CT Image Reconstruction Using Position-Dependent Ray Weighting
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current computed tomography methods using filtered backprojection with cone beam artifacts and redundant data lead to inadequate image quality and inefficient radiation use, especially with 3D methods and large detector arrays, resulting in image artefacts and incomplete utilization of radiation dose.
Innovation Solution
A method that accounts for redundancies in voxel data during backprojection and weights rays based on their position within the beam, particularly favoring centrally positioned rays over marginal ones, and converts initial data from fan beam geometry to parallel beam geometry for improved image quality through rebinning and tailored filtering.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If 3D filtered backprojection method is used with cone beam geometry, then image reconstruction is achieved, but cone beam artefacts occur and image quality deteriorates
Solution Approach 1:
The patent applies local quality by differentiating the treatment of rays based on their position within the beam. Central rays (which pass through the center of the detector array) are weighted differently from peripheral rays (which pass through the edges). This localized differentiation allows the reconstruction algorithm to compensate for the geometric distortions that cause cone beam artifacts, thereby improving image quality without requiring a complete change to the 3D reconstruction approach.
Solution Approach 2:
The patent changes the weighting parameter in the backprojection formula based on ray position. By introducing a position-dependent weight function that varies from central to peripheral rays, the algorithm compensates for the non-uniform sampling density inherent in cone beam geometry. This parameter modification effectively reduces cone beam artifacts while maintaining the efficiency of 3D reconstruction.
2Quantity of substance
If spiral scanning with low table advance is used, then redundant data is produced through multiple irradiation of the same voxel, but radiation dose is used incompletely for imaging
Solution Approach 1:
The patent merges multiple measurements of the same voxel obtained during spiral scanning into a single reconstructed value. By combining the redundant data from multiple irradiations through the weighted backprojection process, the algorithm fully utilizes the radiation dose that would otherwise be wasted. This merging process converts redundant measurements into useful information, improving radiation dose efficiency while maintaining image quality.
3Manufacturing precision
If 2D filtered backprojection method is used with large detector arrays, then preliminary images can be reformatted, but an extremely large number of preliminary images must be processed
Solution Approach 1:
The patent transitions from 2D to 3D reconstruction by incorporating the third dimension (detector array width) directly into the backprojection process. Instead of generating multiple 2D preliminary images and reformatting them, the method performs a single 3D backprojection that directly produces the final volumetric data. This dimensional change eliminates the intermediate processing step and dramatically reduces computational load while maintaining image quality.
4Manufacturing precision
If central rays are weighted more heavily than peripheral rays during backprojection, then image quality improves, but processing complexity increases
Solution Approach 1:
The patent implements local quality by applying position-dependent weighting to rays during backprojection. Central rays, which provide more accurate information about the object's center, are given higher weights, while peripheral rays, which are more prone to geometric distortion, are given lower weights. This localized quality adjustment improves image quality by emphasizing reliable measurements while downweighting unreliable ones, without requiring complex iterative optimization.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
This approach enhances image quality by effectively utilizing the radiation dose and reducing artefacts, achieving high-quality 3D images by weighting rays according to their position and filtering direction, thereby improving the overall image reconstruction process.
Implementation Method 1
a detector array including a plurality of distributed detector elements is adapted to detect rays of the beam and is adapted to supply initial data representing an attenuation of the rays passing through the object under examination
Data Source
AI summary
A method is for image reconstruction for computed tomography with a non-one-dimensional, extended detector. The rays of the detector are weighted during the backprojection as a function of their position in the beam.