Vascular Centerline Extraction via Flow Fields
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing centerline tracing techniques for coronary arteries in medical images are prone to shortcuts and false positives due to imaging artifacts and severe pathologies, making accurate extraction challenging.
Innovation Solution
The method extracts centerlines by formulating the problem as finding maximum flow paths in computational flow fields, using estimated vessel orientation tensors as permeability for flow computations, and employing machine learning-based classification to distinguish true centerlines from leakage.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If minimal path extraction algorithms are used to extract coronary artery centerlines, then the extraction process is simple and fast, but the accuracy deteriorates due to shortcuts through non-coronary structures
Solution Approach 1:
The patent changes the fundamental parameter from minimal path extraction to maximum flow path extraction. By formulating the problem as a flow field computation with pressure gradients, the system extracts centerlines based on flow characteristics rather than simple path length, thereby improving accuracy while maintaining computational efficiency through established flow solvers.
Solution Approach 2:
The patent replaces the mechanical/minimal path extraction approach with a computational flow field approach. Instead of using graph-based shortest path algorithms, the system uses fluid dynamics principles to compute flow fields and extract centerlines as flow paths, substituting a different computational paradigm to achieve better accuracy.
2Quantity of substance
If vesselness masks with high sensitivity are computed to identify vascular structures, then the coverage of vascular structures is improved, but false positive voxels from background or nearby structures increase
Solution Approach 1:
The patent incorporates feedback through iterative flow field computation and centerline extraction. The system computes flow fields based on initial vesselness masks, extracts centerlines, and uses this information to refine the segmentation, allowing the system to learn from initial results and correct false positives through iterative improvement.
Solution Approach 2:
The patent segments the vascular structures from non-vascular structures by computing flow fields that naturally separate true vessels from false positives. By analyzing flow characteristics and pressure gradients, the system can distinguish between actual vascular structures and background or nearby structures, effectively segmenting the true positive vessels from false positives.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
This approach reduces shortcut issues and improves accuracy in extracting coronary artery centerlines, enhancing the robustness to false positives and occlusions, while maintaining computational efficiency.
Implementation Method 1
Embodiments of the present invention utilize an analogy of fluid flowing from regions of high pressure to those of low pressure by setting up a fluid flow problem where the coronary ostium is specified as a high pressure region
Data Source
AI summary
A method and apparatus for extracting centerline representations of vascular structures in medical images is disclosed. A vessel orientation tensor for each of a plurality of voxels associated with the target vessel, such as a coronary artery, in a medical image, such as a coronary tomography angiography (CTA) image, using a trained vessel orientation tensor classifier. A flow field is estimated for the plurality of voxels associated with the target vessel in the medical image based on the vessel orientation tensor estimated for each of the plurality of voxels. A centerline of the target vessel is extracted based on the estimated flow field for the plurality of vessels associated with the target vessel in the medical image by detecting a path that carries maximum flow.