Automated Artery Detection in CTA via Multi-Scale Segmentation

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current medical imaging technologies face challenges in automatically detecting systemic arteries in computed tomography angiography (CTA) due to non-uniform intensity, anatomical variations, and computational demands, particularly in processing large 3D scans with calcifications and stenosis, which limits the accuracy and efficiency of artery centerline extraction.

Innovation Solution

A semi- or fully-automated method for arterial tree reconstruction that uses a multi-scale approach, body region detection, anatomy-driven connection rules, and a continuous genetic algorithm to optimize parameters, enabling the detection of vessels across various body parts and field-of-view CTA scans, independent of image intensity contrast and requiring minimal user interaction.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If complete analysis of vascular network is performed, then measurement precision is improved, but computing time increases

Engineering Contradiction:
Improveaccuracy of artery centerline extractionVSAvoidcomputing time
Core Design Contradiction:
Measurement precisionVSLoss of time

Solution Approach 1:

The patent divides the vascular network analysis into multiple scales, starting with coarse-scale detection to identify major vessels, then progressively refining to finer scales for detailed centerline extraction. This multi-scale segmentation approach reduces the overall computational burden while maintaining accuracy by processing different regions at appropriate levels of detail.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent performs preliminary vessel enhancement and segmentation before centerline extraction. By pre-processing the image to enhance vessel structures and segment them from surrounding tissues, the method prepares the data in advance, making the subsequent centerline extraction faster and more accurate without requiring complete re-analysis of the entire vascular network.

Inventive Principle:
Principle #10Preliminary action

2Ease of operation

If automated artery detection is implemented, then ease of operation is improved, but reliability decreases due to imaging variations and anatomical variations

Engineering Contradiction:
Improveautomation level of artery extractionVSAvoidaccuracy under imaging and anatomical variations
Core Design Contradiction:
Ease of operationVSReliability

Solution Approach 1:

The patent applies different processing strategies and parameters to different regions of the vascular network based on local characteristics. For example, different scales and enhancement techniques are used for large arteries versus small vessels, and parameters are adapted to local imaging conditions and anatomical variations, improving reliability while maintaining automation.

Inventive Principle:
Principle #3Local quality

Solution Approach 2:

The patent dynamically adjusts processing parameters based on detected vessel characteristics, imaging quality, and anatomical context. By changing parameters such as scale, contrast thresholds, and enhancement strength according to local conditions, the method maintains high reliability across diverse imaging variations and anatomical presentations while remaining fully automated.

Inventive Principle:
Principle #35Parameter changes

3Adaptability or versatility

If multi-scale approach is used for vessel detection, then adaptability is improved, but device complexity increases

Engineering Contradiction:
Improvecapability to process various body parts and FOV CTA scansVSAvoidcomplexity of processing algorithm
Core Design Contradiction:
Adaptability or versatilityVSDevice complexity

Solution Approach 1:

The multi-scale approach is implemented through systematic segmentation of the detection process into discrete scale levels, each with specific processing steps. This structured segmentation makes the complex algorithm more manageable and implementable by breaking it into reusable modules that can be applied systematically across different body parts and scan types.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent develops a universal multi-scale framework that can process various body parts and field-of-view CTA scans using the same core algorithmic structure. By creating a multi-functional system that adapts to different anatomical regions and imaging conditions through parameter adjustment rather than separate algorithms, the method achieves high versatility while controlling complexity through code reuse and standardized processing pipelines.

Inventive Principle:
Principle #6Universality (Multi-functionality)

Data Source

PatentEP3497669B1Method for automatically detecting systemic arteries in arbitrary field-of-view computed tomography angiography (CTA).
Publication Date: 2021.02.17 AGFA HEALTHCARE NV
  • EP3497669B1 patent drawingFigure 1
  • EP3497669B1 patent drawingFigure 2
  • EP3497669B1 patent drawingFigure 3

AI summary

Method and system to fully-automatically analyze a medical image represented by a digital image representation. A method is disclosed to automatically detect systemic arteries in arbitrary field-of-view computed tomography angiography (CTA).