Sinogram-Based Motion Correction for Medical Imaging Artifacts
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Diagnostic images obtained using gamma rays or X-rays often suffer from noise and motion blur due to the movement of objects during imaging, which existing technologies fail to effectively correct, especially in medical imaging techniques like PET and CT scans.
Innovation Solution
A method and apparatus that generate sinograms based on object motion, determine a region of interest, extract relevant sinograms, estimate motion information, and correct data to remove noise and motion artifacts, specifically by generating a first sinogram for a state of the object and a second sinogram for its states, and using this information to correct the image data.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If diagnostic imaging is performed using gamma rays or X-rays, then image data can be obtained, but noise and motion blur are introduced due to object movement during imaging
Solution Approach 1:
The patent segments the imaging data by dividing the object into multiple regions of interest (ROIs) and processing each region separately. Sinograms are divided into multiple sets corresponding to different ROIs, allowing motion correction to be applied specifically to affected regions without compromising the entire image, thus reducing motion blur while maintaining overall image quality
Solution Approach 2:
The patent changes the parameter of motion information by estimating and determining actual motion parameters for each ROI based on sinogram data. By calculating motion parameters such as displacement and velocity for specific regions, the system can dynamically adjust correction strategies to compensate for motion blur and noise introduced during imaging
2Manufacturing precision
If motion correction is applied to the entire object, then motion blur may be reduced, but processing time and computational complexity increase significantly
Solution Approach 1:
The patent divides the object into multiple regions of interest and processes only the relevant sinograms corresponding to each ROI. This segmentation approach reduces the volume of data requiring motion correction processing, thereby decreasing computational complexity and processing time while maintaining motion correction accuracy for the most critical regions
Solution Approach 2:
The patent applies different processing strategies to different regions based on their specific characteristics. By identifying ROIs that are most susceptible to motion blur and applying motion correction selectively to these regions, the system optimizes the balance between correction accuracy and processing efficiency, avoiding unnecessary processing of stable regions
3Productivity
If regions of interest are identified and processed separately, then processing efficiency improves, but the complexity of determining accurate ROIs increases
Solution Approach 1:
The patent performs preliminary ROI determination by identifying regions of interest before the main motion correction processing. By pre-segmenting the object and determining which ROIs require correction based on initial analysis of sinogram data, the system simplifies the subsequent processing steps and improves overall efficiency while managing the complexity of ROI identification
Solution Approach 2:
The patent enables the system to automatically determine ROIs by analyzing the sinogram data itself without requiring extensive external input or complex manual configuration. The motion estimation and ROI identification processes utilize the inherent information in the sinogram data to self-determine which regions require correction, reducing operational complexity
Data Source
AI summary
A method and an apparatus for generating images are provided. The method includes generating a first sinogram for a state of an object from among states of the object based on a motion of the object, and a second sinogram for the states, based on data obtained from the object, and determining a region of interest (ROI) of the object based on the first sinogram. The method further includes extracting, from the second sinogram, third sinograms corresponding to the ROI, and estimating motion information of the ROI based on the third sinograms. The method further includes correcting the data based on the motion information.