Histological reconstruction and automated image analysis

• US 6,418,236 B1
• Filed: 06/24/1999
• Issued: 07/09/2002
• Est. Priority Date: 06/24/1999
• Status: Expired due to Term

• Current Assignee: Carl Zeiss Microscopy GmbH (Carl-Zeiss-Stiftung)
• Original Assignee: Chromavision Medical Systems Inc. (NeoGenomics, Inc.)
• Inventors: Decker, William, Ellis, Bob, McLaren, Gina
• Primary Examiner(s): Boudreau, Leo
• Assistant Examiner(s): PATEL, KANJIBHAI B
• Application Number:
  US09/344,308
• Time in Patent Office: 1,111 Days
• Field of Search: 382/308, 382/128, 382/129, 382/130, 382/131, 382/132, 382/133, 382/134, 435/6, 435/7.23, 356/39
• US Class Current: 382/128
• CPC Class Codes: G06V 10/10 Image acquisition document image scanning and transmission H04N1/00; control of digital cameras H04N23/60
• CPC Class Codes: G06V 20/69 Microscopic objects, e.g. biological cells or cellular parts

Abstract

A method for automated image analysis of a biological specimen by histological reconstruction. Also provided is an automated cell image method for analyzing a biological specimen that has consecutively been stained by either an in situ hybridization (ISH) method, or an immunohistochemistry (IHC) method or a nucleic acid stain, and counterstained. The method couples composite images in an automated manner for processing and analysis. To identify structure in tissue that cannot be captured in a single field of view image or a single staining technique, the disclosure provides a method for histological reconstruction to analyze many fields of view on potentially many slides simultaneously.

Claims

1. A method for automated image analysis of a biological specimen, comprising:

(a) providing a biological sample to be analyzed;

(b) automatically scanning the sample at a plurality of different positions with different coordinates;

(c) automatically obtaining an image at each of said different coordinates to produce different images from different locations of the sample;

(d) automatically detecting one or more targets of interest on the sample from the different images of the sample by automatically measuring one or more parameters of the sample and automatically comparing the measured parameters against selected threshold values of the parameters for the targets of interest;

(e) automatically reconstructing an image of a selected target on the sample from the obtained different images, wherein a reconstructed image of a selected target spreading in two or more spatially adjacent and different images of the sample is reconstructed by using image information from the spatially adjacent and different images; and

(f) processing the reconstructed image to obtain information of the selected target of interest.

2. The method of claim 1, further comprising:

(a) automatically identifying a coordinate of the candidate object or area of interest in the reconstructed image; and

(b) automatically acquiring a selected image of the object or area of interest, at the location coordinates obtained from the reconstructed image, wherein the selected image includes image information from two different images obtained at two different coordinates during the scanning.

3. The method of claim 1, wherein the object or area of interest is detected by immunohistochemistry.

4. The method of claim 1, wherein the object or area of interest is detected by in situ hybridization.

5. The method of claim 1, wherein the object or area of interest is detected by a stain.

6. The method of claim 5, wherein the stain is a nucleic acid dye selected from the group consisting of hematoxylin, Giemsa stain, methyl green, Nuclear Fast-Red, Hoechst 33342, Hoechst 33258, thiazole orange, DAPI, ethidium bromide, propidium iodide, TOTO, YOYO-1, SYTOX Blue, SYTOX Green, 7-Aminoactinomycin, 9-Amino-6-chloro-2-methoxyacridine, and acridine homodimer.

7. The method of claim 5, wherein the object or area of interest is stained with a cytoplasmic dye such as eosin or Kleihauer-Betke cytochemical stain or a combination thereof.

8. The method of claim 1, wherein the object or area of interest is a cell specific marker.

9. The method of claim 8, wherein the cell specific marker is detected by a nuclear stain and counterstain.

10. The method of claim 8, wherein the cell specific marker is detected by immunohistochemistry, in situ hybridization, staining or a combination thereof.

11. The method of claim 1, wherein the reconstructed image is a digital image.

12. The method as in claim 1, further comprising:

operating an imaging system at a first magnification to obtain a first set of the different images and a first reconstructed image;

setting the imaging system at a second magnification greater than the first magnification to obtain a second set of the different images and a second reconstructed image;

using the first composite image to identify a target therein; and

using the second composite image to extract desired information about the target.

13. The method as in claim 1, wherein the one or more parameters include the hue, saturation, or intensity of a color image of the sample.

14. A method for histological reconstruction, comprising:

(a) providing a sample of a biological specimen on a series of slides, wherein consecutive samples on said slides are;

(1) stained by a nuclear stain which includes hematoxylin and eosin (H/E), or

(2) detectably labeled by an immunohistochemistry (IHC) method and counterstained, or

(3) detectably labeled by an in situ hybridization (ISH) method and counterstained, or

(4) detectably labeled by a combined (IHC) and (ISH) method and counterstained;

(b) automatically obtaining a sample image from each of the samples by scanning the slides through an imaging field of an imaging system;

(c) automatically reconstructing the sample images to produce a single image frame in which a first sample image is paired with a consecutive sample having a corresponding image;

(d) automatically performing a histological reconstruction of the specimen; and

(e) automatically detecting one or more targets of interest on the sample from the different sample images by automatically measuring one or more parameters of the sample associated with at least one of said stained and said labeled techniques in (a) and automatically comparing the measured parameters against selected threshold values of the parameters for the one or more targets of interest.

15. The method of claim 14, wherein the first sample and consecutive sample have overlapping images.

16. The method as in claim 14, wherein the one or more parameters include the hue, saturation, or intensity of a color image of the sample.

17. The method as in claim 14, further comprising:

operating an imaging system at a first magnification to obtain a first set of the sample images and a first single image frame from reconstructing the first set;

setting the imaging system at a second magnification greater than the first magnification to obtain a second set of the sample images;

using the first single image frame to identify a target therein; and

using the second set of the samples images to extract desired information about the target.

18. A computer program, residing on a computer-readable medium, for automated image analysis of biological specimen, the computer program comprising instructions for causing a computer to:

a) cause a sample to be processed, wherein the sample is stained with a nucleic acid dye and a counterstain;

(b) control a sample stage to scan the sample at a plurality of different coordinates;

(c) control an imaging system to obtain an image at each of said coordinates to produce different images of the sample at the different coordinates;

(d) reconstruct the sample from the individual images to create a reconstructed image of the sample which synthesizes the images into a single image;

(e) automatically identify a coordinate of a candidate object or area of interest in the reconstructed sample by automatically measuring one or more parameters of the sample associated with said stained technique in (a) and automatically comparing the measured parameters against selected threshold values of the parameters for the candidate object or area of interest; and

(f) in response to a user input, select a portion of the reconstructed image corresponding to the object of interest, at the location coordinates obtained from the reconstructed sample, to obtain detailed information about the selected portion.

19. The computer program of claim 18, wherein the reconstructed image of the sample is a digital image.

20. A computer program, residing on a computer-readable medium, for histological reconstruction, the computer program comprising instructions for causing a computer to:

(a) generate, digitize, and store in memory images from a biological specimen on a series of slides, wherein consecutive samples are;

(1) stained by a nuclear stain which includes hematoxylin and with eosin, or

(2) detectably labeled by an immunohistochemical (IHC) method and counterstained, or

(3) detectably labeled by an in situ hybridization (ISH) method and counterstained, or

(4) detectably labeled by a combined IHC and ISH method and counterstained;

(b) place the stored sample images in an order such that a nucleic acid stained and counterstained sample is paired with a consecutive IHC or ISH sample;

(c) automatically perform a histological reconstruction of the specimen by forming a new image of a portion or the entirety of the specimen which includes at least scenes from two different stored sample images; and

(d) automatically detecting one or more targets of interest on the sample from the different sample images by automatically measuring one or more parameters of the sample associated with at least one of said stained and said labeled techniques in (a) and automatically comparing the measured parameters against selected threshold values of the parameters for the targets of interest.

21. A method for automated image analysis of a biological specimen, comprising:

causing a biological sample to be automatically scanned at a plurality of different positions with different coordinates;

causing an image at each of said different coordinates to be automatically obtained to produce different images from different locations of the sample;

causing one or more targets of interest on the sample from the different images of the sample to be automatically detected by automatically measuring one or more parameters of the sample and by automatically comparing the measured parameters against selected threshold values of the parameters for the targets of interest;

causing an image of a selected target on the sample to be automatically reconstructed from the obtained different images, wherein a reconstructed image of a selected target spreading in two or more spatially adjacent and different images of the sample is reconstructed by using image information from the spatially adjacent and different images; and

causing reconstructed image to be processed to obtain information of the selected target of interest.

22. The method of claim 21, further comprising:

causing a coordinate of the candidate object or area of interest in the reconstructed image to be automatically identified; and

causing a selected image of the object or area of interest to be automatically acquired, at the location coordinates obtained from the reconstructed image, wherein the selected image includes image information from two different images obtained at two different coordinates during the scanning.

23. The method as in claim 21, further comprising:

causing an imaging system to operate at a first magnification to obtain a first set of the different images and a first reconstructed image;

causing the imaging system to operate at a second magnification greater than the first magnification to obtain a second set of the different images and a second reconstructed image;

causing a use of the first composite image to identify a target therein; and

causing a use of the second composite image to extract desired information about the target.