Multimodal detection of tissue abnormalities based on raman and background fluorescence spectroscopy

Abstract

Methods and apparatus for classifying tissue use features of Raman spectra and background fluorescent spectra. The spectra may be acquired in the near-infrared wavelengths. Principal component analysis and linear discriminant analysis of reference spectra may be used to obtain a classification function that accepts features of the Raman and background fluorescence spectra for test tissue and yields an indication as to the likelihood that the test tissue is abnormal. The methods and apparatus may be applied to screening for skin cancers or other diseases.

What is claimed is: 
     
         1 . A method for characterizing a tissue, the method comprising:
 obtaining a raw spectrum of the tissue by illuminating the tissue with infrared light that is substantially monochromatic and detecting light backscattered from the tissue at a plurality of infrared wavelengths in a first wavelength range, processing the raw spectrum to obtain both a Raman spectrum of the tissue and a background fluorescence spectrum of the tissue that is a background to the Raman spectrum in the first wavelength range;   processing the Raman spectrum to extract one or more features of the Raman spectrum and processing the background fluorescence spectrum to extract one or more features of the background fluorescence spectrum; and,   characterizing the tissue based upon at least the one or more features of the Raman spectrum and the one or more features of the background fluorescence spectrum.   
     
     
         2 . A method according to  claim 1  wherein processing the raw spectrum to obtain the Raman spectrum comprises fitting a background fitting function to the raw spectrum to yield a fitted background function and subtracting the fitted background function from the raw spectrum. 
     
     
         3 . A method according to  claim 2  wherein processing the background fluorescence to extract one or more features of the background fluorescence spectrum comprises processing the fitted background function. 
     
     
         4 . A method according to  claim 3  wherein the first wavelength range includes wavelengths from about 800 nm to about 1000 nm. 
     
     
         5 . A method according to  claim 3  wherein the infrared light has a wavelength of 785 nm. 
     
     
         6 . A method according to  claim 1  wherein characterizing the tissue based upon at least the Raman spectrum features and the background fluorescence spectrum features comprises applying to one or more datasets a classification function derived from principal components analysis, the one or more datasets collectively including the one or more Raman spectrum features and the one or more background fluorescence spectrum features. 
     
     
         7 . A method according to  claim 6  wherein the one or more datasets include one or more Raman spectrum principal components scores and one or more background fluorescence spectrum principal components scores. 
     
     
         8 . A method according to  claim 6  wherein applying the classification function comprises applying a predetermined principal component to data of the one or more datasets. 
     
     
         9 . A method according to  claim 1  wherein the tissue is skin of a part of a subject's body and the method comprises selecting the classification function corresponding to the part of the subject's body from a plurality of classification functions each corresponding to a different body region. 
     
     
         10 . A method according to  claim 9  wherein the plurality of classification functions includes classification functions corresponding to two or more of the following body parts: head, torso, hand, and arm or thigh. 
     
     
         11 . A method according to  claim 1  wherein characterizing the tissue comprises applying the Raman spectrum features and the background fluorescence spectrum features as inputs to a neural network. 
     
     
         12 . A method according to  claim 1  where the section of tissue is a section of a tissue selected from the group consisting of: skin, lung tissue, and epithelial tissue. 
     
     
         13 . A method according to  claim 12  wherein the tissue comprises epithelial tissue and the epithelial tissue comprises tissue lining the subject's gastrointestinal tract, ear, nose or throat. 
     
     
         14 . A method according to  claim 1  applied to screening for skin cancer. 
     
     
         15 . A method according to  claim 1  applied to screening for one or more conditions selected from the group consisting of: basal cell carcinoma, squamous cell carcinoma, melanoma, actinic keratosis, seborrheic keratosis, sebaceous hyperplasia, keratoacanthoma, lentigo, melanocytic nevi, dysplastic nevi, and blue nevi. 
     
     
         16 . A method according to  claim 1  wherein the first wavelength range includes wavelengths from about 800 nm to about 1000 nm. 
     
     
         17 . A method according to  claim 2  wherein the one or more Raman features include an intensity of the Raman spectrum at a Raman shift of 1445 cm −1  relative to a wavelength of the incident light. 
     
     
         18 . A method according to  claim 17  wherein the one or more Raman features include an intensity of the Raman spectrum at a Raman shift of 1269 cm −1  relative to a wavelength of the incident light. 
     
     
         19 . A method according to  claim 18  wherein characterizing the tissue comprises one or both of indicating whether the tissue is likely affected by melanoma and indicating whether the tissue is likely compound nevus tissue. 
     
     
         20 . A method according to  claim 2  wherein the one or more Raman features include an intensity of the Raman spectrum at a Raman shift of 1269 cm −1  relative to a wavelength of the incident light. 
     
     
         21 . A method according to  claim 2  wherein the one or more Raman features include features within a band having a Raman shift in the range of about 1200 cm −1  to about 1400 cm −1  relative to a wavelength of the incident light. 
     
     
         22 . A method according to  claim 2  wherein the Raman features include features within a band having a Raman shift in the range of about 1500 cm −1  to about 1650 cm −1  relative to a wavelength of the incident light. 
     
     
         23 . A method according to  claim 1  wherein the Raman spectrum includes first and second peaks at Raman shifts of approximately 1368 cm −1  and 1572 cm −1  relative to a wavelength of the incident light, the method further comprising determining a melanin content of the tissue by performing steps comprising:
 computing the melanin content of the tissue based upon intensities of the first and second peaks in the Raman spectrum. 
 
     
     
         24 . Apparatus for characterizing tissues in vivo, the apparatus comprising:
 a light source for illuminating a tissue;   an optical system configured to collect and direct backscattered light from the tissue into a spectrometer;   a data processor connected to receive raw spectrum information for the backscattered light from the spectrometer;   at least one classification function accessible to the data processor, the classification function producing a classification result in response to an input, the input including information about at least one Raman feature and at least one background autofluorescence feature of a tissue spectrum;   wherein the data processor is configured to:
 process the raw spectrum to obtain both a Raman spectrum of the tissue and a background autofluorescence spectrum of the tissue that is a background to the Raman spectrum; 
 process the Raman spectrum to extract one or more features of the Raman spectrum and process the background fluorescence spectrum to extract one or more features of the background fluorescence spectrum; and, 
 characterize the tissue based by applying the one or more features of the Raman spectrum and the one or more features of the background fluorescence spectrum as inputs to the classification function. 
   
     
     
         25 . Apparatus according to  claim 24  comprising a plurality of classification functions and a mechanism for permitting a user to select one of the plurality of classification functions to be used to characterize the tissues. 
     
     
         26 . A method for determining a melanin content of tissue, the method comprising:
 obtaining a spectrum of the tissue, the spectrum including first and second peaks at Raman shifts of approximately 1368 cm −1  and 1572 cm −1 ;   subtracting a background of the spectrum to yield a Raman spectrum;   computing the melanin content of the tissue based upon intensities of the first and second peaks in the Raman spectrum.