Iterative CT Reconstruction with Sharpness-Driven Regularization

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current iterative image reconstruction methods in computed tomography (CT) fail to ensure similar spatial resolution across spectral images, leading to artifacts and incorrect quantitative values due to uniform noise reduction across images.

Innovation Solution

The method involves performing multiple passes of iterative reconstruction, updating regularization parameters based on the sharpness of intermediate photoelectric and Compton scatter images to achieve similar spatial resolution between spectral component images.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If a uniform regularization parameter is used for iterative image reconstruction, then image noise is reduced uniformly across the image, but spatial resolution becomes inconsistent across different spectral images

Engineering Contradiction:
Improveimage noise reductionVSAvoidspatial resolution consistency
Core Design Contradiction:
ReliabilityVSManufacturing precision

Solution Approach 1:

The patent applies different regularization parameters to different spectral images (photoelectric and Compton scatter images) based on their individual sharpness characteristics. This local differentiation allows each spectral image to have optimized noise reduction while maintaining consistent spatial resolution across all spectral components.

Inventive Principle:
Principle #3Local quality

Solution Approach 2:

The patent dynamically adjusts the regularization parameter for each spectral image based on its measured sharpness. The regularization parameter is not fixed but is adapted according to the specific characteristics of each spectral image, enabling optimal balance between noise reduction and spatial resolution consistency.

Inventive Principle:
Principle #15Dynamics

2Loss of information

If spectral images are reconstructed separately with linear combination, then spectral information is preserved, but artifacts are introduced due to different spatial resolutions

Engineering Contradiction:
Improvespectral information preservationVSAvoidartifacts from resolution mismatch
Core Design Contradiction:
Loss of informationVSObject-generated harmful factors

Solution Approach 1:

The patent ensures that each spectral image is reconstructed with appropriate regularization to achieve consistent spatial resolution. By optimizing the regularization parameter for each spectral image based on its sharpness, the patent eliminates resolution mismatches that would otherwise cause artifacts during linear combination, while preserving spectral information.

Inventive Principle:
Principle #3Local quality

3Reliability

If a fixed regularization parameter is used to decrease image noise by 30%, then noise reduction is achieved, but sharpness of edges and low contrast structures is compromised

Engineering Contradiction:
Improveimage noise levelVSAvoidedge sharpness and low contrast structure detail
Core Design Contradiction:
ReliabilityVSManufacturing precision

Solution Approach 1:

The patent dynamically determines the regularization parameter by measuring the sharpness of each spectral image and adjusting the parameter accordingly. This dynamic adaptation allows the system to achieve optimal noise reduction (e.g., 30% decrease) while preserving edge sharpness and low contrast structure details, as each spectral image receives a customized regularization strength based on its specific characteristics.

Inventive Principle:
Principle #15Dynamics

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 ensures that reconstructed spectral component images have consistent spatial resolution, reducing artifacts and improving the accuracy of quantitative values by dynamically adjusting regularization parameters based on image sharpness.

Implementation Method 1

an intermediate photoelectric image and an intermediate Compton scatter image are generated using an iterative reconstruction algorithm

Methodology Applied
Scientific EffectPhotoelectric effect: Photoelectric Effect

Implementation Method 2

an intermediate photoelectric image and an intermediate Compton scatter image are generated using an iterative reconstruction algorithm

Methodology Applied
Scientific EffectCompton scattering: Compton Scattering

Data Source

PatentEP3195265B1Iterative image reconstruction with a sharpness driven regularization parameter
Publication Date: 2018.08.22 KONINKLIJKE PHILIPS NV
  • EP3195265B1 patent drawingFigure 1
  • EP3195265B1 patent drawingFigure 2
  • EP3195265B1 patent drawingFigure 3

AI summary

A method includes performing a first pass of an iterative image reconstruction in which an intermediate first spectral image and an intermediate second spectral image are generated using an iterative image reconstruction algorithm, start first spectral and second spectral images, and initial first spectral regularization and second spectral regularization parameters, updating at least one of the initial first spectral regularization or second spectral regularization parameters, thereby creating an updated first spectral regularization or second spectral regularization parameter, based at least on a sharpness of one of the intermediate first spectral or second spectral images, and performing a subsequent pass of the iterative image reconstruction in which an updated intermediate first spectral and second spectral image is generated using the iterative image reconstruction algorithm, the intermediate first spectral and second spectral images, and the updated first spectral regularization and Compton scatter regularization parameters.