Deformable Registration for 5D Cone Beam CT Motion
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for obtaining volumetric images of a target region moving due to breathing and cardiac motion are inadequate in accurately sorting and reconstructing images that account for both physiological cycles, leading to incomplete or inaccurate representations of the target region's motion patterns.
Innovation Solution
An apparatus and method that utilize registration engines to determine breathing and cardiac correlated images, generating volumetric images by combining these registrations, and a breathing-cardiac image generator to produce images for combined breathing and cardiac stages, allowing for the reconstruction of accurate volumetric images that account for both motion patterns.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If projection images are sorted into phase bins based on breathing or cardiac cycles separately, then the sorting process is simple, but the accuracy of representing combined breathing-cardiac motion is insufficient
Solution Approach 1:
The patent segments the sorting process into two independent stages: first sorting projection images into breathing phase bins, then sorting those into cardiac phase bins. This hierarchical segmentation allows the system to maintain simple sorting operations at each stage while achieving accurate combined motion representation through the combination of both sorting dimensions.
Solution Approach 2:
The patent adds a second sorting dimension by implementing nested phase binning - first by breathing phase, then by cardiac phase within each breathing bin. This dimensional approach transforms a single complex sorting problem into a two-dimensional sorting structure, improving motion representation accuracy while keeping each sorting operation manageable.
2Ease of manufacture
If volumetric images are reconstructed from sorted phase bins, then the reconstruction process is straightforward, but the accuracy of combined breathing-cardiac motion representation is insufficient
Solution Approach 1:
The reconstruction process is segmented into separate breathing and cardiac components. Volumetric images are first reconstructed for each breathing phase bin, then further sorted and reconstructed by cardiac phase within each breathing bin. This segmentation allows straightforward reconstruction operations at each level while achieving accurate combined motion representation through the hierarchical structure.
Solution Approach 2:
The patent implements two-dimensional phase binning where projection images are sorted into breathing phase bins first, then into cardiac phase bins within each breathing bin. This creates a nested structure that enables straightforward reconstruction at each dimension while accurately representing the combined four-dimensional motion (three spatial dimensions plus time with both breathing and cardiac cycles).
Data Source
AI summary
An apparatus includes: an input configured to receive a plurality of images, wherein the images include respective sub-images of a bodily part of a subject, and wherein a position of the bodily part relates to a breathing movement and a cardiac movement of the subject; a first registration engine configured to determine a first registration of at least two breathing correlated images, wherein the at least two breathing correlated images comprise two of the plurality of images or are derived from at least some of the plurality of images; a second registration engine configured to determine a second registration of at least two cardiac correlated images; and a volumetric image generator configured to generate a volumetric image using the first registration and the second registration.


