X-ray CT Image Reconstruction via Adaptive Weighting Coefficient Map
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current image reconstruction methods in X-ray computed tomography struggle to produce high-resolution images, especially when processing mixed pixels at object boundaries, as they apply uniform substance information constraints without distinction, leading to inappropriate effects such as loss of fine structures and unnatural boundaries.
Innovation Solution
Generating a weighting coefficient map based on the reconstruction image to control the prior knowledge applied to each pixel, allowing for differential weighting of substance information, thereby avoiding excess constraint and retaining fine structures while reducing artifacts.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If uniform substance information constraints are applied to all pixels in image reconstruction, then artifacts are reduced, but fine structures and boundaries are lost due to excess constraint
Solution Approach 1:
The patent applies different weighting coefficients to different spatial regions of the image. Specifically, pixels near object boundaries receive lower weighting coefficients for substance information constraints, while pixels in homogeneous regions receive higher weighting coefficients. This local differentiation allows artifact reduction in homogeneous areas while preserving fine structures and boundaries where excessive constraint would be harmful.
2Measurement precision
If iterative approximation method with prior knowledge is used, then reconstruction accuracy is improved, but processing time increases
Solution Approach 1:
The patent applies substance information constraints partially rather than uniformly across all pixels. By introducing a weighting coefficient that modulates the strength of the prior knowledge term in the objective function, the method applies constraints only where beneficial (in homogeneous regions) while omitting them where harmful (at boundaries). This partial application reduces the overall computational burden and processing time while maintaining reconstruction accuracy in critical regions.
3Measurement precision
If high weighting coefficient for prior knowledge is applied, then substance information accuracy is improved, but image resolution deteriorates due to loss of fine structures
Solution Approach 1:
The patent introduces a spatially varying weighting coefficient that adjusts the strength of prior knowledge application locally. In regions where substance information accuracy is critical (homogeneous areas), higher weighting coefficients are applied. In regions where image resolution and fine structure preservation are critical (boundaries and edges), lower weighting coefficients are applied. This local quality differentiation resolves the contradiction by allowing high substance information accuracy where applicable without sacrificing overall image resolution.
Data Source
Figure 1
Figure 2
Figure 3(a)~4(f)
AI summary
This disclosure provides a reconstruction processing method. The method includes an image updating step (Step S2) of updating a reconstruction image by an iterative approximation method, and a weighting coefficient map updating step (Step S4). In the weighting coefficient map updating step, a weighting coefficient map relative to prior knowledge is generated from the reconstruction image obtained by updating an image in the image updating step (Step S2), and a weighting coefficient of the prior knowledge relative to each pixel is controlled in accordance with the weighting coefficient map, whereby a weighting coefficient map is updated. As described above, the weighting coefficient map relative to the prior knowledge is generated from the reconstruction image (during estimation) obtained by updating the image, and the weighting coefficient of the prior knowledge relative to each pixel is controlled in accordance with the weighting coefficient map. Consequently, overcoming of such a drawback is performable that the reconstruction image with high resolution is not obtainable. In other words, controlling the weighting coefficient of the prior knowledge relative to each pixel achieves avoidance of excess constraint to the pixel of the reconstruction image. Accordingly, the reconstruction image of high resolution is obtainable.