Systems and methods for finding regions of in interest in hematoxylin and eosin (HandE) stained tissue images and quantifying intratumor cellular spatial heterogeneity in multiplexed/hyperplexed fluorescence tissue

A method of identifying regions of interest in a hematoxylin and eosin (H&E) stained tissue image, comprising:

• receiving image data representing the H&E stained tissue image;
• quantifying local spatial statistics for the H&E stained tissue image based on the received image data;
• identifying histological structures within the H&E stained tissue image based on the local spatial statistics;
• and generating a segmented H&E stained tissue image using the received image data and the identified histological structures, wherein the image data is H&E image data in an opponent color space obtained by transforming RGB data representing the H&E stained tissue image into the opponent color space, wherein the quantifying local spatial statistics comprises determining mutual information data indicative of statistical associations between neighboring pixels in the H&E image data, wherein the identifying histological structures uses the mutual information data and a graph-based spectral segmentation algorithm, wherein the H&E image data comprises H&E-hue data in the opponent color space, wherein the determining mutual information data comprises estimating a joint distribution of hue angles between neighboring pixels in the H&E-hue data, and calculating a pointwise mutual information (PMI) of the joint distribution, the PMI being the mutual information data, wherein the identifying comprises creating an affinity function from the PMI and detecting boundaries based on the affinity function using spectral clustering, wherein the estimating the joint distribution uses a mixture of bivariate von Mises distribution.

A method of identifying regions of interest in a hematoxylin and eosin (H&E) stained tissue image, comprising:

• receiving image data representing the H&E stained tissue image;
• quantifying local spatial statistics for the H&E stained tissue image based on the received image data;
• identifying histological structures within the H&E stained tissue image based on the local spatial statistics;
• and generating a segmented H&E stained tissue image using the received image data and the identified histological structures, wherein the image data is RGB data representing the H&E stained tissue image, and wherein the quantifying local spatial statistics is based on inter-nuclei distance distributions determined from the RGB data, wherein the quantifying comprises identifying putative nuclei locations from the RGB data in the form of superpixels, building a superpixel graph based on a pointwise distance between each superpixel and a number of its nearest neighbors, and clustering the super pixels into labeled segments, and wherein the identifying comprises merging the labeled segments into the histological structures.

A computerized system for identifying regions of interest in a stained tissue image, comprising:

• a processing apparatus, wherein the processing apparatus includes:
• a quantifying component configured for quantifying local spatial statistics for the stained tissue image based on received image data representing the stained tissue image, wherein the image data is normalized image data in an opponent color space obtained by transforming RGB data in an RGB space that represents the stained tissue image into the opponent color space to create opponent color space data and normalizing the opponent color space data, and wherein the quantifying component is configured for quantifying the local spatial statistics by determining mutual information data indicative of statistical associations between neighboring pixels in the image data;
• an identifying component configured for identifying and detecting boundaries of histological structures within the stained tissue image based on the quantified local spatial statistics using the determined mutual information data and a graph-based spectral segmentation algorithm;
• and
• a segmented tissue image generating component configured for generating a segmented stained tissue image in which the identified histological structures are segmented using the received image data representing the stained tissue image and the detected boundaries of the identified histological structures.

A computerized system for identifying regions of interest in a hematoxylin and eosin (H&E) stained tissue image, comprising:

• a processing apparatus, wherein the processing apparatus includes:
• a quantifying component configured for quantifying local spatial statistics for the H&E stained tissue image based on received image data representing the H&E stained tissue image;
• an identifying component configured for identifying histological structures within the H&E stained tissue image based on the local spatial statistics;
• and a segmented tissue image generating component configured for generating a segmented H&E stained tissue image using the received image data and the identified histological structures, wherein the image data is H&E image data in an opponent color space obtained by transforming RGB data representing the H&E stained tissue image into the opponent color space, wherein the quantifying component is configured for quantifying local spatial statistics by determining mutual information data indicative of statistical associations between neighboring pixels in the H&E image data, and wherein the identifying component is configured for identifying histological structures by using the mutual information data and a graph-based spectral segmentation algorithm, wherein the H&E image data comprises H&E-hue data in the opponent color space, wherein the determining mutual information data comprises estimating a joint distribution of hue angles between neighboring pixels in the H&E-hue data, and calculating a pointwise mutual information (PMI) of the joint distribution, the PMI being the mutual information data, wherein the estimating the joint distribution uses a mixture of bivariate von Mises distribution.

A computerized system for identifying regions of interest in a hematoxylin and eosin (H&E) stained tissue image, comprising:

• a processing apparatus, wherein the processing apparatus includes:
• a quantifying component configured for quantifying local spatial statistics for the H&E stained tissue image based on received image data representing the H&E stained tissue image;
• an identifying component configured for identifying histological structures within the H&E stained tissue image based on the local spatial statistics;
• and a segmented tissue image generating component configured for generating a segmented H&E stained tissue image using the received image data and the identified histological structures, wherein the image data is RGB data representing the H&E stained tissue image, and wherein the quantifying component is configured for quantifying local spatial statistics based on inter-nuclei distance distributions determined from the RGB data, wherein the quantifying comprises identifying putative nuclei locations from the RGB data in the form of superpixels, building a superpixel graph based on a pointwise distance between each superpixel and a number of its nearest neighbors, and clustering the superpixels into labeled segments, and wherein the identifying component is configured for identifying by merging the labeled segments into the histological structures.

A method of identifying regions of interest in a stained tissue image, comprising:

• receiving image data representing the stained tissue image, wherein the image data is normalized RGB data, wherein the normalized RGB data is obtained by transforming RGB data in an RGB space that represents the stained tissue image into an opponent color space to create opponent color space data, normalizing the opponent color space data to create normalized opponent color space data, and converting the normalized opponent color space data to the RGB space;
• quantifying local spatial statistics for the stained tissue image based on inter-nuclei distance distributions determined from the normalized RGB data;
• identifying and detecting boundaries of histological structures within the stained tissue image based on the quantified local spatial statistics;
• and
• generating a segmented stained tissue image in which the identified histological structures are segmented using the received image data representing the stained tissue image and the detected boundaries of the identified histological structures.

A computerized system for identifying regions of interest in a stained tissue image, comprising:

• a processing apparatus, wherein the processing apparatus includes:
• a quantifying component configured for quantifying local spatial statistics for the stained tissue image based on received image data representing the stained tissue image, wherein the image data is normalized RGB data, wherein the normalized RGB data is obtained by transforming RGB data in an RGB space that represents the stained tissue image into an opponent color space to create opponent color space data, normalizing the opponent color space data to create normalized opponent color space data, and converting the normalized opponent color space data to the RGB space, wherein the quantifying component is configured for quantifying the local spatial statistics based on inter-nuclei distance distributions determined from the normalized RGB data;
• an identifying component configured for identifying and detecting boundaries of histological structures within the stained tissue image based on the quantified local spatial statistics;
• and
• a segmented tissue image generating component configured for generating a segmented stained tissue image in which the identified histological structures are segmented using the received image data representing the stained tissue image and the detected boundaries of the identified histological structures.