SYSTEMS AND METHODS FOR QUANTITATIVE ANALYSIS OF HISTOPATHOLOGY IMAGES USING MULTICLASSIFIER ENSEMBLE SCHEMES
3 Assignments
0 Petitions
Accused Products
Abstract
Described herein are systems and methods for performing multi-stage detection and classification of cancer regions from digitized images of biopsy slides. Novel methods for processing the digitized images to improve feature extraction and structure identification are disclosed, including but not limited to the use of quaternions, logarithmic mappings of color channels, and application of wavelets to logarithmic color channel mappings. The extracted features are utilized in improved machine learning algorithms that are further optimized to analyze multiple color channels in multiple dimensions. The improved machine learning algorithms include techniques for accelerating the training of the algorithms, making their application to biopsy detection and classification practical for the first time. The performance of the described systems and methods are further improved by the disclosure of a novel multistage machine learning scheme, in which additional classifiers are utilized to choose among the classes proposed by other classifiers in close cases.
-
Citations
58 Claims
-
1-16. -16. (canceled)
-
17. A method for performing multi-stage classification for detection of cancer regions from one or more digitized images of biopsy slides comprising:
-
pre-processing from the digitized images, biopsy regions of interest or biopsy whole-slides, using color normalization methods and segmentation methods; decomposing the pre-processed digitized images into several segments; describing the image segments by forming feature vectors based on at least one of textural, color, or morphology tissue characteristics; classifying the image segments using one or more classification methods in a first classification stage, wherein the classification methods utilize the feature vectors of the image segments, and further wherein the classification methods are able to produce scores and posterior conditional probabilities or equivalents for each possible class; classifying one or more of the image segments using a second classification stage if one or more conditions are met; classifying, based upon either the first or second classification stage, a portion of the region of interest or biopsy whole slides, as cancerous or non-cancerous; determining a cancer classification, tumor extent, tumor size, or tumor localization based at least in part upon the output of either of the first or second classification stage. - View Dependent Claims (18, 19, 20, 21, 23, 24, 25, 26, 51, 53, 57)
-
-
22. (canceled)
-
27-41. -41. (canceled)
-
42. A method for robust classification of biopsy images using multi-modal expert classifiers and meta-classification, the method comprising:
-
randomizing the available input data, such that the patterns are divided into 2 groups for system training/validation and testing; pre-processing input biopsy regions of interest or biopsy whole-slides using color normalization methods and segmentation methods; describing the image data and forming feature vectors based on textural, color, and morphology tissue characteristics; creating one or more classification blocks, comprising one or more base-level classifiers, grading classifiers, and a unit for combining the predictions of at least some of the base-level classifiers; creating a classification refinement block for use when a classification margin is small; determining a cancer classification or grade, tumor extension, tumor localization using the one or more classification blocks. - View Dependent Claims (43, 44, 45, 46, 47, 48, 49)
-
-
50. A method for automatically detecting, classifying and grading cancerous regions from digitized regions of interest or whole-slides comprising:
-
performing an image color and/or illumination standardization procedure; splitting image into smaller regions if the image is a whole-slide digitized biopsy; determining and grading a location of a tumor in the image by; transforming the image from RGB to RrGrBr in accordance with the following equations; - View Dependent Claims (52)
-
-
54-56. -56. (canceled)
-
58-62. -62. (canceled)
Specification