CT Image Artifact Reduction via Virtual Scan Path Reprojection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional cone-beam CT systems suffer from incomplete data acquisition due to divergent x-ray beams, leading to artifacts in reconstructed images that degrade diagnostic quality, especially when large cone-angles are used in imaging processes.
Innovation Solution
A method and system that reconstructs images using actual CT data with missing information by reprojecting the initial image onto a virtual scan path to acquire virtual data, reducing artifacts through iterative total variation steepest descent optimization and exact image reconstruction algorithms.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If a single circle/arc scanning path is used in cone-beam CT systems, then the scanning path is simplified and easier to implement, but the data acquisition becomes incomplete causing missing data and artifacts in reconstructed images
Solution Approach 1:
The patent transforms the incomplete single-circle scanning data into a complete dataset by adding a temporal dimension through multiple iterations. The iterative reconstruction process generates virtual projections at different time points, effectively converting a 2D scanning problem into a 4D problem (3D space + time) to fill missing data gaps.
Solution Approach 2:
The patent performs preliminary reconstruction to generate an initial image, then uses this initial image to generate virtual projections that fill missing data before performing the final exact reconstruction. This preliminary action of creating virtual data allows the exact reconstruction algorithm to work with complete information.
2Measurement precision
If exact image reconstruction algorithms are used with incomplete data, then the reconstruction accuracy is maximized, but the artifacts from missing data significantly degrade image quality
Solution Approach 1:
The patent introduces an intermediary step between data acquisition and final reconstruction: generating virtual projections from an initial reconstructed image. These virtual projections act as a mediator that fills missing data gaps, allowing the exact reconstruction algorithm to produce artifact-free images.
Solution Approach 2:
The patent implements a feedback loop where the initial reconstructed image is used to generate virtual projections, which are then fed back into the reconstruction process. This feedback mechanism allows the system to iteratively improve the completeness of the data before final reconstruction.
3Object-affected harmful factors
If approximate image reconstruction methods are used to compensate for missing data, then artifacts are reduced, but the methods are only sufficient for very small cone-angles and fail for large cone-angles
Solution Approach 1:
The patent changes the parameter of data completeness by generating virtual projections that fill missing data, rather than changing the reconstruction algorithm itself. This allows the use of exact reconstruction methods across all cone-angles by ensuring complete input data through parameter expansion (adding virtual views).
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
The method effectively reduces artifacts in reconstructed images, improving diagnostic quality by filling in missing data and providing a more accurate representation of the object, regardless of cone-angle or data acquisition process.
Implementation Method 1
The intensity of the transmitted radiation is dependent upon the attenuation of the x-ray beam by the object and each detector produces a separate electrical signal that is a measurement of the beam attenuation
Data Source
AI summary
A system and method for generating an image of an object that is substantially free of artifacts induced by missing data is disclosed. The method includes performing a computed tomography (CT) imaging process using a cone-beam traversed over an actual scan path to acquire actual CT data having missing data. The method also includes reconstructing an initial image of a volume of interest (VOI) using the actual CT data, with the initial reconstructed image having artifacts attributable to the missing data, and reprojecting the reconstructed image of the VOI onto a virtual scan path to at least acquire virtual data corresponding to the missing data. Further, the method includes reconstructing an improved image of the VOI using the actual CT data and the virtual data, with the artifacts reduced in the improved reconstructed image.