Spectral CT Material Decomposition for Metal Artifact Reduction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing CT imaging methods suffer from severe metal artifacts that blur high-attenuation material areas and interfere with normal tissue structures, leading to reduced diagnostic or inspection performance, particularly in critical tissues and organs.
Innovation Solution
A method involving material decomposition (MD) calibration using multiple reference phantoms with known characteristics, constructing a system characteristic model, and employing polynomial approximation and maximum likelihood estimation to generate virtual monoenergetic images through projection-based data recombination, reducing metal artifacts while preserving normal tissue structure.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Object-affected harmful factors
If sinogram painting methods (LMAR or NMAR) are used to reduce metal artifacts, then metal artifact suppression is achieved, but image blurring occurs in areas around high-attenuation materials
Solution Approach 1:
The patent segments the projection data by energy bins, processing different energy ranges separately through material decomposition. This allows selective reconstruction of metal-free regions while preserving metal information in specific energy bins, avoiding the blanket blurring caused by conventional sinogram painting methods.
Solution Approach 2:
The patent changes the energy parameter by generating virtual monoenergetic images at different energy levels (e.g., 40 keV, 60 keV, 80 keV). By adjusting the energy parameter, the system can optimize the balance between metal artifact suppression and image quality preservation, as lower energy bins provide better soft tissue contrast while higher energy bins reduce metal attenuation.
2Measurement precision
If conventional CT imaging is used, then high-resolution three-dimensional structural information is obtained, but serious metal artifacts blur high-attenuation material areas and interfere with normal tissue areas
Solution Approach 1:
The patent introduces an energy dimension to the conventional spatial three-dimensional imaging. By adding energy bin separation, the system transforms the problem from a two-dimensional image processing challenge to a three-dimensional energy-space problem, allowing independent optimization of metal artifact suppression and structural information preservation across different energy levels.
Solution Approach 2:
The patent uses virtual monoenergetic images as an intermediary representation between the raw projection data and the final diagnostic images. These virtual images serve as a mediator that separates metal artifact information from tissue information, allowing the system to reconstruct high-quality images without direct metal interference while maintaining accurate three-dimensional structural data.
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
Effectively reduces metal artifacts in CT images by generating virtual monoenergetic images that maintain normal tissue structure integrity and improve image quantification accuracy.
Implementation Method 1
a photon-counting detector to detect the X-ray photons
Implementation Method 2
spectral CT data of different energy bins based on projection corresponding to the MD calibration phantoms
Data Source
AI summary
A method for reducing metal artifacts in CT images, including the following steps: a material decomposition (MD) calibration step, using multiple MD calibration phantoms with known characteristics and multiple spectral CT data corresponding thereto to construct a system characteristic model of spectral CT; and a MD testing step, including: the following steps: imaging multiple testing objects with a different unknown material and thickness to obtain projection-based multiple spectral CT imaging data of different energy bins; obtaining corresponding multiple basis material images of different materials based on projection data according to the spectral CT imaging data and the system characteristic model of spectral CT; and combining the basis material images and a photon energy information to be recombined with each other to obtain multiple virtual monoenergetic images.