Motion-Corrected X-Ray CT Reconstruction With Noise-Aware Filtering
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing motion-corrected image reconstruction methods in X-ray CT scans fail to accurately distinguish between noise and motion, leading to erroneous motion recognition and excessive smoothing, which impairs image quality and diagnostic accuracy for subjects with motion, such as the heart.
Innovation Solution
An X-ray CT apparatus and method that adjusts smoothing parameters based on the noise relation between image pairs using indicators like tube current values to minimize noise loss and maximize motion correction, preventing false deformations and improving diagnostic accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If noise reduction filtering is applied to images used for motion estimation, then noise is reduced, but motion components may be lost due to excessive smoothing
Solution Approach 1:
The patent dynamically adjusts the smoothing parameter of the filter based on the local variance calculated from the image pair. By changing the parameter (smoothing strength) according to local image characteristics, the filter reduces noise in low-variance regions while preserving motion components in high-variance regions, thus resolving the contradiction between noise reduction and motion preservation
2Device complexity
If motion correction is performed without considering current subject motion, then processing is simplified, but unnatural distortion occurs in images with small motion
Solution Approach 1:
The patent introduces a dynamic judgment mechanism that determines whether motion correction should be applied based on calculated motion information from the image pair. Instead of applying fixed motion correction, the system adaptively activates correction only when motion is detected, preventing unnecessary distortion in static or minimally moving subjects while maintaining correction capability when motion is present
3Device complexity
If smoothing parameter is fixed for noise reduction, then processing is simpler, but difference in noise between images is erroneously recognized as motion
Solution Approach 1:
The patent calculates local variance for each pixel position in the image pair and uses this local characteristic to determine the appropriate smoothing parameter. By making the smoothing parameter spatially varying rather than globally fixed, the filter can adapt to local noise characteristics and motion patterns, preventing erroneous motion detection caused by fixed parameter smoothing
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
The method enhances diagnostic accuracy and efficiency by accurately correcting motion in X-ray CT images, reducing false deformations and maintaining image quality even with varying noise levels.
Implementation Method 1
an X-ray source and an X-ray detector which rotate around a subject, and acquire transmitted X-ray data of the subject
Data Source
AI summary
When motion information of a subject during scanning is acquired from an image pair and motion-corrected image reconstruction is performed to generate a CT image, erroneous recognition and excessive smoothing of a motion component are prevented, and the accuracy of motion correction is improved. A filtering unit for reducing noise of the image pair is provided. The filtering unit acquires a relation between indicators for noise amounts acquired based on information that affects noise amounts of a first image and a second image constituting the image pair, and adjusts smoothness of each image using the relation between the indicators. The presence or absence of a motion is determined for the image pair subjected to noise reduction by filtering, normalization degrees of the two images are made different according to the presence or absence of the motion and the two images are normalized, and the magnitude and a direction of the motion are detected.