Robust stain detection and quantification for histological specimens based on a physical model for stain absorption

• US 20020196964A1
• Filed: 05/29/2002
• Published: 12/26/2002
• Est. Priority Date: 05/29/2001
• Status: Active Grant

• Current Assignee: Ventana Medical Systems Incorporated (Roche Holding AG)
• Original Assignee: TissueInformatics, Inc. (Roche Holding AG)
• Inventors: Fuhrman, Michael, Stone, Ronald, Abdulkarim, Othman
• Granted Patent: US 6,577,754 B2
• Time in Patent Office: Days
• US Class Current: 382/128
• 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: G01N 1/30 Staining; Impregnating ; Fixation; Dehydration; Multistep processes for preparing samples of tissue, cell or nucleic acid material and the like for analysis
• CPC Class Codes: G01N 2015/1472 with colour
• CPC Class Codes: G06T 2207/30024 Cell structures in vitro; Tissue sections in vitro
• CPC Class Codes: G06T 7/0012 Biomedical image inspection
• CPC Class Codes: G06T 7/90 Determination of colour characteristics
• CPC Class Codes: G06V 10/60 relating to illumination properties, e.g. using a reflectance or lighting model
• CPC Class Codes: G06V 20/695 Preprocessing, e.g. image segmentation

Abstract

A physics based model of the absorption of light by histological stains used to measure the amount of one or more stains at locations within tissue is disclosed. The subsequent analysis results in several improvements in the detection of tissue on a slide, improvements to autofocus algorithms so focusing during image acquisition is confined to tissue, improvements to image segmentation and identification of tissued and its features, improvements to the identification of stain where multiple stains are used, and improvements to the quantification of the extent of staining. The invention relates to the application of these improvements to stain detection and quantification to provide for objective comparison between tissues and closer correlation between the presentations of such features and concurrent patterns of gene or protein expression.

Claims

1. A method for analyzing the amount of stain on tissue specimens, comprising the steps of

capturing a color image of a tissue specimen;

gathering a background image;

calculating a mean value in each band of a background image;

randomly sampling a fixed number of points from the color image;

performing a principal component analysis to obtain three vectors; and

transforming the color image to a colorspace in which the colors of the colorspace show the type and amount of stain present.

2. The method of claim 1, wherein the stain is from a hematoxylin dye containing solution.

3. The method of claim 1, wherein the stain is from hematoxylin and eosin dyes containing solution.

4. The method of claim 1, wherein the color image has three color bands.

5. The method of claim 4, wherein the three color bands are red, green and blue.

6. A method for analyzing the amount of stain on tissue specimens, comprising the steps of

capturing a color image of a tissue specimen;

transforming the color image to a colorspace representation in which points are represented by triplets of numbers comprising intensities of red, green, and blue colors;

defining a stain curve defined in terms of a light source and a certain point in the color image;

calculating the distance in colorspace along the stain curve from the triplet representing the light source to the triplet representing the certain point; and

calculating the amount of stain.

7. The method of claim 6, wherein the stain is from a hematoxylin dye containing solution.

8. The method of claim 6 wherein the stain is from hematoxylin and eosin dyes containing solution.

9. A method for analyzing the amount of stain on tissue specimens, comprising the steps of

capturing a color image of a tissue specimen;

forming a three-dimensional colorspace describing the color image;

performing a principal component analysis on the three-dimensional colorspace;

using the results of the principal component analysis to form a transformation matrix; and

calculating the amount of stain by using the transformation matrix.

10. A method for measuring the amount of stain on a stained tissue specimen, comprising the steps of

staining a tissue specimen;

creating a colored image of the tissue specimen, wherein the image is comprised of pixels, wherein each pixel comprises a red, green and blue sub-pixel, wherein each sub-pixel has an intensity and the intensities of each pixel define a triple of intensity values;

gathering a background image;

converting the triple of intensity values into a triple of optical density values;

randomly sampling the optical density values to obtain a number of values large enough to obtain statistically significant results;

performing a principal component analysis on the randomly sampled optical density values to obtain a new coordinate system; and

using the new coordinate system to define a transformation matrix to convert the triple of optical density values into a triple of values, wherein a component of the triple of values is proportional to the amount of stain at that point.

11. A method for measuring the amount of tissue feature on a stained tissue specimen, comprising the steps of

staining a tissue specimen such that a feature of the tissue specimen is stained;

creating a colored image of the tissue specimen, wherein the image is comprised of pixels, wherein each pixel comprises a red, green and blue sub-pixel, wherein each sub-pixel has an intensity and the intensities of each pixel define a triple of intensity values;

gathering a background image;

converting the triple of intensity values into a triple optical density values;

randomly sampling the optical density values to obtain a number of values large enough to obtain statistically significant results;

performing a principal component analysis on the randomly sampled optical density values to obtain a new coordinate system; and

using the new coordinate system to define a transformation matrix to convert the triple of optical density values into a triple of values, wherein a component of the triple of values is proportional to the amount of stain at that point and wherein the amount of stain at that point is proportional to the amount of the tissue feature.

12. A method for measuring the amount of two stains on a stained tissue specimen, comprising the steps of

staining a tissue specimen with two stains, a first stain and a second stain;

creating a colored image of the tissue specimen, wherein the image is comprised of pixels, wherein each pixel comprises a red, green and blue sub-pixel, wherein each sub-pixel has an intensity and the intensities of each pixel define a triple of intensity values;

converting the triple of intensity values into a triple of difference values linearly related to the amount of stains;

randomly sampling the difference values to obtain a number of values large enough to obtain statistically significant results;

performing a principal component analysis on the randomly sampled difference values to obtain a new coordinate system; and

using the new coordinate system to define a transformation matrix to convert the triple of difference values into a triple of values, wherein one component of the triple of values is proportional to the amount of the first stain at that point and a second component of the triple of values is proportional to the amount of the second stain at that point.

13. A method for measuring the amount of stains wherein there are three stains on a stained tissue specimen, comprising the steps of

staining a tissue specimen with three stains, a first stain, a second stain, and a third stain;

creating a colored image of the tissue specimen, wherein the image is comprised of pixels, wherein each pixel comprises a red, green and blue sub-pixel, wherein each sub-pixel has an intensity and the intensities of each pixel define a triple of intensity values;

gathering a background image;

using a filter to filter out the color of the third stain;

converting the triple of intensity values into a triple of difference values linearly related to the amount of stains;

randomly sampling the difference values to obtain a number of values large enough to obtain statistically significant results;

performing a principal component analysis on the randomly sampled difference values to obtain a new coordinate system; and

using the new coordinate system to define a transformation matrix to convert the triple of difference values into a triple of values, wherein one component of the triple of values is proportional to the amount of the first stain at that point and a second component of the triple of values is proportional to the amount of the second stain at that point.