CT Center Shift Estimation Avoiding Metal Artifacts
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for estimating center shift amount in CT devices are inaccurate when dealing with samples containing metal, as extreme pixel value variations at metal boundaries can disrupt total variation calculations.
Innovation Solution
A center shift amount estimating apparatus and method that specifies a region of interest avoiding areas with extreme pixel value differences, calculates total variation indices in temporarily corrected images, and identifies the center shift amount corresponding to extreme index values during 360° or 180° scans.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If total variation is used as an index for estimating center shift amount, then the center shift amount can be estimated, but the estimation becomes inaccurate when metal is present in the sample due to large pixel value variations at metal boundaries
Solution Approach 1:
The patent divides the image into multiple regions and calculates the total variation index for each region separately. By segmenting the image, the method isolates metal boundary regions with extreme pixel value variations from other regions, allowing the index calculation to be performed on regions without such disturbances, thereby improving center shift estimation accuracy
Solution Approach 2:
The patent applies different processing strategies to different regions of the image. Regions with extreme pixel value variations (such as metal boundaries) are identified and excluded from the index calculation, while regions with normal pixel value distributions are used for calculation. This local quality approach ensures that the total variation index reflects true center shift information rather than being dominated by metal artifact variations
2Measurement precision
If a region of interest is specified to avoid areas with extreme pixel value differences, then the center shift amount can be accurately specified, but the device complexity increases due to additional processing steps
Solution Approach 1:
The patent performs preliminary identification of regions with extreme pixel value variations before calculating the total variation index. By pre-identifying and marking these problematic regions, the method prepares the image data in advance, allowing the subsequent index calculation to proceed efficiently on validated regions without requiring complex real-time processing or iterative optimization algorithms
Data Source
Figure 1
Figure 2A~2B
Figure 3A~3B
AI summary
A center shift amount estimating apparatus 300 for estimating a deviation between a rotation axis of the sample and a center of a detector with respect to an X-ray source in a CT device 200, comprises a region specifying unit 330 for specifying a region of interest in a reconstructed uncorrected image to avoid an area where there is an extremely different pixel value from the surrounding area, a temporary correction unit 340 for correcting an assumed center shift amount to reconstruct a temporarily corrected image with respect to the region of interest, an index analyzing unit 360 for searching an extreme value of an index representing variation of pixel values in the temporarily corrected image, and a center shift amount specifying unit 380 for specifying an actual center shift amount with respect to the extreme value.