Non-uniform Discretization for CT Scatter Simulation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for simulating X-ray scatter in computed tomography are time-consuming and often sacrifice accuracy, leading to inefficient scatter correction and poor image quality due to artifacts from scattered X-rays.
Innovation Solution
The method employs non-uniform discretization of the radiative transfer equation (RTE) for accelerated scatter estimation, using spherical harmonics and dual-coordinate systems to simulate scatter flux, and applies different resolutions based on energy and spatial variations to improve computational efficiency without compromising accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If uniform discretization of the radiative transfer equation is used, then accuracy of scatter simulation is maintained, but computational time increases significantly
Solution Approach 1:
The patent applies non-uniform discretization where different spatial regions are discretized at different resolutions. Regions with high scatter contribution are discretized more finely to maintain accuracy, while regions with low scatter contribution use coarser discretization to reduce computational load. This resolves the contradiction by making the discretization quality local rather than uniform throughout the entire domain.
Solution Approach 2:
The computational domain is segmented into multiple regions based on scatter importance, with each region discretized at an appropriate resolution. This segmentation allows the system to focus computational resources on critical regions while using fewer resources in less important areas, thereby maintaining accuracy where needed while reducing overall computational time.
2Measurement precision
If finer spatial discretization is used, then accuracy of scatter flux calculation is improved, but number of computations increases
Solution Approach 1:
Different spatial regions are assigned different discretization resolutions based on their contribution to the scatter flux. Regions that significantly affect the scatter flux are discretized with finer resolution, while other regions use coarser resolution. This maintains calculation accuracy for the most important regions while improving overall computational efficiency.
Solution Approach 2:
The discretization resolution parameter is changed dynamically based on the spatial location and scatter importance. By adjusting this parameter locally rather than keeping it uniform, the system achieves better accuracy-to-computation trade-off, improving productivity without sacrificing necessary precision.
3Measurement precision
If higher resolution is used throughout the domain, then precision of primary flux separation is maintained, but computational resources are wasted in low-scatter regions
Solution Approach 1:
The patent applies different discretization resolutions to different spatial regions based on their scatter characteristics. Regions with high scatter are discretized at higher resolution to maintain precision in primary flux separation, while regions with low scatter use coarser resolution to conserve computational resources. This resolves the contradiction by making resource allocation local rather than uniform.
Solution Approach 2:
Instead of applying full high-resolution discretization uniformly across the entire domain, the patent applies high resolution only partially to regions where it is truly needed for accurate primary flux separation. This partial action approach maintains necessary precision while avoiding waste of computational resources in regions where high resolution would be excessive.
Data Source
AI summary
X-ray scatter simulations to correct computed tomography (CT) data can be accelerated using a non-uniform discretization of the RTE, reducing the number of computations without sacrificing precision. For example, a coarser discretization can be used for higher-order/multiple-scatter flux, than for first-order-scatter flux. Similarly, precision is preserved when coarser angular resolution is used to simulate scatter within a patient, and finer angular resolution used for the scatter flux incident on detectors. Finer energy resolution is more beneficial at lower X-ray energies, and coarser spatial resolution can be applied to regions exhibiting less X-ray scatter (e.g., air and regions with low radiodensity). Further, predefined non-uniform discretization can be learned from scatter simulations on training data (e.g., a priori compressed grids learned from non-uniform grids generated by adaptive mesh methods).