Iterative Image Reconstruction Texture Blending

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Iterative image reconstruction techniques in non-invasive imaging reduce noise but alter the image texture, making images less familiar and potentially confusing for radiologists trained on analytic reconstruction methods, which can affect diagnostic accuracy.

Innovation Solution

The method involves reconstructing images with a tailored image texture by blending analytically-reconstructed and iteratively-reconstructed images in frequency space, using weighting functions and masks to combine higher-frequency components from analytic images with lower-frequency components from iterative images, thereby reducing noise while preserving the familiar texture.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Object-affected harmful factors

If iterative image reconstruction techniques are used, then image noise is reduced, but image texture becomes altered and less familiar to radiologists

Engineering Contradiction:

Improveimage noiseVSAvoidimage texture familiarity

Core Design Contradiction:

Object-affected harmful factorsVSDifficulty of detecting and measuring

Solution Approach 1:

The image frequency spectrum is segmented into different frequency ranges, with high-frequency components preserved from the analytic reconstruction and low-frequency components taken from the iterative reconstruction. This segmentation allows selective combination of beneficial properties from both reconstruction methods.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

Different frequency regions of the image are assigned different qualities: high-frequency regions retain the texture characteristics of analytic reconstruction for familiarity, while low-frequency regions adopt the noise-reduced properties of iterative reconstruction. This local quality differentiation resolves the contradiction between noise reduction and texture familiarity.

Inventive Principle:

Principle #3Local quality

2Difficulty of detecting and measuring

If analytic image reconstruction is used, then image texture remains familiar, but image noise is higher

Engineering Contradiction:

Improveimage texture familiarityVSAvoidimage noise

Core Design Contradiction:

Difficulty of detecting and measuringVSObject-affected harmful factors

Solution Approach 1:

The frequency spectrum is divided into regions where high frequencies contribute to texture and low frequencies contribute to noise. By segmenting and selectively combining components from both analytic and iterative reconstructions, the method achieves both texture familiarity and noise reduction.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The final image is constructed as a composite of frequency components from two different reconstruction methods (analytic and iterative), similar to how composite materials combine properties of different materials. The high-frequency components provide texture familiarity while low-frequency components provide noise reduction.

Inventive Principle:

Principle #40Composite materials

Data Source

PatentUS11270477B2Systems and methods for tailored image texture in iterative image reconstruction

Publication Date: 2022.03.08 GE PRECISION HEALTHCARE LLC

AI summary

Methods and systems are provided for reconstructing images with a tailored image texture. In one embodiment, a method comprises acquiring projection data, and reconstructing an image from the projection data with a desired image texture. In this way, iterative image reconstruction techniques may be used to substantially reduce image noise, thereby enabling a reduction in injected contrast and/or radiation dose, while preserving an image texture familiar from analytic image reconstruction techniques.