Methods and apparatus for processing spectral data for use in tissue characterization

• US 7,282,723 B2
• Filed: 04/18/2003
• Issued: 10/16/2007
• Est. Priority Date: 07/09/2002
• Status: Expired due to Fees

• Current Assignee: Luma Imaging Corp. (SpectraScience, Inc.)
• Original Assignee: MediSpectra, Inc. (SpectraScience, Inc.)
• Inventors: Jiang, Chunsheng, Meese, Thomas M., Schomacker, Kevin T., Zelenchuk, Alex, Griffin, Christopher E., Flewelling, Ross F.
• Primary Examiner(s): Porta; David
• Assistant Examiner(s): Sung; Christine
• Application Number:
  US10/418,415
• Publication Number:
  US 20040010375A1
• Time in Patent Office: 1,642 Days
• Field of Search: 250/458.1, 250/459.1, 250/461.1, 600/476, 600/477, 600/478, 600/317
• US Class Current: 250/458.1
• CPC Class Codes: A61B 1/0655 Control therefor
• CPC Class Codes: A61B 5/0059 using light, e.g. diagnosis by transillumination, diascopy, fluorescence A61B5/0093, A61B5/1455, A61B5/02416 and A61B5/0261 take precedence
• CPC Class Codes: A61B 5/725 using specific filters therefor, e.g. Kalman or adaptive filters A61B5/316 take precedence
• CPC Class Codes: A61B 5/7257 using Fourier transforms
• 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 2021/6423 Spectral mapping, video display
• CPC Class Codes: G01N 21/31 Investigating relative effect of material at wavelengths characteristic of specific elements or molecules, e.g. atomic absorption spectrometry G01N21/72 takes precedence
• CPC Class Codes: G01N 21/6456 Spatial resolved fluorescence measurements; Imaging
• CPC Class Codes: G06V 20/698 Matching; Classification
• CPC Class Codes: G16H 50/20 for computer-aided diagnosis, e.g. based on medical expert systems

Abstract

The invention provides methods for processing tissue-derived spectral data for use in a tissue classification algorithm. Methods of the invention comprise application of spectral and/or image masks for automatically separating ambiguous or unclassifiable spectral data from valid spectral data. The invention improves the accuracy of tissue classification, in part, by properly identifying and accounting for spectral data from tissue regions that are affected by an obstruction and/or regions that lie outside a diagnostic zone of interest.

Claims

1. A method of processing spectral data obtained from a plurality of regions of a tissue sample for use in a tissue classification scheme, the method comprising the steps of:

(a) automatically identifying a subset of spectral data obtained from said plurality of regions of said tissue sample, said subset of data affected by one or more obstructions selected from the group consisting of glare, shadow, mucus, a smoke tube, and a speculum, and identifying one or more regions of said tissue sample from which said subset was obtained;

(b) processing spectral data obtained from said plurality of regions in said tissue classification scheme, wherein said subset of spectral data identified in step (a) is accounted for in said tissue classification scheme by at least one of (i), (ii), (iii) as follows;

(i) disqualifying said subset of spectral data from use in said tissue classification scheme;

(ii) classifying said one or more identified regions from which said subset of spectral data was obtained as indeterminate; and

(iii) weighting at least a portion of said subset of spectral data in said tissue classification scheme to penalize said data in favor of other spectral data determined to be unaffected by said one or more obstructions; and

(c) classifying said plurality of regions of said tissue sample according to said tissue classification scheme.

2. The method of claim 1, wherein step (a) comprises:

(i) applying one or more spectral masks; and

(ii) optionally, applying one or more image masks.

3. The method of claim 2, wherein said one or more spectral masks comprises a cervical edge mask.

4. The method of claim 3, wherein said cervical edge mask is based at least in part on a ratio of a first reflectance intensity at a first wavelength to a second reflectance intensity at a second wavelength.

5. The method of claim 4, wherein said first wavelength is about 700 nm and said second wavelength is about 540 nm.

6. The method of claim 3, wherein said cervical edge mask is based at least in part on a ratio of a first fluorescence intensity at a first wavelength to a second fluorescence intensity at a second wavelength.

7. The method of claim 6, wherein said first wavelength is about 530 nm and said second wavelength is about 410 nm.

8. The method of claim 3, wherein said applying step comprises comparing F(530 nm)/F(410 nm) with a threshold of about 4.75.

9. The method of claim 3, wherein said cervical edge mask is based at least in part on (i) a ratio of a first reflectance intensity to a second reflectance intensity and (ii) a ratio of a first fluorescence intensity to a second fluorescence intensity.

10. The method of claim 3, wherein said cervical edge mask comprises a metric as follows:

BB(450 nm)BB(700 nm)/BB(540 nm)<

0.30 OR F(530 nm)/F(410 nm)>

4.75.

11. The method of claim 2, wherein said one or more spectral masks comprises a mucus mask.

12. The method of claim 11, wherein said mucus mask is based at least in part on a ratio of a first reflectance intensityat a first wavelength to a second reflectance intensity at a second wavelength.

13. The method of claim 12, wherein said first wavelength and said second wavelength maximize a discrimination function comprising the ratio.

14. The method of claim 12, wherein said first wavelength is about 456 nm and said second wavelength is about 542 nm.

15. The method of claim 11, wherein said applying step comprises comparing BB(456 nm)/BB(542 nm) with a threshold of about 1.1.

16. The method of claim 11, wherein said applying step comprises comparing BB(594 nm)/BB(610 nm) with a threshold of about 0.74.

17. The method of claim 11, wherein said mucus mask comprises a metric as follows:

BB(456 nm)/BB(542 nm)<

1.06 OR BB(594 nm)/BB(610 nm)>

0.74.

18. The method of claim 1, wherein said processing step comprises disqualifying said subset of spectral data from use in said tissue classification scheme.

19. The method of claim 1, wherein said subset of spectral data identified in step (a) is accounted for in said tissue classification scheme by classifying said one or more identified regions from which said subset was obtained as indeterminate.

20. The method of claim 1, wherein said processing step comprises determining an overlap between said one or more identified regions from which said subset was obtained and an area identified by one or more image masks.

21. The method of claim 20, wherein said processing step comprises classifying at least a portion of said overlap as indeterminate.

22. The method of claim 20, wherein said processing step comprises weighting spectral data from said overlap in said tissue classification scheme according to a likelihood said data are affected by an artifact.

23. The method of claim 20, wherein said one or more image masks comprise a member selected from the group consisting of a smoke tube mask, a speculum mask, a region-of-interest mask, and a vaginal wall mask.

24. The method of claim 1, wherein said spectral data comprises fluorescence spectral data and reflectance spectral data.

25. The method of claim 1, wherein step (a) comprises:

(i) applying one or more spectral masks; and

(ii) applying one or more image masks.

26. The method of claim 25, wherein step (ii) comprises automatically identifying pixels of an image of said tissue sample corresponding to said one or more obstructions.

27. The method of claim 25, wherein step (i) comprises automatically identifying one or more interrogations points at which fluorescence data, reflectance data, or both fluorescence data and reflectance data are affected by said one or more obstructions.

28. The method of claim 25, wherein said one or more spectral masks comprises a cervical edge mask, a mucus mask, an NED mask, and a necrosis mask.

29. The method of claim 25, wherein said one or more image masks comprises a glare mask, a mucus mask, a blood mask, an os mask, a region of interest mask, a smoke tube mask, a speculum mask, a vaginal wall mask, and a fluid mask.

30. The method of claim 1, wherein said subset of spectral data identified in step (a) is accounted for in said tissue classification scheme by weighting said identified subset of spectral data in said tissue classification scheme according to a likelihood said data are affected by an artifact.

31. The method of claim 1, wherein step (a) further comprises automatically identifying spectral data from a region outside a diagnostic zone of interest, said region comprising at least one member selected from the group consisting of a vaginal wall, an os, and a cervical edge.

32. The method of claim 1, wherein step (a) further comprises automatically identifying spectral data affected by at least one member selected from the group consisting of blood, fluid, and foam.

33. The method of claim 1, further comprising the step of applying a spectral mask to identify NED tissue prior to step (b).

34. The method of claim 1, further comprising the step of applying a spectral mask to identify necrotic tissue prior to step (b).

35. The method of claim 1, further comprising the step of applying a low signal mask to account for low signal data prior to step (a).

36. The method of claim 1, wherein said tissue sample comprises in vivo tissue.

37. The method of claim 1, wherein said tissue sample comprises epithelial cells.

38. The method of claim 37, wherein said tissue sample comprises at least one tissue type selected from the group consisting of colorectal, gastroesophageal, urinary bladder, lung, and skin tissue.

39. The method of claim 1, wherein said tissue sample comprises cervical tissue.

40. The method of claim 1, wherein said tissue classification scheme identifies CIN 2/3 tissue.

41. The method of claim 1, wherein said tissue classification scheme identifies a probability of CIN 2/3 associated with a portion of said tissue sample.

42. The method of claim 1, further comprising the step of:

(d) displaying results of said tissue classification scheme.

53. The method of claim 42, wherein said one or more image masks comprise a glare mask, a mucus mask, a blood mask, an os mask, a region of interest mask, a smoke tube mask, a speculum mask, a vaginal wall mask, and a fluid mask.

54. The method of claim 42, wherein said spectral data from said one or more regions identified in step (a), in step (b), or in both steps (a) and (b) are accounted for in said tissue classification scheme by classifying said one or more identified regions as indeterminate.

55. The method of claim 42, wherein said spectral data from said one or more regions identified in step (a), in step (b), or in both steps (a) and (b) are accounted for in said tissue classification scheme by weighting said spectral data from said one or more identified regions according to a likelihood said data are affected by an artifact.

56. The method of claim 42, further comprising at least one of the steps of:

(i) applying a spectral mask to identify NED tissue prior to step (c);

(ii) applying a spectral mask to identify necrotic tissue prior to step (c); and

(iii) applying a low signal mask to account for low signal data prior to step (a).

43. A method of processing spectral data obtained from a plurality of regions of a tissue sample for use in a tissue classification scheme, the method comprising the steps of:

(a) applying one or more spectral masks to automatically identify one or more regions of said tissue sample from which spectral data is non-representative of a zone of interest of said tissue sample;

(b) applying one or more image masks to automatically identify one or more regions of said tissue sample from which spectral data is non-representative of a zone of interest of said tissue sample;

(c) processing spectral data obtained from said plurality of regions in said tissue classification scheme, wherein spectral data from said one or more regions identified in steps (a) and (b) are accounted for in said tissue classification scheme by at least one of (i) (ii) (iii), as follows;

(i) disqualifying spectral data from one or more of said identified regions from use in said tissue classification scheme;

(ii) classifying one or more of said identified regions as indeterminate; and

(iii) weighting spectral data from one or more of said identified regions in said tissue classification scheme to penalize said data in favor of other spectral data that is determined to be representative of said zone of interest of said tissue sample; and

(d) classifying said plurality of regions of said tissue sample according to said tissue classification scheme.

44. The method of claim 43, wherein step (a) comprises automatically identifying one or more interrogation points at which fluorescence data, reflectance data, or both fluorescence data and reflectance data are non-representative of said zone of interest of said tissue sample.

45. The method of claim 43, wherein step (b) comprises automatically identifying pixels of an image of said tissue sample corresponding to an area outside said zone of interest of said tissue sample.

46. The method of claim 43, wherein said spectral data that are non-representative of a zone of interest of said tissue sample comprise one or both of the following:

(i) spectral data from tissue regions affected by one or more obstructions; and

(ii) spectral data from regions lying outside a diagnostic zone of interest.

47. The method of claim 46, wherein said one or more obstructions comprises at least one of the following:

mucus, fluid, foam, a medical instrument, glare, shadow, and blood.

48. The method of claim 46, wherein said regions lying outside a diagnostic zone of interest comprises at least one of the following:

a tissue wall, an os, an edge surface of a tissue, and tissue in the vicinity of a smoke tube.

49. The method of claim 43, wherein said spectral data that are non-representative of a zone of interest of said tissue sample comprise:

(i) spectral data from tissue regions affected by one or more obstructions; and

(ii) spectral data from regions lying outside a diagnostic zone of interest.

50. The method of claim 43, wherein said one or more spectral masks comprises at least one of the following:

a cervical edge mask, and a mucus mask.

51. The method of claim 43, wherein said one or more spectral masks comprise a cervical edge mask and a mucus mask.

52. The method of claim 43, wherein said one or more image masks comprises at least one of the following:

a glare mask, a mucus mask, a blood mask, an os mask, a region of interest mask, a smoke tube mask, a speculum mask, a vaginal wall mask, and a fluid mask.

57. The method of claim 43, wherein said tissue sample comprises in vivo tissue.

58. The method of claim 43, wherein said tissue sample comprises at least one tissue type selected from the group consisting of colorectal, gastroesophageal, urinary bladder, lung, and skin tissue.

59. The method of claim 43, wherein said tissue sample comprises cervical tissue.

60. The method of claim 43, wherein said tissue classification scheme identifies CIN 2/3 tissue.

61. The method of claim 43, wherein said tissue classification scheme identifies a probability of CIN 2/3 associated with a portion of said tissue sample.

62. The method of claim 43, further comprising the step of:

(e) displaying results of said tissue classification scheme.

64. The apparatus of claim 62, wherein said processor is configured to automatically identify said subset of spectral data by:

(i) applying one or more spectral masks; and

(ii) optionally, applying one or more image masks.

65. The apparatus of claim 64, wherein said one or more spectral masks comprises a cervical edge mask.

66. The apparatus of claim 64, wherein said one or more spectral masks comprises a mucus mask.

67. The apparatus of claim 62, wherein said processor is configured to determine an overlap between said one or more identified regions from which said subset was obtained and an area identified by one or more image masks.

68. The apparatus of claim 67, wherein said processor is configured to classify at least a portion of said overlap as indeterminate.

69. The apparatus of claim 67, wherein said processor is configured to weight spectral data from said overlap in said tissue classification scheme according to a likelihood said data are affected by an artifact.

70. The apparatus of claim 67, wherein said one or more image masks comprise a member selected from the group consisting of a smoke tube mask, a speculum mask, a region-of-interest mask, and a vaginal wall mask.

71. The apparatus of claim 62, wherein said spectral data comprise fluorescence spectral data and reflectance spectral data.

72. The apparatus of claim 62, wherein said processor is configured to automatically identify said subset of spectral data by:

(i) applying one or more spectral masks; and

(ii) applying one or more image masks.

73. The apparatus of claim 72, wherein said applying one or more image masks comprises automatically identifying pixels of an image of said tissue sample corresponding to said one or more obstructions.

74. The apparatus of claim 72, wherein said applying one or more spectral masks comprises automatically identifying one or more interrogation points at which fluorescence data, reflectance data, or both fluorescence data and reflectance data are affected by said one or more obstructions.

75. The apparatus of claim 72, wherein said one or more spectral masks comprises a cervical edge mask, a mucus mask, an NED mask, and a necrosis mask.

76. The apparatus of claim 72, wherein said one or more image masks comprises a glare mask, a mucus mask, a blood mask, an os mask, a region of interest mask, a smoke tube mask, a speculum mask, a vaginal wall mask, and a fluid mask.

77. The apparatus of claim 62, wherein said tissue sample comprises in vivo tissue.

78. The apparatus of claim 62, wherein said tissue sample comprises epithelial cells.

79. The apparatus of claim 78, wherein said tissue sample comprises at least one tissue type selected from the group consisting of colorectal, gastroesophageal, urinary bladder, lung, and skin tissue.

80. The apparatus of claim 62, wherein said tissue sample comprises cervical tissue.

81. The apparatus of claim 62, wherein said tissue classification scheme identifies CIN 2/3 tissue.

82. The apparatus of claim 62, wherein said tissue classification scheme identifies a probability of CIN 2/3 associated with a portion of said tissue sample.

63. An apparatus for processing spectral data obtained from a plurality of regions of a tissue sample for use in a tissue classification system, the apparatus comprising:

a probe configured to obtain spectral data from said plurality of regions of said tissue sample; and

a processor configured to;

automatically identify a subset of spectral data obtained from said plurality of regions of said tissue sample, said subset of data affected by one or more obstructions selected from the group consisting of glare, shadow, mucus, a smoke tube, and a speculum, and identifying one or more regions of said tissue sample from which said subset was obtained; and

process spectral data obtained from said plurality of regions in said tissue classification scheme, wherein said identified subset of spectral data is accounted for in said tissue classification scheme by at least one of (i), (ii), (iii) as follows;

(i) disqualifying said subset of spectral data from use in said tissue classification scheme;

(ii) classifying said one or more identified regions from which said subset of spectral data was obtained as indeterminate; and

(iii) weighting at least a portion of said subset of spectral data in said tissue classification scheme to penalize said data in favor of other spectral data determined to be unaffected by said one or more obstructions.

83. An apparatus for processing spectral data obtained from a plurality of regions of a tissue sample for use in a tissue classification system, the apparatus comprising:

a probe configured to obtain spectral data from said plurality of regions of said tissue sample; and

a processor configured to;

(i) apply one or more spectral masks to automatically identify one or more regions of said tissue sample from which spectral data is non-representative of a zone of interest of said tissue sample;

(ii) apply one or more image masks to automatically identify one or more regions of said tissue sample from which spectral data is non-representative of a zone of interest of said tissue sample; and

(iii) process spectral data obtained from said plurality of regions in said tissue classification scheme, wherein spectral data from said one or more regions identified in (i) and (ii) are accounted for in said tissue classification scheme by at least one of (A), (B), (C) as follows;

(A) disqualifying spectral data from one or more of said identified regions from use in said tissue classification scheme;

(B) classifying one or more of said identified regions as indeterminate; and

(C) weighting spectral data from one or more of said identified regions in said tissue classification scheme to penalize said data in favor of other spectral data that is determined to be representative of said zone of interest of said tissue sample.

84. The apparatus of claim 83, wherein said processor is configured to apply said one or more spectral masks to automatically identify one or more interrogation points at which fluorescence data, reflectance data, or both fluorescence data and reflectance data are non-representative of said zone of interest of said tissue sample.

85. The apparatus of claim 83, wherein said processor is configured to apply said one or more image masks to automatically identify pixels of an image of said tissue sample corresponding to an area outside said zone of interest of said tissue sample.

86. The apparatus of claim 83, wherein said spectral data that are non-representative of a zone of interest of said tissue sample comprise one or both of the following:

(A) spectral data from tissue regions affected by one or more obstructions; and

(B) spectral data from regions lying outside a diagnostic zone of interest.

87. The apparatus of claim 86, wherein said one or more obstructions comprises at least one of the following:

mucus, fluid, foam, a medical instrument, glare, shadow, and blood.

88. The apparatus of claim 86, wherein said regions lying outside a diagnostic zone of interest comprises at least one of the following:

a tissue wall, an os, an edge surface of a tissue, and tissue in the vicinity of a smoke tube.

89. The apparatus of claim 83, wherein said spectral data that are non-representative of a zone of interest of said tissue sample comprise:

(A) spectral data from tissue regions affected by one or more obstructions; and

(B) spectral data from regions lying outside a diagnostic zone of interest.

90. The apparatus of claim 83, wherein the processor is configured to perform at least one of the following:

(A) apply a spectral mask to identify NED tissue;

(B) apply a spectral mask to identify necrotic tissue; and

(C) apply a low signal mask to account for low signal data.

91. The apparatus of claim 83, wherein said tissue sample comprises in vivo tissue.

92. The apparatus of claim 83, wherein said tissue sample comprises at least one tissue type selected from the group consisting of colorectal, gastroesophageal, urinary bladder, lung, and skin tissue.

93. The apparatus of claim 83, wherein said tissue sample comprises cervical tissue.

94. The apparatus of claim 83, wherein said tissue classification scheme identifies CIN 2/3 tissue.

95. The apparatus of claim 83, wherein said tissue classification scheme identifies a probability of CIN 2/3 associated with a portion of said tissue sample.