Parallel CT Image Reconstruction Using Block-Separable Surrogates
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing CT imaging technologies face long computation times and inefficiencies due to data transfer bottlenecks in parallel processing methods, particularly when using model-based image reconstruction (MBIR) techniques for large helical scans.
Innovation Solution
The implementation of a distributed processing system with a central processor and multiple distributed processors that perform parallel processing by alternating between updating images using sinogram and image domain information separately, allowing concurrent computation and data transfer, and utilizing block-separable surrogates to reduce communication overhead.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If model based image reconstruction (MBIR) or iterative approaches are used to reduce noise and artifacts in CT imaging, then image quality is improved, but computation time increases excessively
Solution Approach 1:
The patent divides the image reconstruction task into multiple independent blocks that can be processed in parallel. The cost function is segmented into block-separable components, allowing distributed processors to work simultaneously on different blocks without requiring continuous data transfer, thus reducing computation time while maintaining image quality.
Solution Approach 2:
The patent implements an iterative reconstruction process where useful computation continues throughout the entire process. By using block-separable surrogates and alternating updates between sinogram and image domain information, the system maintains continuous productive computation without idle waiting periods, significantly reducing total computation time.
2Productivity
If parallel processing is implemented to reduce computation time, then productivity is improved, but data transfer bottlenecks cause processors to remain idle
Solution Approach 1:
The patent segments the reconstruction problem into independent block-separable components that can be processed in parallel by distributed processors. Each processor works on its assigned block independently, eliminating the need for frequent data transfers and synchronization, thus reducing idle time while maintaining high productivity.
Solution Approach 2:
The patent performs preliminary organization of data into block-separable surrogates before distributed processing begins. This preliminary structuring allows processors to independently minimize their respective surrogate components without requiring continuous communication during the iterative process, minimizing idle time caused by data transfer bottlenecks.
3Loss of time
If distributed processors are used to perform parallel processing, then computation time is reduced, but communication overhead increases
Solution Approach 1:
The patent divides the optimization problem into block-separable surrogates that can be minimized independently by distributed processors. This segmentation reduces communication overhead because each processor only needs to minimize its local surrogate component, requiring minimal coordination and data exchange compared to traditional distributed optimization approaches.
Solution Approach 2:
The patent introduces block-separable surrogates as intermediary functions that bridge the original cost function and the distributed optimization process. These surrogates allow processors to work independently with simplified local objectives, reducing the complexity of inter-processor communication while still achieving the global optimization goal.
Data Source
AI summary
A method for iteratively reconstructing an image is provided. The method includes acquiring, with a detector, computed tomography (CT) imaging information. The method also includes generating, with at least one processor, sinogram information from the CT imaging information. Further, the method includes generating, with the at least one processor, image domain information from the CT imaging information. Also, the method includes updating the image using the sinogram information. The method further includes updating the image using the image domain information. Updating the image using the sinogram information and updating the image using the image domain information are performed separately and alternately in an iterative fashion.