CT Edge Reconstruction via Numerical Derivative Clipping
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current X-ray computed tomography (CT) image reconstruction methods struggle to effectively enhance edge detection and suppress non-edge regions, leading to artifacts and noise, particularly in applications where interface edges are the primary interest.
Innovation Solution
Applying image filters such as Prewitt, Zerocross, and Sobel filters to CT image data, followed by a numerical derivative to emphasize edges and suppress non-edge regions, and clipping derivative values to reduce artifacts and noise, resulting in improved edge reconstruction.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional image reconstruction methods are used, then the reconstruction process is simple, but edge detection is poor and non-edge regions are not suppressed effectively
Solution Approach 1:
The patent applies image filters (Prewitt, Zerocross, Sobel) and numerical derivatives to the sinogram data before reconstruction. This preliminary processing enhances edge information and suppresses non-edge regions in the frequency domain, allowing the reconstruction algorithm to focus computational resources on edge features rather than processing all regions equally.
Solution Approach 2:
The patent applies different processing treatments to different regions of the sinogram. By using numerical derivatives and filtering operations, the method selectively enhances regions containing edge information while suppressing regions without edges, creating locally optimized quality characteristics across the image data.
2Object-affected harmful factors
If conventional reconstruction methods are used, then computational resources are saved, but artifacts and noise are not reduced effectively
Solution Approach 1:
The patent transforms the sinogram data through numerical differentiation and filtering operations that change the frequency domain parameters of the image data. This parameter transformation converts noise and artifacts into distinguishable patterns that can be identified and suppressed during reconstruction, improving image quality without requiring excessive computational resources.
Solution Approach 2:
The method extracts and isolates edge information from the sinogram data using numerical derivatives and filtering operations. By separating edge features from non-edge regions before reconstruction, the method eliminates the need to process and reconstruct noise-prone regions, thereby reducing artifacts while maintaining computational efficiency.
3Measurement precision
If edge enhancement is applied, then edge detection improves, but post-processing requirements increase
Solution Approach 1:
The patent performs edge enhancement operations (filtering and numerical differentiation) on the sinogram data before reconstruction. This preliminary edge enhancement ensures that edge information is already optimized in the reconstructed image, eliminating the need for subsequent post-processing operations and reducing overall processing time.
Data Source
AI summary
A computer-implemented method of reconstructing boundaries from a computed tomography (CT) image dataset is provided. The method comprises receiving CT image data and applying a number of image filters to the CT image data. A numerical derivative is applied to the filtered CT image data to flatten and suppress non-edge regions that do not correspond to edges. Numerical derivative values are clipped to a specified range to reduce artifacts and noise, and an X-ray image is reconstructed according to the image filters and clipped numerical derivative values.