Spectral CT Noise Removal via Structure Model Fitting
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Spectral CT images are inherently noisier than conventional CT images due to higher radiation doses and narrow angle projections, leading to blurred images and increased risks from contrast materials used in contrast-enhanced studies.
Innovation Solution
A method and apparatus for estimating local noise values in spectral images, fitting local structure models to de-noise voxels based on noise models, and selecting appropriate models to reduce noise while preserving image structure, thereby enhancing image quality and reducing radiation dose.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If spectral CT imaging is performed with higher radiation dose, then image quality and material decomposition accuracy improve, but patient radiation exposure and cancer risk increase
Solution Approach 1:
The patent applies preliminary action by performing denoising operations before material decomposition. The system first estimates noise characteristics from the spectral images, then removes noise components prior to conducting material decomposition. This preliminary noise removal improves the accuracy of subsequent material decomposition while allowing for reduced radiation dose acquisition.
Solution Approach 2:
The patent replaces the mechanical approach of increasing radiation dose to improve signal quality with a computational approach. Instead of mechanically increasing photon flux, the system uses algorithms to estimate and remove noise components from lower-dose spectral images, substituting physical dose increase with computational noise reduction.
2Object-affected harmful factors
If radiation dose is reduced to decrease patient exposure, then radiation risk decreases, but image noise increases leading to blurred images
Solution Approach 1:
The patent replaces the mechanical relationship between radiation dose and image quality with a computational noise modeling approach. The system models noise characteristics as a function of radiation dose and uses these models to remove noise from low-dose images, decoupling the traditional trade-off between dose and image quality.
Solution Approach 2:
The patent changes the parameter relationship by introducing noise estimation and removal as an intermediate processing step. Instead of directly relating radiation dose to image quality, the system introduces noise parameters that can be independently estimated and removed, allowing low-dose acquisition with high-quality output.
3Illumination intensity
If contrast material volume is increased to improve contrast to noise ratio, then image contrast improves, but patient risk from contrast material increases
Solution Approach 1:
The patent applies preliminary action by performing noise removal on spectral images before contrast material administration or using images acquired at earlier time points. This preliminary denoising improves the baseline image quality, allowing for reduced contrast material volume while maintaining adequate contrast to noise ratio.
Solution Approach 2:
The patent creates virtual contrast-enhanced images by processing non-contrast spectral images through noise removal and material decomposition algorithms. This virtual copying of contrast enhancement effects reduces or eliminates the need for actual contrast material administration while maintaining diagnostic image quality.
Data Source
Figure 1
Figure 2~4
Figure 5~6
AI summary
A method includes estimating structure models for a voxel(s) of a spectral image based on a noise model, fitting structure models to a 3D neighborhood about the voxel(s), selecting one of the structure models for the voxel(s) based on the fittings and predetermined model selection criteria, and de-noising the voxel(s) based on the selected structure model, producing a set of de-noised spectral images. Another method includes generating a virtual contrast enhanced intermediate image for each energy image of a set of spectral images corresponding to different energy ranges based on de-noised spectral images, decomposed de-noised spectral images, an iodine map, and a contrast enhancement factor; and generating final virtual contrast enhanced images by incorporating a simulated partial volume effect with the intermediate virtual contrast enhanced images. Also described herein are approaches for generating a virtual non-contrasted image, a bone and calcification segmentation map, and an iodine map for multi-energy imaging studies.