Spectral CT Ordered Subsets Reconstruction Using Expanded Data Subsets
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional ordered subsets (OS) methodology cannot be directly applied to spectral CT reconstruction due to the difference in the number of projection views between single-energy CT and spectral CT data, leading to inefficiencies in image reconstruction.
Innovation Solution
A method for spectral CT ordered subsets reconstruction is developed, which involves obtaining two sets of projection data with different numbers of views, partitioning the data into equal-sized subsets, including the second set in each subset, and performing OS reconstruction to generate a reconstructed CT image, while addressing issues like beam hardening and missing data through specific matrix calculations and updates.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If conventional ordered subsets methodology is applied to spectral CT reconstruction, then reconstruction speed can be improved, but the method cannot be directly applied due to different numbers of projection views between single-energy and spectral CT data
Solution Approach 1:
The patent segments the projection data by partitioning the first set of projection data (single-energy CT data) into N equal-sized subsets. Each subset is then combined with the second set of projection data (spectral CT data) to form expanded subsets. This segmentation allows the ordered subsets methodology to be applied despite the different numbers of projection views between the two data sets, enabling faster reconstruction while maintaining adaptability to spectral CT.
2Loss of information
If two datasets with different numbers of projection views are used in reconstruction, then spectral CT information can be incorporated, but the conventional OS scheme cannot be directly applied
Solution Approach 1:
The patent merges the first set of projection data (single-energy CT data with more views) and the second set of projection data (spectral CT data with fewer views) by including the second set in each of the N subsets of the first set. This combining approach retains all spectral CT information while creating a unified data structure that can be processed by the ordered subsets methodology, avoiding the need for separate reconstruction algorithms for each data type.
Solution Approach 2:
The patent changes the parameter structure of the projection data by expanding each subset to include both single-energy and spectral CT projection views. This parameter transformation allows the reconstruction algorithm to handle datasets with different numbers of projection views uniformly, reducing algorithmic complexity while preserving all spectral information.
3Loss of time
If ordered subsets reconstruction is performed with expanded subsets, then convergence speed is improved, but data partitioning and subset expansion are required
Solution Approach 1:
The patent performs preliminary actions by partitioning the first set of projection data into N equal-sized subsets and including the second set of projection data in each subset before performing the ordered subsets reconstruction. This pre-processing step organizes the data in a way that enables faster convergence during reconstruction, and the implementation complexity is manageable through systematic data organization rather than complex computational algorithms.
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 enables efficient reconstruction of CT images by leveraging the differences in projection views, improving the robustness of the OS method and achieving faster convergence, as demonstrated by comparisons with non-OS schemes.
Implementation Method 1
partitioning the first set of projection data into N equal-sized subsets of projection data; including the second set of projection data in each of the N subsets of projection data to generate N expanded subsets of projection data; performing OS reconstruction with the N expanded subsets of projection data
Implementation Method 2
an X-ray beam traversing an object and a detector relating the overall attenuation per ray
Implementation Method 3
The attenuation is derived from a comparison of the same ray with and without the presence of the object
Data Source
AI summary
A spectral computed tomography (CT) ordered subsets (OS) reconstruction method using two sets of projection data having a different number of views is provided. The method includes obtaining a first set of projection data that includes single-energy CT data for a first number of views, obtaining a second set of projection data that includes spectral CT data for a second number of views, wherein the first and second numbers are integers and the second number is smaller than the first number, partitioning the first set of projection data into N equal-sized subsets of projection data, including the second set of projection data in each of the N subsets of projection data to generate N expanded subsets of projection data, and performing OS reconstruction with the N expanded subsets of projection data to generate a reconstructed CT image.