Image-Based Biological Particle Classification with Cascade Analysis

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing automated biological particle classification systems face challenges in balancing computational efficiency and accuracy, often requiring heavy preprocessing and cumbersome troubleshooting due to complex operations, especially in time-sensitive applications with limited computational resources.

Innovation Solution

A system utilizing an image-based feature extraction and classification architecture with a cascade classifier that includes two-level analysis, where simpler features are used for common classifications and more complex features are reserved for difficult cases, allowing for efficient and accurate classification of biological particles.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If heavy preprocessing and complex operations are used in automated biological particle classification, then measurement precision is improved, but device complexity increases

Engineering Contradiction:
Improveclassification accuracyVSAvoidsystem complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The classification system is divided into multiple sequential stages: initial filtering stage, intermediate classification stage, and detailed analysis stage. Each stage processes only the necessary subset of particles with appropriate complexity level, avoiding unnecessary computation for all particles while maintaining high classification accuracy.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

Different levels of feature extraction and analysis are applied to different particles based on their classification needs. Simple particles receive minimal processing while complex or ambiguous particles undergo more rigorous analysis, optimizing the balance between accuracy and computational complexity.

Inventive Principle:
Principle #3Local quality

2Measurement precision

If complex feature extraction is performed on all particles, then measurement precision is improved, but productivity decreases

Engineering Contradiction:
Improveclassification accuracyVSAvoidprocessing speed
Core Design Contradiction:
Measurement precisionVSProductivity

Solution Approach 1:

The particle population is segmented into different groups based on initial simple features, with only specific groups requiring complex feature extraction. This segmentation enables high-speed processing of the majority of particles while applying detailed analysis only where necessary.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

Full complex feature extraction is performed only on a partial subset of particles that require detailed classification, while other particles are classified using simpler, faster methods. This partial action approach maintains high accuracy for critical cases while preserving overall processing throughput.

Inventive Principle:
Principle #16Partial or excessive action

3Measurement precision

If a single complex classifier is used, then measurement precision is improved, but ease of operation worsens

Engineering Contradiction:
Improveclassification accuracyVSAvoidtroubleshooting difficulty
Core Design Contradiction:
Measurement precisionVSEase of operation

Solution Approach 1:

The monolithic classifier is segmented into multiple modular classification components arranged in a pipeline. Each component performs a specific classification function and can be independently configured, trained, and troubleshooted, making the overall system easier to operate and maintain while achieving high classification accuracy through the combination of components.

Inventive Principle:
Principle #1Segmentation

Data Source

PatentEP3500964B1System and method of classification of biological particles
Publication Date: 2025.07.09 IRIS INTERNATIONAL INC
  • EP3500964B1 patent drawingFigure 1A~1B
  • EP3500964B1 patent drawingFigure 2
  • EP3500964B1 patent drawingFigure 3

AI summary

A method and system for classification of cells and particles in a biological sample using an automated image-based feature extraction and classification architecture. A method operates by applying a mask or series of masks to an image, extracting features from the unmasked portions of the image based on the content and location of colored pixels, selecting a subset of the extracted features, and mapping the subset of the extracted features into a classifier architecture. In a majority of cases, the first level model architecture provides an accurate identification of the cell or particle. In a minority of cases, the classification of the cell or particle requires a second level step requiring the use of numerical or categorical values from the first level in combination with a second level model.