Cardiac Image Reconstruction Using Phase-Sorted Projection Subsets
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for reconstructing cross-sectional images of cyclically moving objects, such as the heart, using X-ray projections result in blurred images due to heart movement, and current iterative reconstruction algorithms are slow and prone to artifacts.
Innovation Solution
A data processing unit that sorts X-ray projections into subsets based on cardiac phase similarity, using electrocardiographic signals to weight projections and process them simultaneously in each iteration step, with an aperture function to reduce artifacts.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If all projections are used for tomographic reconstruction, then the reconstruction speed is fast, but the image quality deteriorates due to heart movement causing blurred images
Solution Approach 1:
The patent segments the projection data into multiple subsets, where each subset contains projections from a specific cardiac phase range. This segmentation allows the reconstruction algorithm to process projections corresponding to similar heart phases together, maintaining image sharpness while enabling faster iterative reconstruction through subset-based parallel processing
2Manufacturing precision
If iterative reconstruction algorithms are used to improve image quality, then the image quality improves, but the reconstruction speed deteriorates becoming slow
Solution Approach 1:
By dividing all projections into multiple subsets and processing each subset in separate iteration steps, the algorithm achieves faster convergence. Each iteration step processes a manageable subset of projections, reducing computational burden per step while maintaining the iterative refinement process needed for high image quality
Solution Approach 2:
The patent implements periodic action by cycling through different subsets in successive iteration steps. The algorithm systematically processes subset 1, then subset 2, and so on, repeating this periodic pattern until convergence. This periodic subset processing accelerates reconstruction while preserving image quality through multiple iterative refinements
3Productivity
If subsets of projections are used to improve reconstruction speed, then the reconstruction speed improves, but image artifacts increase
Solution Approach 1:
The patent applies local quality by assigning different weights to projections based on their cardiac phase similarity to the target phase. Projections closer to the target cardiac phase receive higher weights, while those farther away receive lower weights. This local weighting within each subset minimizes artifacts by emphasizing relevant projections while still utilizing the subset structure for speed
Data Source
AI summary
The invention relates to a method and a device for the iterative reconstruction of cross-sectional images of the heart (7) of a patient based on projections (P1, . . . P5) from different directions which are for example generated with a helical cone-beam CT scanner. A cardiac weight function (f) quantifies how near the projections (P1, . . . ) are to a given observation phase (To) of the heart cycle based on simultaneously recorded electrocardiographic signals (ECG). The whole set of projections (P1, . . . ) is divided into subsets (S1, . . . ) which each contain only projections corresponding to a similar cardiac weight (f), and an iterative reconstruction algorithm like ART uses in one update or iteration step all projections of such a subset (S1, . . . ) simultaneously.