Adaptive Forward Projection Ray Spacing for CT Artifact Reduction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Iterative reconstruction methods using forward projection in computed tomography systems often lead to image artifacts, degrading the quality of reconstructed images due to fixed ray spacing and interpolation kernel widths.
Innovation Solution
A forward projection apparatus that varies the number of rays, ray spacing, and interpolation kernel width based on the effective image element spacing, using fictive rays and adaptive interpolation to reduce artifacts, particularly aliasing and varying kernel width issues.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If fixed ray spacing and fixed interpolation kernel width are used in forward projection, then the computational algorithm is simple, but image artifacts appear and image quality deteriorates
Solution Approach 1:
The patent applies dynamics by making the ray spacing and interpolation kernel width variable rather than fixed. The ray spacing is adjusted dynamically based on the distance from the radiation source, and the kernel width is adapted according to the effective image element spacing at different positions. This dynamic adjustment resolves the contradiction by allowing the algorithm to maintain simplicity in structure while achieving high image quality through adaptive parameter variation.
Solution Approach 2:
The patent implements parameter changes by varying the ray spacing and interpolation kernel width according to specific conditions. The ray spacing is changed as a function of distance from the radiation source, and the kernel width is modified based on the effective image element spacing. These parameter changes enable the system to reduce image artifacts while maintaining computational efficiency.
2Productivity
If uniform ray distribution is used, then the forward projection is computationally efficient, but aliasing artifacts occur in regions with varying effective image element spacing
Solution Approach 1:
The patent applies local quality by making the ray spacing non-uniform, adapted to local conditions of effective image element spacing. In regions where the effective image element spacing is smaller, the ray spacing is reduced accordingly, while in regions with larger spacing, the ray spacing is increased. This local adaptation eliminates aliasing artifacts while maintaining overall computational efficiency by avoiding unnecessary dense sampling in all regions.
3Device complexity
If fixed interpolation kernel width is used, then the interpolation process is simple, but varying effective kernel widths cause artifacts in the projection data
Solution Approach 1:
The patent implements parameter changes by making the interpolation kernel width variable. The kernel width is adjusted according to the effective image element spacing at different positions in the image. This ensures that the effective kernel width remains consistent across different rays, preventing artifacts in the projection data while keeping the interpolation process relatively simple through a systematic parameter adaptation rule.
Data Source
AI summary
The invention relates to a forward projection apparatus for performing a forward projection through an image (22), wherein at least one of a number of rays (20, 21) for performing the forward projection, a ray spacing between the rays and a kernel width of an interpolation kernel for calculating interpolated values located on the rays is varied depending on the ray width relative to an effective image element spacing between image elements (24) of the image. This allows reducing artifacts in simulated projection data and, thus, in an image, which is iteratively reconstructed by using the simulated projection data. For example, if the number of provided rays and/or the ray spacing between the provided rays is varied, aliasing artifacts can be reduced. Moreover, if the ray spacing between the provided rays and/or the kernel width of the interpolation kernel is varied, artifacts caused by varying effective kernel widths may be reduced.