Iterative Image Reconstruction Using Spatially Non-Homogeneous Element Selection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Iterative reconstruction algorithms for CT images face challenges in computational efficiency and resource demand, particularly in helical multi-slice CT systems, where conventional methods require extensive computation time and resources due to the need for iterative coordinate descent (ICD) algorithms, which can be slow and resource-intensive.
Innovation Solution
The method involves selecting a spatially non-homogeneous set of image elements for iterative optimization, focusing on regions furthest from convergence, and applying non-homogeneous iterative algorithms to reduce computation time and improve image quality, using techniques such as image element selection maps and varying update criteria to prioritize regions needing most updates.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional iterative reconstruction algorithms (such as ICD) are applied to all image elements, then image reconstruction quality is improved, but computation time and computational resources increase significantly
Solution Approach 1:
The patent applies different reconstruction strategies to different regions of the image based on their convergence characteristics. Regions that converge quickly use conventional algorithms, while regions that converge slowly receive additional iterative updates. This local differentiation maintains high image quality in critical areas while reducing overall computation time by avoiding uniform application of computationally intensive algorithms across the entire image.
Solution Approach 2:
The patent segments the image into different regions based on convergence behavior, identifying fast-converging and slow-converging areas. This segmentation allows the system to apply computational resources selectively - using full iterative reconstruction only where necessary (slow-converging regions) and simpler methods where sufficient (fast-converging regions), thereby resolving the contradiction between quality and computation time.
2Measurement precision
If iterative algorithms update every voxel sequentially, then convergence accuracy is improved, but computational resources and processing time increase
Solution Approach 1:
The patent applies partial action by updating only a subset of voxels in each iteration rather than all voxels sequentially. The system identifies and prioritizes updates for voxels in slow-converging regions, applying iterative updates selectively rather than uniformly. This approach maintains convergence accuracy in critical regions while improving overall reconstruction speed by reducing the number of sequential operations required.
Solution Approach 2:
The patent introduces dynamic adaptivity by adjusting the update strategy based on real-time convergence behavior. The system monitors convergence rates and dynamically modifies which voxels receive updates and how many iterations they require. This dynamic approach allows the reconstruction process to adapt to local characteristics, maintaining accuracy where needed while speeding up processing in regions that converge quickly.
Data Source
AI summary
A method for reconstructing an image of an object, the image comprising a plurality of image elements, is disclosed. The method includes accessing image data associated with the plurality of image elements, applying a first algorithm to the plurality of image elements, selecting a spatially non-homogenous set of the plurality of image elements, and applying an iterative algorithm to the set of image elements to reduce an amount of time necessary for reconstructing the image, or to improve an image quality at a fixed computation time, or both.