Systems and methods for segmentation and processing of tissue images and feature extraction from same for treating, diagnosing, or predicting medical conditions

• US 9,971,931 B2
• Filed: 07/26/2017
• Issued: 05/15/2018
• Est. Priority Date: 07/30/2010
• Status: Active Grant

• Current Assignee: Fundação D. Anna De Sommer Champalimaud E Dr. Carlos Montez Champalimaud
• Original Assignee: Fundação D. Anna De Sommer Champalimaud E Dr. Carlos Montez Champalimaud
• Inventors: Ajemba, Peter, Scott, Richard, Ramachandran, Janakiramanan, Zeineh, Jack, Donovan, Michael, Fernandez, Gerardo, Liu, Qiuhua, Khan, Faisal
• Primary Examiner(s): Rahmjoo, Manuchehr
• Application Number:
  US15/660,727
• Publication Number:
  US 20180025212A1
• Time in Patent Office: 293 Days
• Field of Search: 382128
• CPC Class Codes: G06T 2207/10056 Microscopic image
• CPC Class Codes: G06T 2207/20041 Distance transform
• CPC Class Codes: G06T 2207/20212 Image combination
• CPC Class Codes: G06T 2207/30024 Cell structures in vitro; Tissue sections in vitro
• CPC Class Codes: G06T 5/94 based on local image properties, e.g. for local contrast enhancement
• CPC Class Codes: G06T 7/0012 Biomedical image inspection
• CPC Class Codes: G06T 7/11 Region-based segmentation
• CPC Class Codes: G06T 7/155 involving morphological operators
• CPC Class Codes: G06T 7/45 using co-occurrence matrix computation
• CPC Class Codes: G06V 20/698 Matching; Classification

Abstract

Apparatus, methods, and computer-readable media are provided for segmentation, processing (e.g., preprocessing and/or postprocessing), and/or feature extraction from tissue images such as, for example, images of nuclei and/or cytoplasm. Tissue images processed by various embodiments described herein may be generated by Hematoxylin and Eosin (H&E) staining, immunofluorescence (IF) detection, immunohistochemistry (IHC), similar and/or related staining processes, and/or other processes. Predictive features described herein may be provided for use in, for example, one or more predictive models for treating, diagnosing, and/or predicting the occurrence (e.g., recurrence) of one or more medical conditions such as, for example, cancer or other types of disease.

Claims

1. A system for predicting the occurrence of a medical condition, the system comprising:

(1) a database configured to store patient data, including at least one sample image of a tissue sample treated with a plurality of flurochrome labeled antibodies; and

(2) a processor configured by code executing therein to perform the following;

(a) generate a patient image dataset, using the at least one sample image, that includes values for one or more texture features selected from a group of features consisting of (i) homogeneity and (ii) correlation;

(b) evaluate at least the patient image dataset with a Support Vector Regression for Censored Data (SVRc) algorithm executed as code by the processor, where the SVRc algorithm is configured to output a value corresponding to a risk score for a medical condition occurrence based on the patient image dataset, wherein the SVRc algorithm is generated by performing regression, using code executed in the processor, on a population dataset, where each member of the population has measurement values corresponding to each feature of the patient image dataset;

(c) assign the patient to a high probability of a medical condition occurrence where the output value is below a pre-determined threshold and assign the patient to a low probability of a medical condition occurrence where the output value is above the pre-determined threshold; and

(d) generate a report based on the updated patient dataset.

2. The system of claim 1, wherein the processor is configured to determine homogeneity according to:

3. The system of claim 1, wherein the processor is configured to determine correlation according to:

4. The system of claim 1, wherein the processor is configured to evaluate the patient dataset using the SVRC algorithm, the patient dataset comprising the patient image dataset, one or more clinical features, and one or more molecular features, and the SVRC algorithm is generated by performing regression on a population dataset where each member of the population dataset has data corresponding to each of the features found in the patient dataset.

5. A system of predicting occurrence of a medical condition, the system comprising:

(1) a database configured to store patient data, including at least one sample image of a tissue sample treated with a plurality of flurochrome labeled antibodies; and

(2) a processor configured by code executing therein to perform the following;

(a) generate a patient morphometric dataset, using the at least one sample image, that includes values for at least one of the following features (1) one or more values selected from a group of ring measurements of one or more ring metrics; and

(2) one or more values derived from one or more ring metrics, wherein the one or more values represents an adjacency relationship between rings;

(b) evaluate at least the patient morphometric dataset with a Support Vector Regression for Censored Data (SVRc) algorithm executed as code by the processor, where the SVRc algorithm is configured to output a value corresponding to a risk score for a medical condition occurrence based on the morphometric dataset, wherein the SVRc algorithm is generated by performing regression, using code executed in the processor, on a population dataset, where each member of the population has measurement values corresponding to each feature of the patient morphometric dataset;

(c) assign the patient to a high probability of a medical condition occurrence where the output model value is below a pre-determined threshold and assign the patient to a low probability of a medical condition occurrence where the output value is above the pre-determined threshold; and

(d) generate a report based on the updated patient dataset.

6. The system of claim 5, wherein the medical condition is prostate cancer.

7. The system of claim 5, wherein the patient morphometric dataset further includes at least one value corresponding to (a) an outer diameter of a ring, inner diameter of a ring, a border gap, lumen or clearing diameter, a border density, a lumen ratio, a proportion of border touching inner clearing, a proportion of border touching stroma, a ratio of border less than a predefined number of pixels from stroma, a mean distance of border pixels from stroma, a width of epithelial padding between ring and stroma;

(b) feature(s) derived from any of the foregoing values; and

(c) feature(s) representing an adjacency relationship between rings.

8. The system of claim 5, wherein the patient morphometric dataset includes at least one value corresponding to one or more features derived from a ring segmentation and/or ring classification.

9. The system of claim 5, wherein the processor is configured to evaluate the patient dataset using the SVRC algorithm, the patient dataset comprising the patient morphometric dataset, one or more clinical features, and one or more molecular features, and the SVRC algorithm is generated by performing regression on a population dataset where each member of the population dataset has data corresponding to each of the features found in the patient dataset.

10. The system of claim 9, wherein the patient dataset further includes at least one value corresponding to one or more features selected from the group of ring combination features consisting of (i) a feature generated by localizing a feature per ring, (ii) a set of values characterizing morphological components of a ring, and (iii) a combined image feature generated by combining one or more localized features, localized per ring, with one or more morphological components of a ring.

11. The system of claim 10, wherein the one or more ring combination features comprises a biomarker feature in combination with a morphological ring feature.

12. The system of claim 10, wherein the one or more ring combination features comprises an androgen receptor (AR) biomarker intensity value in combination with a morphological ring feature.

13. The system of claim 10, wherein the one or more ring combination features comprises a Ki67 area feature in combination with a morphological ring feature.

14. The system of claim 10, wherein the one or more ring combination features comprises a morphologic or architectural marker in combination with a morphological ring feature.

15. The system of claim 14, wherein the morphologic or architectural marker comprises a high molecular weight cytokeratin, HMWCK, or AMACR marker.

16. A system of predicting occurrence of a medical condition, the system comprising:

(1) a database configured to store patient data, including at least one sample image of a tissue sample treated with a plurality of flurochrome labeled antibodies; and

(2) a processor configured by code executing therein to perform the following;

(a) generate a patient image dataset, using the at least one sample image, that includes values for at least one of the following features;

(1) a feature generated based upon a comparison of histograms, said histograms corresponding to compartments or sub-compartments of cellular objects and (2) a feature generated from an intensity index corresponding to image signal intensity;

(b) evaluate at least the patient image dataset with a Support Vector Regression for Censored Data (SVRc) algorithm executed as code by the processor, where the SVRc algorithm is configured to output a value corresponding to a risk score for a medical condition occurrence based on the morphometric dataset, wherein the SVRc algorithm is generated by performing regression, using code executed in the processor, on a population dataset, where each member of the population has measurement values corresponding to each feature of the patient image dataset;

(c) assign the patient to a high probability of a medical condition occurrence where the output model value is below a pre-determined threshold and assign the patient to a low probability of a medical condition occurrence where the output value is above the pre-determined threshold; and

(d) generate a report based on the updated patient dataset.

17. The system of claim 16, wherein the processor is configured to evaluate the patient dataset using the SVRC algorithm, the patient dataset comprising the patient image dataset, one or more clinical features, and one or more molecular features, and the SVRC algorithm is generated by performing regression on a population dataset where each member of the population dataset has data corresponding to each of the features found in the patient dataset.