Automatic assay assessment and normalization for image processing
Inventors
Nie, Yao • Sainz de Cea, Maria V.
Assignees
Interested in licensing this patent?
MTEC can help explore whether this patent might be available for licensing for your application.
Publication Number
US-12293822-B2
Publication Date
2025-05-06
Expiration Date
Abstract
Disclosed herein are systems and methods for of assessing stain titer levels. An exemplary method includes generating a set of field of views for the image or the region of the image, selecting field of views from the set of field of views that meet predefined criteria, creating a series of patches within each of the selected field of views, retaining patches from the series of patches that meet predefined criteria indicative of a presence of the stain for which the titer is to be estimated, deriving stain color features and stain intensity features pertaining to the stain from the retained patches, estimating a titer score for each of the retained patches based on the stain color features and the stain intensity features, and calculating a weighted average score for the titer of the stain based on the estimated titer score for each of the retained patches.
Core Innovation
The invention determines an estimated titer level of a query image of a biological sample stained with a stain by classifying stain image features derived from image patches of the query image. The invention then normalizes the titer level of the query image by deriving chromatic distribution coordinates and density distribution coordinates within a color model that incorporates density information.
Based in-part on parameter values associated with the estimated titer level, the invention aligns chromatic distribution coordinates in the query image with template image chromatic distribution coordinates and provides transformed chromatic distribution coordinates. Based in-part on a second parameter value associated with the estimated titer level, the invention scales density distribution coordinates in the query image with template image density distribution coordinates to provide transformed density distribution coordinates.
The invention reconstructs an RGB image by inversely transforming the query image within the color model incorporating the density information. The inverse transformation uses weighted transformed chromatic distribution coordinates and weighted transformed density distribution coordinates, with weighting derived from probabilities that pixels are stain pixels.
Claims Coverage
The provided partial content includes three independent claims: a method claim, a system claim, and a non-transitory computer-readable medium claim. Each independent claim includes the same core inventive sequence of estimating a stain titer level and normalizing it using chromatic and density distribution coordinate transformations within a density-incorporating color model, followed by RGB reconstruction using weighted transformed coordinates.
Estimating a stain titer level from patched stain features
Determining an estimated titer level of a query image by classifying stain image features derived from image patches of the query image.
Normalizing titer level via chromatic and density coordinate transformations in a density-incorporating color model
Normalizing the titer level of the query image by deriving chromatic distribution coordinates and density distribution coordinates in a color model incorporating density information, aligning chromatic distribution coordinates based in-part on a first parameter value, and scaling density distribution coordinates based in-part on a second parameter value.
Reconstructing RGB image using inverse transformation with weighted transformed coordinates
Reconstructing an RGB image by inversely transforming the query image within the color model incorporating the density information using weighted transformed chromatic distribution coordinates and weighted transformed density distribution coordinates.
Across the independent claims, titer estimation from patched stain image features is combined with density-incorporating color-model coordinate transformations, including chromatic alignment and density scaling using parameter values associated with the estimated titer, and RGB reconstruction using weighted transformed chromatic and density distribution coordinates.
Stated Advantages
Improved consistency of cell-detection outputs across varying HTX titers while preserving non-HTX stains.
Documented Applications
Digital pathology system evaluation of cell detection algorithm outputs across varying HTX titers.
Interested in licensing this patent?