MRI Distortion Correction Using Shift and Jacobian Maps
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for correcting image distortions in magnetic resonance imaging using a cost mathematical function are prone to corrupting anatomical structures when images have strong distortions or unexpected differences, particularly due to pulsation and body movements.
Innovation Solution
An image processing apparatus that generates a shift map by optimizing a cost function using differences between two magnetic resonance images and edge images, employing a fitting model to correct distortions while maintaining anatomical structures, and applying a Jacobian map for brightness correction.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If a cost mathematical function is used to correct image distortions by solving an optimization problem, then image distortions in phase encoding directions are reduced, but anatomical structures are corrupted when images have strong distortions or unexpected differences
Solution Approach 1:
The patent segments the correction process into two distinct components: a shift map for positional distortion correction and a Jacobian map for brightness/intensity correction. This segmentation allows each component to address specific aspects of distortion independently, preventing the corruption of anatomical structures while maintaining correction effectiveness.
Solution Approach 2:
The patent introduces a Jacobian map as an intermediary element that mediates between the shift map and the final corrected image. The Jacobian map compensates for brightness changes caused by the shift operation, acting as a buffer that prevents direct transmission of distortion artifacts to the corrected image, thereby preserving anatomical structure integrity.
2Quantity of substance
If pixel compression is applied to reduce image size, then storage and processing efficiency are improved, but distortion differences between forward and backward images increase
Solution Approach 1:
The patent addresses the distortion measurement problem by operating in the frequency domain (k-space) rather than directly in the spatial domain. This dimensional transformation allows for more accurate distortion measurement and correction while working with compressed image data, as the frequency domain representation preserves structural information that may be lost in spatial compression.
Data Source
AI summary
An image processing apparatus according to an embodiment includes processing circuitry. The processing circuitry is configured to obtain two magnetic resonance images corresponding to two phase encoding directions opposite to each other. The processing circuitry is configured to generate a shift map related to shifting a plurality of pixels in the two magnetic resonance images, by optimizing a cost function using a first difference between the two magnetic resonance images and a second difference between two edge images generated on the basis of the two magnetic resonance images. The processing circuitry is configured to generate a correction image obtained by correcting distortions of the two magnetic resonance images on the basis of the two magnetic resonance images and the shift map.