CT Image Reconstruction Using Weighting Windows and Regularization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Computed tomography (CT) image reconstruction techniques face challenges in reducing or removing motion artifacts and noise, especially when objects move during scanning, which affects the quality of iterative reconstruction images.
Innovation Solution
A method and system for image reconstruction that utilize a computing device with a processor and storage medium to determine a weighting window based on detector array parameters, and an objective function including a model representing the difference between target and image data, with regularization items for denoising, to reconstruct images by performing iterations based on this function.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If iterative reconstruction techniques are used in CT imaging, then image reconstruction quality is improved, but motion artifacts and noise are introduced when objects move during scanning
Solution Approach 1:
The patent applies preliminary action by determining a weighting window and regularization parameters before performing iterative reconstruction. The weighting window is calculated based on detector array parameters and imaging conditions in advance, and the regularization item is prepared with predefined models. This preliminary preparation allows the reconstruction algorithm to account for potential motion artifacts and noise before they fully degrade the image quality, thereby improving reconstruction quality while mitigating harmful effects.
Solution Approach 2:
The patent utilizes parameter changes by dynamically adjusting the weighting window parameters and regularization strength based on the specific imaging conditions and detector array configuration. The weighting window parameters are modified according to the motion characteristics and noise levels detected during scanning. This adaptive parameter adjustment allows the reconstruction algorithm to optimize its performance for different motion scenarios, reducing motion artifacts while maintaining image quality.
2Manufacturing precision
If a weighting window and regularization models are applied to reduce motion artifacts and noise, then image quality is improved, but computational complexity increases
Solution Approach 1:
The patent applies local quality by implementing a weighting window that applies different weights to different regions of the detector array data. Rather than uniformly processing all data, the weighting window selectively emphasizes or de-emphasizes specific regions based on their reliability and contribution to the final image. This localized approach reduces computational complexity by focusing processing resources on the most informative regions while still achieving improved image quality through targeted regularization.
3Object-affected harmful factors
If multiple iterations are performed based on the objective function, then motion artifacts and noise are reduced, but reconstruction time increases
Solution Approach 1:
The patent applies partial action by performing a limited number of iterative reconstruction steps rather than exhaustive iterations. The regularization item with predefined models provides sufficient noise and artifact reduction within a practical number of iterations, avoiding the diminishing returns of excessive computation. This approach achieves adequate suppression of motion artifacts and noise while maintaining clinically acceptable reconstruction times.
Data Source
AI summary
The present disclosure relates to methods, systems, and non-transitory computer readable mediums for reconstructing an image. Image data may be obtained, wherein the image data may be generated by a detector array. A weighting window may be determined based on at least one parameter relating to the detector array. A first set of data may be determined based on the image data and the weighting window. An objective function associated with a target image may be determined based on the first set of data, wherein the objective function may include a first model, the first model may represent a difference between the target image and the first set of data, and the first model may be identified based on the first set of data. The target image may be reconstructed by performing a plurality of iterations based on the objective function.