Image Unmixing via Local Focus-Based Reference Vector Segmentation

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing methods for unmixing image data in biological specimens struggle with accurately adjusting reference vectors due to variations in imaging focus and other local properties, leading to inconsistent unmixing results.

Innovation Solution

The method involves classifying multi-channel images into regions based on local imaging characteristics, determining reference vectors specific to each region, and unmixing each region separately to produce consistent stain channel images, even with varying focus quality.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If a single set of reference vectors is used for unmixing the entire image, then the unmixing process is simple and fast, but the unmixing results are inconsistent in regions with varying focus quality

Engineering Contradiction:
Improveunmixing consistencyVSAvoidunmixing process complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The image is segmented into multiple regions based on local focus quality metrics. Each region is then processed separately with its own optimized reference vectors, allowing the unmixing process to adapt to local variations in focus while maintaining overall consistency across the entire image.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

Different reference vectors are applied to different regions of the image based on their local focus quality characteristics. This local adaptation ensures that each region is unmixed with the most appropriate reference vectors for its specific conditions, improving overall unmixing consistency.

Inventive Principle:
Principle #3Local quality

2Measurement precision

If reference vectors are adjusted to match local focus quality, then unmixing accuracy improves in varying focus conditions, but the computational complexity increases

Engineering Contradiction:
Improveunmixing accuracyVSAvoidprocessing time
Core Design Contradiction:
Measurement precisionVSLoss of time

Solution Approach 1:

Focus quality metrics are calculated and regions are identified before the unmixing process begins. This preliminary classification allows the system to prepare appropriate reference vectors for each region in advance, avoiding iterative adjustments during unmixing and reducing overall processing time.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The system automatically determines focus quality metrics and selects appropriate reference vectors for each region without requiring manual intervention. This self-service approach streamlines the process and reduces computational overhead compared to manual reference vector selection.

Inventive Principle:
Principle #25Self-service

3Reliability

If multiple sets of reference vectors are determined for different regions, then unmixing results become consistent across varying image qualities, but the overall processing complexity increases

Engineering Contradiction:
Improveunmixing result consistencyVSAvoidprocessing system complexity
Core Design Contradiction:
ReliabilityVSDevice complexity

Solution Approach 1:

The reference vectors are modified as a function of local focus quality parameters. By systematically varying the reference vectors based on measured focus metrics, the system achieves consistent unmixing results across different focus conditions while maintaining a structured approach that limits processing complexity.

Inventive Principle:
Principle #35Parameter changes

Data Source

PatentEP3391288B1Systems and methods of unmixing images with varying acquisition properties
Publication Date: 2023.07.12 VENTANA MEDICAL SYSTEMS INC
  • EP3391288B1 patent drawingFigure 1A~1B
  • EP3391288B1 patent drawingFigure 2A~2B
  • EP3391288B1 patent drawingFigure 3

AI summary

Systems and methods for unmixing of multichannel image data in the presence of locally varying image characteristics. Feature images created from a multi-channel input image form a feature vector for each pixel in the input image. Feature vectors are classified based on the local image characteristics, and areas are formed in the input image that share local image characteristics. Each area is unmixed separately using reference vectors that were obtained from regions in example images that have the same image characteristics. The unmixing results are combined to form a final unmixing result image created from the individually unmixed image areas.