Methods, systems, and apparatuses for quantitative analysis of heterogeneous biomarker distribution
Inventors
Barnes, Michael • Chafin, David • Garsha, Karl • Grogan, Thomas M. • Roberts, Esteban • Stevens, Benjamin • Ventura, Franklin • Chefd'hotel, Christophe • Shanmugam, Kandavel • Gray, Joe • Ramunno-Johnson, Damien • Vu, Tothu Tania • Druker, Brian J. • Jacob, Thomas
Assignees
Oregon Health & Science University • Ventana Medical Systems Inc
Interested in licensing this patent?
MTEC can help explore whether this patent might be available for licensing for your application.
Publication Number
US-12579642-B2
Publication Date
2026-03-17
Expiration Date
Abstract
Methods, systems, and apparatuses for detecting and describing heterogeneity in a cell sample are disclosed herein. A plurality of fields of view (FOV) are generated for one or more areas of interest (AOI) within an image of the cell sample are generated. Hyperspectral or multispectral data from each FOV is organized into an image stack containing one or more z-layers, with each z-layer containing intensity data for a single marker at each pixel in the FOV. A cluster analysis is applied to the image stacks, wherein the clustering algorithm groups pixels having a similar ratio of detectable marker intensity across layers of the z-axis, thereby generating a plurality of clusters having similar expression patterns.
Core Innovation
The invention provides a quantitative multiplex imaging workflow to image a cell sample using a multi-spectral imaging system and to obtain multi-spectral pixels from the cell sample. The acquired multi-spectral pixels are spectrally unmixed to provide a number of n single channel images, and a cluster analysis is performed by applying an unsupervised, non-parametric, density-based clustering algorithm in an n+2 dimensional space defined by the n single channel images, thereby generating a plurality of clusters and a representation of a clustering result.
The invention characterizes heterogeneity in a cell sample comprising one or more analytes labeled with a detectable marker by analyzing an image of the cell sample on a computer apparatus. The analysis creates a cluster map comprising a plurality of clusters of expression patterns, and the dataset includes an image stack for at least a portion of the image, with an x-axis and a y-axis representing spatial coordinates and a z-axis comprising n layers, where each z-axis layer comprises intensity data for a single detectable marker at a plurality of x,y coordinates.
The cluster analysis applies an unsupervised, non-parametric, density-based clustering algorithm to the image stack such that the points, defined by x,y coordinates and n intensity data values from the n layers at the respective x,y coordinate, are clustered in the at least n+2 dimensional space. The method outputs output data representative of the result of the cluster analysis, the output data being indicative of the heterogeneity in the cell sample, and the n layers are obtained by unmixing the multi-channel image.
Claims Coverage
The partial content contains two independent claims: clm-00001 and clm-00007. Across these independent claims, the core coverage centers on unsupervised, non-parametric density-based clustering applied in at least n+2 dimensional space derived from n marker intensity layers over spatial coordinates.
Multi-spectral imaging, spectral unmixing, and n+2 dimensional density-based clustering
Acquiring image data with a multi-spectral imaging system comprising multi-spectral pixels, spectrally unmixing the multi-spectral pixels to provide n single channel images, and applying an unsupervised, non-parametric, density-based clustering algorithm in an n+2 dimensional space given by the n single channel images to generate a plurality of clusters and a representation of a clustering result.
Cluster-map heterogeneity characterization from an image stack using density-based clustering
Analyzing an image on a computer apparatus to apply cluster analysis that creates a cluster map with a plurality of clusters of expression patterns, where the dataset comprises an image stack with x,y spatial coordinates and n z-axis intensity layers for a single detectable marker, the clustering applies an unsupervised, non-parametric, density-based clustering algorithm to cluster points in an at least n+2 dimensional space defined by x,y coordinates and n intensity data values, and outputting output data representative of the clustering result indicative of heterogeneity, with the n layers obtained by unmixing a multi-channel image.
Independent claim clm-00001 covers imaging a cell sample via multi-spectral acquisition, spectral unmixing into n single-channel images, and unsupervised non-parametric density-based clustering in n+2 dimensional space to generate multiple clusters. Independent claim clm-00007 covers heterogeneity characterization by constructing an image stack with x,y spatial coordinates and n marker intensity layers, applying unsupervised non-parametric density-based clustering in at least n+2 dimensions to generate clusters of expression patterns, and outputting cluster-map data indicative of heterogeneity.
Stated Advantages
Ability to detect and represent heterogeneity in a cell sample using output data indicative of heterogeneity.
Ability to generate cluster maps comprising clusters of expression patterns.
Ability to output a clustering result representation and clustering-derived output data.
Documented Applications
Characterizing heterogeneity in a cell sample comprising one or more analytes labeled with a detectable marker by analyzing an image on a computer apparatus.
Interested in licensing this patent?