Modified Ordered Subsets Attenuation Weighted Reconstruction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current medical imaging techniques, particularly iterative reconstruction methods in emission tomography, face challenges in achieving high-resolution images of targeted fields of view (FOV) within the scan FOV, leading to artifacts and inaccuracies due to unaccounted signals from outside the targeted FOV, and conventional subset selection methods result in computational inefficiencies and artifacts such as block structure and noise ringing.
Innovation Solution
A modified subset method groups projections with similar attenuation path lengths to form 'pie' shape subsets, providing limited angle information and reducing artifacts by selecting subset numbers not based on block periodicity, thereby improving image reconstruction accuracy and resolution.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional iterative reconstruction techniques are used for targeted FOV less than scan FOV, then reconstruction is generally straightforward, but artifacts and quantitative inaccuracies occur at the periphery of the reconstructed image
Solution Approach 1:
The patent divides the projection data into multiple ordered subsets, where each subset corresponds to a specific angular range. This segmentation allows the reconstruction algorithm to process different angular information separately and combine them systematically, reducing artifacts at the periphery of the targeted FOV while maintaining quantitative accuracy.
2Productivity
If conventional ordered subsets method is used, then computational burden is reduced, but block structure artifacts and noise ringing occur
Solution Approach 1:
The patent applies different processing strategies to different subsets based on their local characteristics. By ordering subsets according to their angular ranges and applying attenuation correction specific to each subset's geometry, the method maintains computational efficiency while reducing block structure artifacts and noise ringing that occur with conventional uniform subset processing.
3Measurement precision
If full scan FOV is reconstructed at high resolution, then targeted FOV image quality is improved, but computational resources and time increase
Solution Approach 1:
The patent applies ordered subsets methodology to process only the necessary projection data for the targeted FOV at high resolution, rather than processing the entire scan FOV. This partial action approach maintains high resolution for the region of interest while significantly reducing computational resources and reconstruction time compared to reconstructing the full scan FOV.
Data Source
AI summary
A method and system is provided for performing medical imaging. The method and system includes at least one radiation detector to detect radiation from a subject, and an image processor which determines attenuation paths for an image point, groups substantially similar attenuation path lengths for the same image point to form a modified subset group, and processing image data using the modified subset group in order to provide a reconstructed image substantially similar to an original image.