Variance-Stabilizing X-Ray Image Denoising
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing X-ray imaging technologies face challenges in reducing noise levels while using low X-ray dosages, as conventional denoising methods are inadequate for signal-dependent noise and mixed noise situations, particularly at very low dose levels.
Innovation Solution
A method involving a variance-stabilizing transformation parameterized by X-ray imaging device properties and measurement parameters, followed by a noise reduction algorithm and inverse transformation, to generate denoised X-ray images, which can handle mixed noise situations and improve denoising quality at low dosages.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Object-affected harmful factors
If low X-ray intensity is used to lower the X-ray dose, then the radiation dosage is reduced, but the noise level in the images increases
Solution Approach 1:
The patent applies a variance-stabilizing transformation before noise reduction to prepare the data in advance. This preliminary processing step transforms the signal-dependent noise into approximately Gaussian noise with stabilized variance, making the subsequent noise reduction more effective and enabling better noise removal at low doses
Solution Approach 2:
The patent changes the parameter space by applying a variance-stabilizing transformation that modifies the statistical properties of the noise. This transformation changes the noise from being signal-dependent (Poisson-like) to having approximately constant variance (Gaussian), allowing standard denoising algorithms to work more effectively at low doses
2Reliability
If conventional denoising methods are applied to low-dose X-ray images, then noise reduction is attempted, but the methods are inadequate for signal-dependent noise and mixed noise situations
Solution Approach 1:
The patent introduces a variance-stabilizing transformation as an intermediary step between the raw low-dose image and the noise reduction algorithm. This intermediary transformation adapts the data to make it suitable for standard Gaussian-based denoising methods, bridging the gap between signal-dependent noise and Gaussian noise assumptions
Solution Approach 2:
The patent segments the noise reduction process into distinct stages: variance-stabilizing transformation, noise reduction algorithm application, and inverse transformation. This segmentation allows each stage to be optimized independently, with the transformation stage handling the signal-dependent noise characteristics and the noise reduction stage focusing on Gaussian noise removal
Data Source
AI summary
A method for processing at least one X-ray image is provided. A variance of noise is signal dependent. The method includes applying a variance-stabilizing transformation to image data of the at least X-ray image to generate variance-stabilized data. At least one transform parameter of the variance-stabilizing transformation is dependent on a property of the at least one X-ray image that depends on an X-ray imaging device and/or a measurement parameter used to record the at least one X-ray image. A noise reduction algorithm is applied to the variance-stabilized data to generate noise-reduced data, and an inverse transformation of the variance stabilizing transformation is applied to the noise-reduced data to generate a denoised X-ray image.