Sinogram Domain Registration for Spectral CT Motion Correction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Spectral CT systems face errors and artifacts in material decomposition due to patient motion during energy scans, which are not accurately corrected by existing registration processes, leading to degraded image quality.
Innovation Solution
A motion-compensated reprojection-based material-decomposition method that registers projection data in the sinogram domain, maintaining beam hardening effects to ensure accurate material decomposition without introducing non-physical artifacts, using techniques like split-step and cost-function methods for projection data alignment.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If registration is performed between CT images at different X-ray energies, then material differentiation is improved, but patient motion between scans introduces errors and artifacts
Solution Approach 1:
The patent performs registration of projection data in the sinogram domain before material decomposition, rather than registering reconstructed images after acquisition. This preliminary registration aligns the projection data from different energy scans before they are processed into material decomposition results, thereby correcting for patient motion artifacts that would otherwise degrade the accuracy of material differentiation.
2Reliability
If registration is performed on reconstructed images, then motion correction is achieved, but beam hardening effects are lost and non-physical artifacts are introduced
Solution Approach 1:
The patent performs registration in the sinogram domain before material decomposition and image reconstruction, preserving the beam hardening information in the projection data. By registering the projection data rather than reconstructed images, the method maintains the physical characteristics of the X-ray attenuation data while still correcting for patient motion, thus avoiding the introduction of non-physical artifacts.
Solution Approach 2:
The patent uses the sinogram domain as an intermediary representation between the raw projection data and the final reconstructed images. Registration is performed in this intermediate domain, allowing motion correction to be applied while preserving the physical characteristics of the projection data. This intermediary approach enables both motion correction and preservation of beam hardening effects.
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 achieves accurate material decomposition by correcting for patient motion while preserving the physical characteristics of the projection data, resulting in improved image quality and reduced artifacts.
Implementation Method 1
A radiation source, such as an X-ray tube, irradiates the body from one side
Implementation Method 2
spectral CT technology, which can provide improvement in material differentiation and beam-hardening correction
Data Source
AI summary
A method and apparatus is provided to perform material decomposition based on spectral computed tomography (CT) projection data generated using registered reconstructed images. Registration is performed in the image domain, whereas material decomposition is performed in the sinogram domain. In the sinogram domain, material decomposition can include beam-hardening corrections. For at least two energy components, CT images are reconstructed, and registration is performed among the CT images. In certain implementations, the registered images are forward projected, and material decomposition is based on the resultant forward projections. In other implementations, motion images are generated from differences between the reconstructed CT images pre- and post-registration. The projection data is then corrected using forward projections of the motion images, and material decomposition is performed using the motion-corrected projection data.