Spatially Varying 3D CT Artifact Removal by Iterative Thresholding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods struggle to effectively remove spatially varying artifacts such as laminographic and high-angle cone beam artifacts in 3D computed tomography, particularly in setups that violate Tuy's or Orlov's conditions, leading to inaccurate and obscured reconstructions.
Innovation Solution
A method involving thresholding current reconstructions to create thresholded reconstructions, simulating and subtracting these from the original to iteratively reduce artifacts, combined with optional neural network training to enhance accuracy and speed.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If conventional reconstruction algorithms (FBP, FDK) are used, then reconstruction speed and simplicity are maintained, but spatially varying artifacts appear in the reconstructed images
Solution Approach 1:
The method performs preliminary thresholding on the initial reconstruction to identify artifact regions before final reconstruction. By pre-identifying artifact locations and characteristics, the algorithm can针对性地 remove artifacts during the reconstruction process, resolving the contradiction between speed and accuracy.
Solution Approach 2:
The patent introduces an intermediate artifact removal step between the initial FBP/FDK reconstruction and the final reconstruction. This intermediary process uses thresholded reconstructions to create artifact-free reference data, which then guides the final high-accuracy reconstruction, effectively mediating between speed and precision requirements.
2Manufacturing precision
If specialized artifact removal algorithms are used, then reconstruction accuracy improves, but computational complexity and processing time increase
Solution Approach 1:
The artifact removal process is segmented into distinct steps: initial reconstruction, thresholding to identify artifacts, creating thresholded reconstructions, and final artifact-free reconstruction. This segmentation allows each step to be optimized independently, reducing overall complexity while maintaining accuracy.
Solution Approach 2:
The method applies thresholding at multiple stages and uses iterative refinement where necessary. By applying partial artifact removal in the initial stage and then refining with more comprehensive methods, the algorithm achieves high accuracy without requiring the full complexity of advanced algorithms from the start.
3Productivity
If limited-angle laminographic geometry is used, then signal-to-noise ratio increases and throughput improves, but spatially varying artifacts are introduced
Solution Approach 1:
The patent converts the harmful spatially varying artifacts into useful information by using thresholding to identify them. The artifacts, which result from limited-angle geometry, are systematically identified and removed through the thresholded reconstruction process, transforming a disadvantage into a solvable problem that improves final image quality.
Solution Approach 2:
The method changes the parameter space by working in the thresholded reconstruction domain rather than directly in the projection data domain. By transforming the problem into identifying and removing artifacts in the reconstructed space, the algorithm can effectively handle the limited-angle geometry artifacts while maintaining the throughput benefits of the laminographic setup.
Data Source
Figure 1
Figure 2A~2E
Figure 3A
AI summary
A method for removing spatially varying artifacts such laminographic artifacts and/or high-angle cone beam artifacts for 3D computed tomography (CT) involves thresholding current reconstructions to create thresholded reconstructions and then creating simulated reconstructions from the thresholded reconstructions. These simulated reconstructions are subtracted from the current reconstructions to create the current reconstructions for a next iteration. A final reconstruction is then created by summing the thresholded reconstructions. This approach can progressively remove the artifacts. In addition, the method can be used to generate high quality training data to further improve the speed and robustness. These methods will work for other non-Orlov complete computed tomography in general, such as high cone angle, missing views.