Scatter Correction for CT Imaging Using Hybrid Kernel Models
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current scatter correction methods in CT imaging are not optimized to effectively model scatter from patient support structures like the patient table, leading to underestimation of scatter and artifacts such as cupping in CT images, especially in half-fan geometry with an offset detector.
Innovation Solution
The method involves a Piercing Point Equalization technique and an Analytical Hybrid Kernel model to estimate and correct for scattered radiation from adjacent objects, using a combination of empirical and analytical approaches to improve scatter correction, particularly for patient table scatter, by applying correction factors and hybrid kernels that adapt to projection angles and object geometry.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional scatter correction methods are used, then the imaging process is simple, but scatter from patient table is underestimated causing cupping artifacts
Solution Approach 1:
The scatter correction method segments the scattering sources into distinct categories: patient body scatter and patient table scatter. Each source type is modeled separately with dedicated correction factors and kernels, allowing accurate estimation of scatter from both the patient and the supporting table structure without confusion between the two sources.
Solution Approach 2:
The method applies different scatter correction characteristics to different regions and sources. Patient body scatter uses one set of correction factors and kernels, while patient table scatter uses another set optimized for the table's specific scattering properties. This localized approach ensures each scatter source is corrected according to its unique characteristics.
2Area of stationary object
If half-fan geometry with offset detector is used, then field-of-view is increased, but scatter correction is worsened due to geometric complexity
Solution Approach 1:
The scatter correction method dynamically adapts to the half-fan geometry by using projection-angle-dependent correction factors and kernels. The correction parameters vary based on the specific geometric configuration at each projection angle, allowing the system to maintain accurate scatter correction while preserving the extended field-of-view capability of the offset detector geometry.
3Manufacturing precision
If scatter from adjacent objects is not corrected, then the imaging process is simple, but artifacts and loss of spatial resolution occur
Solution Approach 1:
The method introduces intermediary correction factors and kernels that mediate between the raw projection data and the final corrected images. These intermediary elements (correction factors, kernels) process the scatter information separately and then integrate it back into the image formation process, preserving spatial resolution while removing artifacts caused by uncorrected scatter from adjacent objects like the patient table.
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
These methods significantly reduce artifacts and improve image quality by accurately estimating and correcting for scattered radiation, resulting in enhanced spatial and contrast resolution in CT images, as demonstrated by improved pelvis CT slices with reduced cupping artifacts.
Implementation Method 1
transmitting radiation from a 'point source' 105 through the object 102, which will absorb some of the radiation based on its size, density, and atomic composition
Implementation Method 2
when a quantum of radiation is absorbed by a portion of the object, one or more scattered rays are often generated that deviate from the transmission path
Data Source
Figure 1a
Figure 1b
Figure 2
AI summary
A system and method for forming an adjusted estimate of scattered radiation in a radiographic projection of a target object, which incorporates scattered radiation from objects adjacent to the target object, such as a patient table. A piercing point equalization method is disclosed, and a refinement of analytical kernel methods which utilizes hybrid kernels is also disclosed.