 (en) * | 2004-07-20 | 2006-02-23 | Prescient Medical, Inc. | Systems and methods for medical interventional optical monitoring with molecular filters |
US8787630B2 (en) | 2004-08-11 | 2014-07-22 | Lumidigm, Inc. | Multispectral barcode imaging |
EP1815375A2 (en) * | 2004-11-15 | 2007-08-08 | Novo Nordisk A/S | Method and apparatus for monitoring long term and short term effects of a treatment |
DE102004059473A1 (en) * | 2004-12-10 | 2006-06-22 | Roche Diagnostics Gmbh | Method and device for the examination of medically relevant fluids and tissue samples |
US7460895B2 (en) * | 2005-01-24 | 2008-12-02 | University Of Iowa Research Foundation | Method for generating a net analyte signal calibration model and uses thereof |
US7801338B2 (en) | 2005-04-27 | 2010-09-21 | Lumidigm, Inc. | Multispectral biometric sensors |
US8251904B2 (en) | 2005-06-09 | 2012-08-28 | Roche Diagnostics Operations, Inc. | Device and method for insulin dosing |
US20060281982A1 (en) * | 2005-06-14 | 2006-12-14 | Diasense, Inc. | Method and apparatus for the non-invasive sensing of glucose in a human subject |
US8140139B2 (en) | 2005-06-14 | 2012-03-20 | Dominion Assets, Llc | Method and apparatus for the non-invasive sensing of glucose in a human subject |
WO2007057016A1 (en) * | 2005-11-16 | 2007-05-24 | Chemometec A/S | Determination of chemical or physical properties of sample or component of a sample |
US20070129690A1 (en) | 2005-12-02 | 2007-06-07 | Joel Rosenblatt | Catheter with polymeric coating |
ITTO20050887A1 (en) * | 2005-12-20 | 2007-06-21 | Consiglio Nazionale Ricerche | SPECTROPHOTOMETER AND SPECTROPHOTOMETRIC PROCEDURE USING A FABRY-PEROT CAVITY RESONATOR |
TW200734462A (en) * | 2006-03-08 | 2007-09-16 | In Motion Invest Ltd | Regulating stem cells |
US8175346B2 (en) | 2006-07-19 | 2012-05-08 | Lumidigm, Inc. | Whole-hand multispectral biometric imaging |
US7995808B2 (en) | 2006-07-19 | 2011-08-09 | Lumidigm, Inc. | Contactless multispectral biometric capture |
WO2008100329A2 (en) | 2006-07-19 | 2008-08-21 | Lumidigm, Inc. | Multibiometric multispectral imager |
US8355545B2 (en) | 2007-04-10 | 2013-01-15 | Lumidigm, Inc. | Biometric detection using spatial, temporal, and/or spectral techniques |
US7804984B2 (en) | 2006-07-31 | 2010-09-28 | Lumidigm, Inc. | Spatial-spectral fingerprint spoof detection |
US7801339B2 (en) | 2006-07-31 | 2010-09-21 | Lumidigm, Inc. | Biometrics with spatiospectral spoof detection |
US7446878B2 (en) * | 2006-11-16 | 2008-11-04 | Trutouch Technologies, Inc. | Method and apparatus for improvement of spectrometer stability, and multivariate calibration transfer |
JP5358890B2 (en) * | 2007-03-20 | 2013-12-04 | 株式会社島津製作所 | Interference spectrophotometer |
WO2008134135A2 (en) | 2007-03-21 | 2008-11-06 | Lumidigm, Inc. | Biometrics based on locally consistent features |
FR2923905B1 (en) * | 2007-11-19 | 2014-09-05 | Centre Nat Etd Spatiales | METHOD AND DEVICE FOR INTERFEROMETRIC INVERSION WITH FREE SAMPLING |
MY179050A (en) | 2008-04-04 | 2020-10-26 | Colgate Palmolive Co | Analysis of substrates having agents deposited thereon |
US20100082097A1 (en) * | 2008-10-01 | 2010-04-01 | Joel Rosenblatt | Article Containing Segregated Biguanide and Lewis Acid |
US20100160749A1 (en) * | 2008-12-24 | 2010-06-24 | Glusense Ltd. | Implantable optical glucose sensing |
US10475529B2 (en) | 2011-07-19 | 2019-11-12 | Optiscan Biomedical Corporation | Method and apparatus for analyte measurements using calibration sets |
WO2011028620A1 (en) | 2009-08-26 | 2011-03-10 | Lumidigm, Inc. | Multiplexed biometric imaging and dual-imager biometric sensor |
US20110082711A1 (en) | 2009-10-06 | 2011-04-07 | Masimo Laboratories, Inc. | Personal digital assistant or organizer for monitoring glucose levels |
US8570149B2 (en) | 2010-03-16 | 2013-10-29 | Lumidigm, Inc. | Biometric imaging using an optical adaptive interface |
US7884933B1 (en) | 2010-05-05 | 2011-02-08 | Revolutionary Business Concepts, Inc. | Apparatus and method for determining analyte concentrations |
US9037205B2 (en) | 2011-06-30 | 2015-05-19 | Glusense, Ltd | Implantable optical glucose sensing |
NL1039667C2 (en) * | 2012-06-11 | 2013-12-12 | Purac Biochem Bv | Quantification of lactide amounts in a polymeric matrix. |
US9351671B2 (en) | 2012-07-16 | 2016-05-31 | Timothy Ruchti | Multiplexed pathlength resolved noninvasive analyzer apparatus and method of use thereof |
US9351672B2 (en) | 2012-07-16 | 2016-05-31 | Timothy Ruchti | Multiplexed pathlength resolved noninvasive analyzer apparatus with stacked filters and method of use thereof |
US9585604B2 (en) | 2012-07-16 | 2017-03-07 | Zyomed Corp. | Multiplexed pathlength resolved noninvasive analyzer apparatus with dynamic optical paths and method of use thereof |
US20150018646A1 (en) | 2013-07-12 | 2015-01-15 | Sandeep Gulati | Dynamic sample mapping noninvasive analyzer apparatus and method of use thereof |
US9429474B2 (en) | 2012-10-08 | 2016-08-30 | Si-Ware Systems | Fourier transform micro spectrometer based on spatially-shifted interferogram bursts |
US9693694B2 (en) | 2012-11-21 | 2017-07-04 | Folim G. Halaka | Cancer cell detection using dielectrophoretic dynamic light scattering (DDLS) spectroscopy |
US8843186B2 (en) | 2012-11-21 | 2014-09-23 | Folim G. Halaka | Non-invasive reagentless glucose determination |
CN103808688B (en) * | 2014-01-22 | 2016-08-10 | 重庆医科大学 | The quality conformance of near infrared spectroscopy Fast nondestructive evaluation final drug |
CN104048939A (en) * | 2014-06-05 | 2014-09-17 | 中国肉类食品综合研究中心 | Near infrared rapid detection method for blood sugar content in live pig blood |
US9459201B2 (en) | 2014-09-29 | 2016-10-04 | Zyomed Corp. | Systems and methods for noninvasive blood glucose and other analyte detection and measurement using collision computing |
EP3206567A1 (en) | 2014-10-13 | 2017-08-23 | Glusense, Ltd. | Analyte-sensing device |
US9554738B1 (en) | 2016-03-30 | 2017-01-31 | Zyomed Corp. | Spectroscopic tomography systems and methods for noninvasive detection and measurement of analytes using collision computing |
US10871487B2 (en) | 2016-04-20 | 2020-12-22 | Glusense Ltd. | FRET-based glucose-detection molecules |
KR20180050948A (en) | 2016-11-07 | 2018-05-16 | 삼성전자주식회사 | Spectrometer and operation method of spectrometer |
US20180271417A1 (en) * | 2017-03-21 | 2018-09-27 | Wipro Limited | Method and a device for non-invasive monitoring of a blood glucose level of a user |
KR20200058845A (en) | 2018-11-20 | 2020-05-28 | 삼성전자주식회사 | Electronic device and method for obtaining information regarding blood glucose of user |
US11678819B2 (en) * | 2019-04-24 | 2023-06-20 | Roxanne Abul-Haj | Noninvasive applied force / optical glucose concentration determination analyzer apparatus and method of use thereof |
EP3858238A1 (en) * | 2020-01-31 | 2021-08-04 | trinamiX GmbH | Portable device and method for providing treatment data to a user |
Family Cites Families (11)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JPS5524004A (en) * | 1978-06-22 | 1980-02-20 | Minolta Camera Kk | Oxymeter |
US4882492A (en) * | 1988-01-19 | 1989-11-21 | Biotronics Associates, Inc. | Non-invasive near infrared measurement of blood analyte concentrations |
JPH0227226A (en) * | 1988-07-18 | 1990-01-30 | Fuji Electric Co Ltd | Data processing system for fourier transform spectroscope |
US5028787A (en) * | 1989-01-19 | 1991-07-02 | Futrex, Inc. | Non-invasive measurement of blood glucose |
US5086229A (en) * | 1989-01-19 | 1992-02-04 | Futrex, Inc. | Non-invasive measurement of blood glucose |
DE3922670C2 (en) * | 1989-04-10 | 1993-10-14 | Milchwirtschaftliche Foerderun | Quality assurance procedures for milk or milk products |
US5070874A (en) * | 1990-01-30 | 1991-12-10 | Biocontrol Technology, Inc. | Non-invasive determination of glucose concentration in body of patients |
US5197470A (en) * | 1990-07-16 | 1993-03-30 | Eastman Kodak Company | Near infrared diagnostic method and instrument |
AU2245092A (en) * | 1991-12-31 | 1993-07-28 | Vivascan Corporation | Blood constituent determination based on differential spectral analysis |
US5321265A (en) * | 1992-07-15 | 1994-06-14 | Block Myron J | Non-invasive testing |
US5459317A (en) * | 1994-02-14 | 1995-10-17 | Ohio University | Method and apparatus for non-invasive detection of physiological chemicals, particularly glucose |
-
1994
- 1994-02-14 US US08/195,709 patent/US5459317A/en not_active Expired - Lifetime
-
1995
- 1995-02-07 WO PCT/US1995/001556 patent/WO1995022046A1/en not_active Application Discontinuation
- 1995-02-07 AU AU18395/95A patent/AU703370B2/en not_active Ceased
- 1995-02-07 CA CA002183085A patent/CA2183085A1/en not_active Abandoned
- 1995-02-07 US US08/860,850 patent/US6061582A/en not_active Expired - Lifetime
- 1995-02-07 JP JP7521296A patent/JPH09509739A/en not_active Ceased
- 1995-02-07 EP EP95910197A patent/EP0745216A1/en not_active Ceased
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US7711411B2 (en) * | 2002-10-25 | 2010-05-04 | Nir Technologies Inc. | Method of in-vivo measurement of fat content of a body and apparatus therefor |
Also Published As
Publication number | Publication date |
---|---|
AU1839595A (en) | 1995-08-29 |
US5459317A (en) | 1995-10-17 |
JPH09509739A (en) | 1997-09-30 |
US6061582A (en) | 2000-05-09 |
EP0745216A1 (en) | 1996-12-04 |
AU703370B2 (en) | 1999-03-25 |
WO1995022046A1 (en) | 1995-08-17 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CA2183085A1 (en) | Method and apparatus for non-invasive detection of physiological chemicals, particularly glucose | |
EP0877926B1 (en) | Apparatus for multi-spectral analysis in noninvasive nir spectroscopy | |
EP0877925B1 (en) | Method and apparatus for multi-spectral analysis in noninvasive infrared spectroscopy | |
US5750994A (en) | Positive correlation filter systems and methods of use thereof | |
US5379764A (en) | Non-invasive determination of analyte concentration in body of mammals | |
WO1997028437A9 (en) | Method and apparatus for multi-spectral analysis in noninvasive infrared spectroscopy | |
CA2019511C (en) | Method of and apparatus for determining the similarity of a biological analyte from known biological fluids | |
AU6425899A (en) | Methods and apparatus for tailoring spectroscopic calibration models | |
Ham et al. | Determination of glucose concentrations in an aqueous matrix from NIR spectra using optimal time-domain filtering and partial least-squares regression | |
US6615151B1 (en) | Method for creating spectral instrument variation tolerance in calibration algorithms | |
JPH09159606A (en) | Liquid-correlation spectrometry | |
AU713502C (en) | Method and apparatus for multi-spectral analysis in noninvasive NIR spectroscopy | |
CA2169881C (en) | Liquid correlation spectrometry | |
Cohen et al. | Paper Session II-B-Optical Diagnostics: Reagentless Chemistry for Extended Space Flights |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
EEER | Examination request | ||
FZDE | Discontinued |