Non-invasive in vivo tissue classification using near-infrared measurements

• US 6,405,065 B1
• Filed: 01/19/2000
• Issued: 06/11/2002
• Est. Priority Date: 01/22/1999
• Status: Expired due to Fees

• Current Assignee: GLT Acquisition Corporation
• Original Assignee: Instrumentation Metrics Inc.
• Inventors: Ruchti, Timothy L., Malin, Stephen F., Rennert, Jessica
• Primary Examiner(s): WINAKUR, ERIC FRANK
• Application Number:
  US09/487,547
• Time in Patent Office: 874 Days
• Field of Search: 600/310, 600/314, 600/473, 600/475, 128/920, 128/923, 128/924, 128/925, 356/402, 250/340, 250/341.1, 250/339.01
• US Class Current: 600/310
• CPC Class Codes: A61B 2560/0223 of calibration, e.g. protocols for calibrating sensors
• CPC Class Codes: A61B 5/0075 by spectroscopy, i.e. measuring spectra, e.g. Raman spectroscopy, infrared absorption spectroscopy A61B5/0071 takes precedence
• CPC Class Codes: A61B 5/1075 for measuring dimensions by non-invasive methods, e.g. for determining thickness of tissue layer A61B8/0858 takes precedence
• CPC Class Codes: A61B 5/14532 for measuring glucose, e.g. by tissue impedance measurement
• CPC Class Codes: A61B 5/14546 for measuring analytes not otherwise provided for, e.g. ions, cytochromes
• CPC Class Codes: A61B 5/1455 using optical sensors, e.g. spectral photometrical oximeters
• CPC Class Codes: A61B 5/7264 Classification of physiological signals or data, e.g. using neural networks, statistical classifiers, expert systems or fuzzy systems neural networks per se G06N3/00; expert systems per se G06N5/00
• CPC Class Codes: A61B 5/7267 involving training the classification device
• CPC Class Codes: G01N 21/359 using near infrared light
• CPC Class Codes: G01N 21/4785 Standardising light scatter apparatus; Standards therefor
• CPC Class Codes: G16H 50/20 for computer-aided diagnosis, e.g. based on medical expert systems
• CPC Class Codes: Y10S 128/923 by comparison of patient data to other data

Abstract

An in vivo, non-invasive method of tissue classification using near-IR (NIR) spectral measurements. A classification model is based on NIR spectral absorbance measurements from an exemplary population. Spectral features representing variation between tissue types are identified. Analytic techniques enhance the features of interest and correct spectral interference to improve the predictive ability of the classification model. A classification routine defines classes based on variation between tissue types, such that variation within a class is small compared to variation between classes. A decision rule assigns individual samples from the exemplary population to classes. An in-vivo, non-invasive procedure applies the classification model to individual tissue samples. A preferred embodiment of the invention distinguishes transgenic mice from non-transgenic individuals based on variation in fat composition within muscle tissue.

Claims

1. A method of developing a classification model for classification of tissue samples comprising the steps of:

providing a set of measured absorbance spectra from a population of exemplary subjects;

visually inspecting said spectra;

comparing said spectra with known spectral absorbance patterns;

selecting at least one feature of interest wherein variation according to issue type may be found based on observed similarities between said sample spectra and said known spectral absorbance patterns

enhancing said one or more features of interest by correcting said spectra for light scatter;

extracting at least one feature of interest relevant to classification;

selecting factors of said features of interest related to structural and chemical properties of said samples;

defining classes for said tissue samples on the basis of structural and state similarity, wherein variation within a class is small compared to variation between classes; and

assigning class membership.

2. The method of claim 1, wherein each of said measured absorbance spectra are represented as a vector mε

^N of absorbance values pertaining to a set of N wavelengths λ

ε

^N spanning the wavelength region of 700 to 2500 nm.

3. The method of claim 1, wherein said step of selecting at least one feature of interest further comprises:

truncating said spectra at the boundaries of said at least one selected feature of interest.

4. The method of claim 1, wherein said step of correcting said spectra comprises:

employing multiplicative scatter correction.

5. The method of claim 4, wherein scatter for each of said sample spectra is estimating by rotating said sample spectra to a reference spectrum {overscore (m)} according to:

6. The method of claim 1, wherein said feature extraction step further comprises the steps of:

applying a mathematical transformation wherein said feature-enhanced sample spectra are decomposed to distinct factors that represent underlying variation within the data set;

employing factor-based methods to determine which of said factors are attributable to a known spectral absorbance pattern; and

including the measured contribution of said known spectral absorbance pattern to the sample spectral absorbance as features.

7. The method of claim 6, wherein said extracted features are represented from said scatter-corrected measurements, x, in a vector zε

^M through;

8. The method of claim 7 further comprising the steps of:

including factors represented as vectors in the data set; and

excluding those factors represented as scalars from the data set.

9. The method of claim 1, wherein said factor selection step further comprises the steps of:

representing variation within said measured spectra as factor loadings; and

representing the weight of a particular sample on said spectral variation as factor scores corresponding to said factor loadings.

10. The method of claim 9, wherein a clustering of said factor scores represents a variation according to tissue type and said factor loadings represent a feature responsible for tissue-specific variation.

11. The method of claim 1, wherein said set of defining classes comprises:

employing means for defining classes, wherein within-class variation is minimized and between-class variation is maximized.

12. The method of claim 11 wherein said means for defining classes comprises linear discriminant analysis.

13. The method of claim 12, wherein said discriminant analysis employs a criterion function:

$J<br><br>\left(w\right)=\frac{w^t<br><br>S_b<br><br>w}{w^t<br><br>S_w<br><br>w},$

where w is a directional unit vector, S_b is the between class scatter matrix, S_w is the within-class scatter matrix, wherein a vector w is selected such that the between-class variation/within-class variation ratio is maximized, wherein said vector w represents the separation between classes.

14. The method of claim 13, further comprising the step of:

applying said criterion function on the basis of the first M scores, where M is an arbitrary number, wherein each of said samples are projected onto said vector w;

wherein said projections onto said vector w are scalars.

15. The method of claim 14, wherein said decision rule represents said scalars of said samples as L, and {overscore (L)} represents a boundary between said classes;

and

wherein assignment to a first class is based on the condition L>

{overscore (L)}; and

wherein assignment to a second class is based on the condition L<

{overscore (L)}.

16. The method for classifying tissue comprising the steps of:

providing a new set of spectral absorbance measurements from a test subject;

extracting features corresponding to tissue-specific variation;

employing a classification model to compare said extracted features with a set of exemplary measurements; and

assigning class membership through application of a decision rule, wherein said features are represented in a vector zε

^M that is determined through;

17. The method of claims 16, wherein said feature extraction step comprises any mathematical transformation that enhances a quality or aspect of sample measurement for interpretation to represent concisely structural properties and physiological state of a tissue measurement site, wherein a resulting set of features is used to classify a subject.

18. The method of claim 16, wherein said classification model comprises means for determining a set of similarity measures with predefined classes.

19. The method of claim 18, further comprising the step of:

providing a classification system that assumes that said classes are mutually exclusive and that forces each measurement to be assigned to a single class.

20. The method of claim 16 wherein said decision rule comprises means for assigning class membership on the basis of a set of measures calculated by a decision engine.

21. The method of claim 16, wherein said pattern classification step further comprises the steps of:

measuring the similarity of a feature to predefined classes; and

assigning class membership.

22. The method of claim 21, wherein said measuring step uses mutually exclusive classes and assigns each measurement to one class.

23. The method of claim 22, further comprising the step of:

using measurements and class assignments to determine a mapping from features to class assignments.

24. A method of developing a classification model for distinguishing transgenic mice from non-transgenic mice based on fat composition in the tissue comprising the steps of:

providing a set of spectral absorbance measurements from an exemplary population of subject animals;

selecting one or more features of interest within said spectral measurements wherein variation according to tissue type may be found;

enhancing said one or more features of interest;

extracting at least one feature of interest relevant to classification;

selecting factors of said features of interest related to structural and chemical properties of said samples;

defining classes for said tissue samples on the basis of structural and state similarity, wherein variation within a class is small compared to variation between classes; and

assigning class membership.

25. The method of claim 24, wherein said transgenic mice are foreign growth hormone transgenic mice.

26. The method of claim 24, wherein each of said measured absorbance spectra are represented as a vector mε

^N of absorbance values pertaining to a set of N wavelengths λ

ε

^N spanning the wavelength region of 700 to 2500 nm.

27. The method of claim 24, wherein said feature extraction step further comprises the steps of:

visually inspecting said absorbance spectra;

comparing said absorbance spectra with the spectral absorbance pattern of animal fat;

selecting said one or more features of interest based on observed similarities between said sample spectra and said spectral absorbance pattern of animal fat; and

truncating said sample spectra at the boundaries of said one or more selected features of interest.

28. The method of claim 27, wherein said feature extraction step further comprises the steps of:

applying a mathematical transformation wherein said feature-enhanced sample spectra are decomposed to distinct factors that represent underlying variation within the data set;

employing factor-based methods to determine which of said factors are attributable to said animal fat spectral absorbance pattern; and

including the measured contribution of said animal fat spectral absorbance pattern to the sample spectral absorbance as features.

29. The method of claim 28, wherein said extracted features are represented from said scatter-corrected measurements, x, in a vector zε

^M through;

30. The method of claim 28, wherein loading of said factors represent variation according to fat composition of tissue, and scores of said factors identify different subject classes based on fat composition in the tissue.

31. The method of claim 24, wherein said feature enhancement step comprises:

employing means for correcting said sample spectra for light scatter.

32. The method of claim 31 wherein said means for correcting said spectra comprises multiplicative scatter correction.

33. The method of claim 32, wherein scatter for each of said sample spectra is estimating by rotating said sample spectra to a reference spectrum {overscore (m)} according to:

34. The method of claim 24, wherein said class definition step defines classes for said population based on variation according to fat composition in the tissue, such that within-class variation is minimized and between-class variation is maximized.

35. The method of claim 24, wherein said class assignment step employs a decision rule to assign membership to individuals from said sample population.

36. A method of distinguishing transgenic mice from non-transgenic mice based on variation according to fat composition in the tissue comprising the steps of:

providing a new set of spectral absorbance measurements from a subject;

extracting features corresponding to tissue-specific variation;

performing pattern classification, wherein a classification model compares said extracted features with a set of exemplary measurements; and

assigning class membership through application of a decision rule.

37. The method of claim 36, wherein said transgenic mice are foreign growth hormone transgenic mice.

38. The method of claim 36, wherein said feature extraction step comprises any mathematical transformation that enhances a quality or aspect of sample measurement for interpretation to represent concisely structural properties and physiological state of a tissue measurement site, wherein a resulting set of features is used to classify said subject according to fat composition in the tissue.

39. The method of claim 36, wherein said classification model comprises means for determining a set of similarity measures with predefined classes.

40. The method of claim 39, wherein said classes are defined on the basis of fat composition in the tissue.

41. The method of claim 39, further comprising the step of:

providing a classification system that assumes that said classes are mutually exclusive and that forces each measurement to be assigned to a single class.

42. The method of claim 41 wherein said features are represented in a vector zε

^M that is determined through;

43. The method of claim 36, wherein said decision rule comprises means for assigning class membership on the basis of a set of measures calculated by a decision engine.

44. The method of claim 36, wherein said pattern classification step further comprises the steps of:

measuring the similarity of a feature to predefined classes; and

assigning class membership.

45. The method of claim 44, wherein said predefined classes are any of transgenic mice and non-transgenic mice.

46. The method of claim 45, wherein said classes are mutually exclusive.

47. The method of claim 46, further comprising the step of:

using measurements and class assignments to determine a mapping from features to class assignments.