Vascular Flow Segmentation for AVM DSA Dataset Clarity
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for acquiring 3D or 4D digital subtraction angiography (DSA) datasets of arteriovenous malformations (AVMs) are inadequate due to incomplete opacification of the nidus, leading to increased X-ray dose and complexity in treatment planning, especially when multiple blood vessels supply the AVMs.
Innovation Solution
A computer-implemented method for acquiring and classifying sub-portions of branched hollow organs, such as vascular trees, based on contrast medium flow direction and confluences, to create partial datasets with dedicated representations, reducing the need for multiple recordings and simplifying analysis.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If multiple 3D-DSAs are performed to achieve complete opacification of the nidus, then the completeness of the dataset improves, but the X-ray dose to the patient increases
Solution Approach 1:
The patent segments the vascular tree into multiple sub-portions based on confluences and flow directions. By classifying sub-portions as single-fed or multi-fed, the system can selectively process and evaluate each segment independently, achieving complete coverage of the nidus through a single 3D-DSA while maintaining data completeness without requiring multiple recordings.
2Reliability
If multiple 3D-DSAs are performed to achieve complete opacification of the nidus, then the completeness of the dataset improves, but the workflow complexity increases
Solution Approach 1:
The patent performs preliminary classification of sub-portions into single-fed and multi-fed categories before final evaluation. This preliminary segmentation and classification enables automated processing and simplifies the workflow by organizing the complex vascular structure into manageable segments that can be systematically evaluated, reducing the need for manual assembly of multiple datasets.
3Reliability
If contrast medium is injected in ascending thoracic aorta or venous injection to achieve complete opacification, then the completeness of the dataset improves, but the interpretability of the dataset deteriorates due to mixing of contrast medium in different afferent vessels
Solution Approach 1:
The patent applies local quality analysis by evaluating contrast medium flow characteristics specifically in each classified sub-portion independently. By determining flow directions and identifying confluences at the local level of each sub-portion, the system can interpret the contrast medium distribution in a single-fed sub-portion without interference from mixing in other vessels, thereby maintaining high interpretability while achieving complete opacification.
Data Source
AI summary
A computer-implemented method for providing a results dataset includes: acquiring an image dataset of an examination object; identifying portions of a hollow organ in the image dataset based on a mapped contrast medium flow; classifying sub-portions of the identified portions into single-fed and multi-fed sub-portions based on a mapped flow direction of the contrast medium flow and an identification of confluences of the identified portions of the hollow organ mapped in the image dataset, wherein sub-portions of the hollow organ arranged downstream relative to a confluence are classified as multi-fed sub-portions; and providing the results dataset based on the image dataset and the classified sub-portions of the hollow organ, wherein the results dataset has a partial dataset for each classified sub-portion of the hollow organ, and wherein, the partial datasets have a dedicated representation of the contrast medium flow in each classified sub-portion of the hollow organ.