Tubular Object Recognition Using Centerline Features in Medical Imaging

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing medical imaging technologies for analyzing tubular structures, such as blood vessels and nerves, are inefficient and inaccurate, requiring manual analysis of two-dimensional layers in three-dimensional images, which hinders precise diagnosis and treatment of medical conditions.

Innovation Solution

A system utilizing a trained neural network model to automatically determine the centerline of tubular structures in medical images, enabling efficient and accurate recognition and analysis of target objects, including detection of abnormalities.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If manual analysis of two-dimensional layers is used to view tubular structures, then the doctor can analyze the structures, but the efficiency and accuracy are low

Engineering Contradiction:
Improveaccuracy of tubular structure analysisVSAvoidefficiency of tubular structure analysis
Core Design Contradiction:
Measurement precisionVSProductivity

Solution Approach 1:

The patent replaces the manual mechanical analysis process with an automated computer-based system that uses image processing algorithms and machine learning models to automatically identify and analyze tubular structures in medical images, thereby improving both efficiency and accuracy simultaneously

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

Solution Approach 2:

The patent introduces an intermediary software system that acts as a bridge between the raw medical images and the doctor's analysis, providing automated preprocessing, enhancement, and measurement functions that improve both the speed and precision of tubular structure evaluation

Inventive Principle:
Principle #24Intermediary (Mediator)

2Ease of operation

If manual analysis of each two-dimensional layer is performed, then the tubular structures can be viewed, but the process is time-consuming and inefficient

Engineering Contradiction:
Improveease of viewing tubular structuresVSAvoidtime required for analysis
Core Design Contradiction:
Ease of operationVSLoss of time

Solution Approach 1:

The patent transforms the analysis from examining individual two-dimensional layers to a three-dimensional volumetric view, allowing tubular structures to be visualized and analyzed in their complete spatial context, which dramatically reduces the time required while maintaining ease of operation

Inventive Principle:
Principle #17Another dimension (Dimensionality change)

Solution Approach 2:

The patent performs preliminary automated processing of medical images including noise reduction, contrast enhancement, and tubular structure segmentation before presentation to the doctor, thereby reducing the time required for actual analysis while making the process easier to operate

Inventive Principle:
Principle #10Preliminary action

3Productivity

If automated recognition systems are implemented, then efficiency and accuracy improve, but the system complexity increases

Engineering Contradiction:
Improveefficiency of tubular structure analysisVSAvoidcomplexity of recognition system
Core Design Contradiction:
ProductivityVSDevice complexity

Solution Approach 1:

The patent divides the complex automated recognition system into distinct functional modules including image preprocessing, tubular structure detection, feature extraction, and analysis components, which can be independently optimized and maintained, thereby managing system complexity while maintaining high productivity

Inventive Principle:
Principle #1Segmentation

Data Source

PatentEP4128040B1Systems and methods for object recognition
Publication Date: 2026.01.21 SHANGHAI UNITED IMAGING HEALTHCARE
  • EP4128040B1 patent drawingFigure 1
  • EP4128040B1 patent drawingFigure 2
  • EP4128040B1 patent drawingFigure 3

AI summary

Systems and methods for object recognition. The systems may obtain image data captured by an imaging device. The image data may include one or more objects. The systems may determine a centerline of a target object in the one or more objects based on the image data. The systems may determine a recognition result of the target object using a trained neural network model based on at least one feature parameter of the centerline of the target object. The recognition result may include a name of the target object. The systems may perform an anomaly detection on the target object based on the recognition result of the target object.