Medical Image Processing Reducing Motion Artifacts
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
X-ray CT image reconstruction is hindered by motion artifacts when scanning moving objects, leading to degraded image quality, particularly in high-resolution imaging of fast-moving body parts like the heart, where current methods either increase hardware complexity or fail to accurately represent motion relationships between adjacent times.
Innovation Solution
A medical image processing method and apparatus that generates motion information using partial angle reconstruction (PAR) image pairs, updates this information to minimize motion artifacts, and applies it to raw data to produce a reconstructed image, enabling motion compensation and reducing inter-cyclic motion artifacts in 3D imaging.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional X-ray CT image reconstruction is used to scan moving objects, then image acquisition speed is maintained, but motion artifacts occur and image quality is degraded
Solution Approach 1:
The system performs preliminary motion estimation by comparing PAR image pairs obtained at different phase sections before final image reconstruction. Motion information is estimated in advance and used to correct the raw data, preventing motion artifacts from occurring in the final reconstructed image
Solution Approach 2:
The system iteratively updates motion information by comparing PAR images at different phases, refines the motion estimation, and applies corrections to reduce motion artifacts. This feedback loop continues until motion artifacts are minimized in the reconstructed image
2Productivity
If scan time is reduced and pitch is increased to improve scanning efficiency, then productivity increases, but motion artifacts worsen due to faster movement
Solution Approach 1:
Motion compensation is performed preliminarily by estimating motion from PAR image pairs before final reconstruction, allowing the system to use faster scanning parameters (reduced scan time, increased pitch) while still correcting for the resulting motion artifacts
Solution Approach 2:
The system changes the temporal sampling parameters by acquiring data at multiple phase sections throughout the cardiac cycle and reconstructs images at different time points, effectively capturing motion information even during fast scanning with reduced scan time and increased pitch
3Measurement precision
If motion compensation methods are applied to reduce motion artifacts, then image quality improves, but computational complexity increases
Solution Approach 1:
The computational task is segmented into manageable parts: first generating PAR images from raw data, then comparing specific PAR image pairs at different phases to estimate motion, and finally applying motion correction. This segmentation reduces overall computational complexity compared to attempting full motion compensation at once
Solution Approach 2:
PAR images serve as intermediary representations that simplify motion estimation. Instead of directly analyzing raw projection data for motion, the system uses PAR images as an intermediate step, making motion compensation computationally more efficient while maintaining image quality
Data Source
AI summary
Provided is a medical image processing apparatus including a processor. The processor obtains raw data in a first phase section and generates first motion information by using at least one partial angle reconstruction (PAR) image pair including two PAR images respectively obtained in two phase sections in the first phase section that face each other. The processor also generates a summed image by summing a plurality of PAR images obtained at different phases within the first phase section by using the first motion information Second motion information is generated by updating the first motion information such that an image metric representing motion artifacts is minimized when being calculated from the summed image and a reconstructed image is generated by applying the second motion information to the raw data.