CT Image Processing Edge-Preserving Denoising Spike Suppression
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Heavy-tailed noise in CT images, particularly in Photon Counting CT systems and low-dose scans, leads to increased spike occurrence after denoising, resulting in false-positive spike identifications and loss of texture and contrast.
Innovation Solution
A computer-implemented method for image-processing of CT images that involves pre-processing steps using an edge-preserving denoising algorithm, followed by an adaptive spike suppression algorithm to reduce the number of spikes while maintaining image texture.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If edge-preserving denoising is applied to remove noise, then noise reduction is achieved, but the number of spikes increases
Solution Approach 1:
The patent applies preliminary action by performing edge-preserving denoising before spike suppression. The denoising step is performed first to reduce noise and improve the signal-to-noise ratio, which makes subsequent spike identification and suppression more effective. This preliminary processing prepares the image data for the following spike removal step.
Solution Approach 2:
The patent extracts and removes spikes from the image data using an adaptive spike suppression algorithm. After denoising, the algorithm identifies and extracts spike artifacts from the pre-processed image, separating them from the actual tissue structures. This extraction process removes the harmful spike components while preserving the underlying image information.
2Object-generated harmful factors
If spike suppression is applied to remove spikes, then spike number is reduced, but texture and contrast are lost
Solution Approach 1:
The patent performs edge-preserving denoising as a preliminary step before spike suppression. This preliminary denoising improves the signal-to-noise ratio and enhances the distinguishability of spikes from actual tissue structures. As a result, the subsequent spike suppression algorithm can more accurately identify and remove only the spike artifacts while preserving genuine tissue texture and contrast information.
Solution Approach 2:
The patent employs an adaptive spike suppression algorithm that uses feedback mechanisms to dynamically adjust its processing based on the image characteristics. The algorithm continuously monitors the image data and adapts its suppression strength and criteria, allowing it to remove spikes while preserving important tissue structures and maintaining texture and contrast information.
3Reliability
If denoising is applied to improve signal-to-noise ratio, then noise is reduced, but false-positive spike identifications increase
Solution Approach 1:
The patent applies edge-preserving denoising as a preliminary step that improves the signal-to-noise ratio by removing random noise while preserving edges and structures. This enhanced signal quality provides a better foundation for subsequent spike identification, reducing false-positive identifications by making the true spike structures more distinguishable from the background.
Solution Approach 2:
The adaptive spike suppression algorithm incorporates feedback mechanisms that continuously evaluate the image characteristics and adjust identification criteria accordingly. This feedback loop allows the algorithm to learn from the denoised image structure and refine its spike detection, improving accuracy by adapting to the specific tissue patterns and reducing false positives.
Data Source
AI summary
The invention provides a computer-implemented method for image-processing of CT images, the method comprising performing one or more pre-processing steps on a CT image so as to obtain a pre-processed CT image, wherein the one or more pre-processing steps comprise applying an edge-preserving denoising algorithm; and performing an adaptive spike suppression algorithm on the pre-processed CT image to obtain a processed CT image, the adaptive spike suppression algorithm being configured such that the processed CT image has a reduced number of spikes as compared to the pre-processed CT image.