Assessing risk of breast cancer recurrence

Patent metadata

Field	Value
Publication number	US-11288795-B2
Application number	US-201916659545-A
Country	US
Kind code	B2
Filing date	Oct 21, 2019
Priority date	Jun 10, 2014
Publication date	Mar 29, 2022
Grant date	Mar 29, 2022

Abstract

The subject disclosure presents systems and computer-implemented methods for assessing a risk of cancer recurrence in a patient based on a holistic integration of large amounts of prognostic information for said patient into a single comparative prognostic dataset. A risk classification system may be trained using the large amounts of information from a cohort of training slides from several patients, along with survival data for said patients. For example, a machine-learning-based binary classifier in the risk classification system may be trained using a set of granular image features computed from a plurality of slides corresponding to several cancer patients whose survival information is known and input into the system. The trained classifier may be used to classify image features from one or more test patients into a low-risk or high-risk group.

First claim

The invention claimed is:

1. A system comprising:
a processor; and
a memory coupled to the processor, the memory storing computer-readable instructions that, when executed by the processor, cause the processor to perform operations comprising:
accessing a plurality of images corresponding to a tissue sample, wherein a first image of the plurality of images is a hematoxylin-and-eosin (H&E) image and a second image of the plurality of images is an individually stained or a multiplex immunohistochemical (IHC) image;
detecting, for each image of the plurality of images, a plurality of marker-specific features, wherein each marker-specific feature of plurality of the marker-specific features corresponds to a tissue object represented in an image of the plurality of images;
computing a plurality of inter-marker features, wherein each of the inter-marker features is computed based on a first value derived from a marker-specific feature of the first image and a second value derived from a marker-specific feature of the second image;
training a breast-cancer recurrence model using the plurality of inter-marker features and a clinical outcome data of a subject associated with the tissue sample; and
processing another image using the trained breast-cancer recurrence model to estimate a breast-cancer-risk recurrence score; and
outputting the breast-cancer-risk recurrence score.

2. The system of claim 1 , wherein the H&E image comprises an image of the tissue sample that is stained with one or more stains that highlight one or more structures in the tissue sample.

3. The system of claim 1 , wherein the H&E image depicts morphological features within the tissue sample.

4. The system of claim 1 , wherein said plurality of images corresponds to more than one tissue sample, each tissue sample of the more than one tissue sample being associated with a different subject and clinical outcome data corresponding to the different subject.

5. The system of claim 1 , wherein the training the breast-cancer recurrence model further comprises using features from additional tissue samples and clinical outcome data corresponding to each of the additional tissue samples.

6. The system of claim 1 , wherein the training the breast-cancer recurrence model comprises using a machine-learning operation.

7. The system of claim 6 , wherein the machine-learning operation includes using an L1-regularized logistic regression.

8. The system of claim 1 , wherein training the breast-cancer recurrence model comprises performing a bootstrap analysis.

9. The system of claim 1 , wherein training the breast-cancer recurrence model further comprise performing a cross-validation of the breast-cancer recurrence model on a subset of the plurality of images.

10. The system of claim 1 , wherein the plurality of marker-specific features includes a nucleus blob shape feature, a nucleus blob area feature, a nucleus blob compactness feature, a nucleus blob density feature, a normalized color feature, an absolute color feature, a center vote strength, an annular region based feature, or a blob topography feature.

11. The system of claim 1 , wherein the plurality of marker-specific features and the plurality of inter-marker features includes a count of lymphocytes in a vicinity of a tumor region, a proportion of Ki67− stained tumor region within an ER-stained tumor region, a size variation of Ki67 positive tumor cells, an average eccentricity of Ki67 positive tumor blobs, an average density of Ki67 and neighbors within a 50 um radius around a Ki67+ cell, a heterogeneity of a Ki67+ staining intensity, a heterogeneity of a hematoxylin stain intensity within one or more Ki67+ tumor cells, or an interquartile range of a PR+ tumor cell size.

12. The system of claim 1 , wherein the operations further comprise performing a registration operation between at least two of the plurality of images.

13. The system of claim 1 , further comprising detecting tissue objects in an image of the plurality of images by counting a number of tumor nuclei or lymphocytes in the image.

14. A computer-implemented method comprising:
accessing a plurality of images corresponding to a tissue sample, wherein a first image of the plurality of images is a hematoxylin-and-eosin (H&E) image and a second image of the plurality of images is an individually stained or a multiplex immunohistochemical (IHC) image;
detecting, for each image of the plurality of images, a plurality of marker-specific features, wherein each marker-specific feature of plurality of the marker-specific features corresponds to a tissue object represented in an image of the plurality of images;
computing a plurality of inter-marker features, wherein each of the inter-marker features is computed based on a first value derived from a marker-specific feature of the first image and a second value derived from a marker-specific feature of the second image;
training a breast-cancer recurrence model using the plurality of inter-marker features and a clinical outcome data of a subject associated with the tissue sample; and
processing another image using the trained breast-cancer recurrence model to estimate a breast-cancer-risk recurrence score; and
outputting the breast-cancer-risk recurrence score.

15. The computer-implemented method of claim 14 , wherein the H&E image comprises an image of the tissue sample that is stained with one or more stains that highlight one or more structures in the tissue sample.

16. The computer-implemented method of claim 14 , wherein the H&E image depicts morphological features within the tissue sample.

17. The computer-implemented method of claim 14 , wherein said plurality of images corresponds to more than one tissue sample, each tissue sample of the more than one tissue sample being associated with a different subject and clinical outcome data corresponding to the different subject.

18. A non-transitory computer-readable storage medium storing instructions executable by a processor to perform operations comprising:
accessing a plurality of images corresponding to a tissue sample, wherein a first image of the plurality of images is a hematoxylin-and-eosin (H&E) image and a second image of the plurality of images is an individually stained or a multiplex immunohistochemical (IHC) image;
detecting, for each image of the plurality of images, a plurality of marker-specific features, wherein each marker-specific feature of plurality of the marker-specific features corresponds to a tissue object represented in an image of the plurality of images;
computing a plurality of inter-marker features, wherein each of the inter-marker features is computed based on a first value derived from a marker-specific feature of the first image and a second value derived from a marker-specific feature of the second image;
training a breast-cancer recurrence model using the plurality of inter-marker features and a clinical outcome data of a subject associated with the tissue sample; and
processing another image using the trained breast-cancer recurrence model to estimate a breast-cancer-risk recurrence score; and
outputting the breast-cancer-risk recurrence score.

19. The non-transitory computer-readable storage medium of claim 18 , wherein the plurality of marker-specific features includes a nucleus blob shape feature, a nucleus blob area feature, a nucleus blob compactness feature, a nucleus blob density feature, a normalized color feature, an absolute color feature, a center vote strength, an annular region based feature, or a blob topography f

Assignees

• Ventana Med Syst Inc
• Univ Leland Stanford Junior

Inventors

• Barnes Michael
• Chukka Srinivas
• Knowles David

Classifications

• G06V10/806

  of extracted features · CPC title

• G06V10/771

  Feature selection, e.g. selecting representative features from a multi-dimensional feature space · CPC title

• G06F18/211

  Selection of the most significant subset of features · CPC title

• G06F18/253

  of extracted features · CPC title

• G06T2207/10056

  Microscopic image · CPC title