Automated IHC Nuclei Classification via H&E Registration

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

The traditional workflow for immunoscore computation in immunohistochemical (IHC) slide analysis is highly subjective and biased, as it relies on manual selection and counting of fields of view (FOVs) or regions of interest (ROIs) by pathologists, leading to non-reproducible results.

Innovation Solution

A computer-implemented method that receives input images, including a biomarker image and an H&E image, performs analysis to derive features, registers the images to form a merged image, and classifies nuclei based on the merged features set, thereby reducing subjectivity and improving reproducibility.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Ease of operation

If manual selection and counting of FOVs or ROIs is performed by pathologists, then the process allows for expert judgment and interpretation, but the results become highly subjective and biased, reducing reproducibility

Engineering Contradiction:
Improvemanual interpretation capabilityVSAvoidreproducibility
Core Design Contradiction:
Ease of operationVSReliability

Solution Approach 1:

The patent replaces the manual mechanical process of FOV/ROI selection and cell counting with an automated image analysis system using computer algorithms. The system automatically identifies and counts immune cells in IHC images without human intervention, eliminating subjective bias while maintaining analytical capability through computational methods

Inventive Principle:
Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

The system enables self-service by allowing the image analysis process to perform its own FOV/ROI selection and cell counting functions autonomously. The automated algorithm independently identifies regions of interest and quantifies immune cells without requiring pathologist input, making the process self-sufficient and reproducible

Inventive Principle:
Principle #25Self-service

2Reliability

If automated image analysis is implemented, then reproducibility and objectivity are improved, but the system complexity and computational requirements increase

Engineering Contradiction:
ImprovereproducibilityVSAvoidsystem complexity
Core Design Contradiction:
ReliabilityVSDevice complexity

Solution Approach 1:

The patent segments the complex image analysis task into distinct modular components: image preprocessing, FOV/ROI identification, immune cell detection, and quantification. Each module performs a specific function independently, making the overall complex system manageable through functional decomposition and reducing implementation difficulty

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The system introduces an intermediary computational layer that bridges the gap between raw IHC images and final quantitative results. This intermediary processing stage applies standardized algorithms to objectively transform images into measurable data, reducing the need for complex manual intervention while maintaining analytical rigor

Inventive Principle:
Principle #24Intermediary (Mediator)

Data Source

PatentEP4567728A1Computer scoring based on primary stain and immunohistochemistry images
Publication Date: 2025.06.11 VENTANA MEDICAL SYSTEMS INC
  • EP4567728A1 patent drawingFigure 1
  • EP4567728A1 patent drawingFigure 2
  • EP4567728A1 patent drawingFigure 3A

AI summary

Described herein are computer-implemented methods for analysis of a tissue sample. An example method includes: annotating the whole tumor regions or set of tumorous sub-regions either on a biomarker image or an H&E image (e.g. from an adjacent serial section of the biomarker image); registering at least a portion of the biomarker image to the H&E image; detecting different cellular and regional tissue structures within the registered H&E image; computing a probability map based on the different detected structures within the registered H&E image; deriving nuclear metrics from each of the biomarker and H&E images; deriving probability metrics from the probability map; and classifying tumor nuclei in the biomarker image based on the computed nuclear and probability metrics.