Adaptive Scaling Factor for Cone Beam Artifact Reduction in Circular CT
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Cone beam artifacts in CT imaging, particularly in circular cone beam CT systems, remain a challenge due to missing data in the Radon domain, leading to image inaccuracies and increased radiation exposure from additional scans required for complete data collection.
Innovation Solution
The implementation of an adaptive scaling factor technique that optimizes the combination of circular and line data to reduce cone beam artifacts, using filtered backprojection and iterative reconstruction methods like OS-SART and TV minimization, along with forward projection and rebinning to generate corrected images with reduced artifacts.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If exact reconstruction is applied using a combination of circular and line trajectories, then cone beam artifacts are substantially removed, but the additional line scan increases patient radiation exposure and causes data inconsistency due to motion or agent enhancement changes between scans
Solution Approach 1:
The patent creates a virtual copy of the line trajectory data by estimating it from the circular trajectory data through mathematical transformations in the Radon domain. This estimated line data is then combined with the circular data to achieve exact reconstruction without requiring an actual additional physical scan, thereby avoiding additional radiation exposure while maintaining image accuracy
Solution Approach 2:
The patent introduces an intermediary estimation process that transforms circular trajectory data into estimated line trajectory data through Radon domain operations. This intermediary estimated data serves as a bridge that allows exact reconstruction algorithms to work with circular data alone, eliminating the need for separate line scans and their associated radiation exposure
2Object-affected harmful factors
If a scanogram is used to estimate line data, then patient radiation dose is not increased, but cone beam artifacts remain observable and motion or agent enhancement changes cause inaccuracy
Solution Approach 1:
The patent applies parameter changes by transforming the estimation approach from direct spatial domain methods (scanogram) to frequency domain methods (Radon domain transformations). This changes the mathematical parameters and operations used to estimate line data, resulting in more accurate estimates that better preserve image quality and reduce artifacts while maintaining the benefit of avoiding additional radiation exposure
3Reliability
If an additional line scan is performed to achieve theoretically complete trajectory, then exact reconstruction is possible, but the additional scan increases device complexity and exposes patient to additional radiation
Solution Approach 1:
The patent makes the circular trajectory data serve multiple functions: it is used both for direct image reconstruction and for estimating the line trajectory data needed for exact reconstruction. This multi-functionality allows the system to achieve complete reconstruction accuracy using only a single circular scan, eliminating the need for additional line scans and reducing device complexity
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 effectively minimizes cone beam artifacts in CT images, improving image quality and reducing radiation exposure by adaptively adjusting the scaling factor based on cone beam artifact metrics, resulting in more accurate and diagnostic-quality images from circular data alone.
Implementation Method 1
The X-ray source cone angle in most advanced CT systems such as Toshiba AquilionONE is quite large, and CFK images tend to suffer cone beam artifacts due to missing data in radon domain
Data Source
AI summary
Cone beam artifacts arise in circular CT reconstruction. The cone beam artifacts are substantially removed by reconstructing a reference image from measured data at circular source trajectory, generating synthetic data by forward projection of the reference image along a pre-determined source trajectory, which supplements the circular source trajectory to a theoretically complete trajectory, reconstructing a correction image from the synthetic data and applying a scaling factor whose value is adaptively determined and optimized based upon the minimization of a predetermined cone beam artifact metric. Ultimately, the cone beam artifact is substantially reduced by generating a corrected image using the reference image and the correction image that has been optimally scaled based upon the adaptively determined scaling factor value.