Four-Dimensional Angiography Vessel Overlap Artifact Reduction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Four-dimensional digital subtraction angiography faces challenges in reducing vessel overlap artifacts, which affect image quality and clinical significance due to unclear contrast information assignment in overlapping vessels during multiplicative back projection.
Innovation Solution
A method that performs a plausibility check to verify the connection of vessel sections filled with contrast medium to admissible source points, using physically motivated criteria to reduce overlap artifacts by ensuring only physically plausible fillings are displayed, and employing graph theory algorithms to identify connected components and correct partial image data sets.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If multiplicative back projection is used to reconstruct four-dimensional angiography data sets, then temporal resolution and image quality are improved, but vessel overlap artifacts increase due to non-unique contrast information assignment
Solution Approach 1:
The patent segments the vascular tree into distinct vascular segments based on anatomical boundaries. By dividing the continuous vascular system into discrete segments, the algorithm can uniquely assign contrast information to specific segments, resolving the ambiguity in overlapping vessels. This segmentation enables precise temporal resolution while eliminating vessel overlap artifacts through clear segment identification.
Solution Approach 2:
The patent introduces an intermediary confidence map that mediates between the projection images and the three-dimensional vessel data set. This confidence map assigns reliability values to different vascular segments based on the clarity of contrast information in projection images. The intermediary structure allows the system to distinguish between reliable contrast information and ambiguous overlapping regions, enabling accurate temporal reconstruction without artifacts.
2Object-generated harmful factors
If confidence values are used to identify and remove unreliable vessel sections, then overlap artifacts are reduced, but risk of incorrect removal of legitimate contrast fillings increases
Solution Approach 1:
The patent applies local quality by assigning different confidence values to different vascular segments based on their specific anatomical locations and the quality of contrast information in corresponding projection images. Each segment is evaluated independently with its own confidence score, allowing selective removal of unreliable segments while preserving legitimate contrast fillings in high-confidence regions. This localized approach maintains detection accuracy while reducing artifacts.
Solution Approach 2:
The patent performs preliminary action by pre-calculating confidence values for all vascular segments before the main reconstruction process. These pre-computed confidence maps guide subsequent processing steps, enabling the system to identify and handle unreliable segments in advance. This preliminary evaluation ensures that only segments with sufficient confidence are processed, preventing incorrect removal of legitimate contrast fillings while still eliminating overlap artifacts.
Data Source
AI summary
Vessel overlap artifacts are reduced in a four-dimensional angiography data set a blood vessel system of a patient with a contrast medium. A three-dimensional vessel data set of the blood vessel system is reconstructed from two-dimensional projection images of digital subtraction angiography showing the blood vessel system, determined by multiplicative back projection of the projection images into the vessel data set or a base data set of the four-dimensional angiography data set derived vessel data set. A plausibility check is performed with vessel sections displayed as filled with contrast medium in partial image data sets of the angiography data set assigned to all individual, and different instants of the covered period are checked against a plausibility check criterion checking for a contrast medium-filled connection to an admissible source point, after which a corrected partial image data set is determined containing only vessel sections satisfying the plausibility check criterion.