B0 Image Calculation via Optimization for MRI Gradient Correction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for correcting gradient nonlinearity bias in diffusion weighted imaging require a direct measurement of the b0 image, which is sensitive to noise, reducing the accuracy of the correction process.
Innovation Solution
Calculating the b0 image using an optimization process with a set of equations constructed from multiple diffusion weighted images acquired at different b-values and directions, allowing for improved noise resistance and accuracy in gradient nonlinearity correction.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a b0 image is directly measured, then the correction process can be performed, but noise in the b0 image reduces the accuracy of the gradient nonlinearity correction
Solution Approach 1:
The patent introduces an optimization process as an intermediary between the measured b0 image and the final correction application. This optimization process uses the measured b0 image along with multiple diffusion-weighted images to calculate an optimized b0 value that filters out noise while preserving the underlying signal, thereby improving correction accuracy without requiring a perfect direct measurement
Solution Approach 2:
The patent creates an optimized copy of the b0 image through calculation using multiple diffusion-weighted images with different b-values and gradient directions. This calculated b0 image serves as a noise-reduced replica that can be used for correction purposes, eliminating the need to rely solely on the noisy directly-measured b0 image
2Measurement precision
If multiple diffusion weighted images are acquired at different b-values and directions, then the b0 image calculation accuracy is improved, but the acquisition time and complexity increase
Solution Approach 1:
The patent makes the acquired diffusion-weighted images serve multiple purposes: they are used both for calculating the optimized b0 values and for the subsequent gradient nonlinearity correction process. This multi-functionality maximizes the utility of each acquired image, reducing the need for additional separate acquisitions and thereby minimizing time loss despite the multiple b-values and directions required
Data Source
AI summary
The invention provides for a medical imaging system (100, 300). The execution of the machine executable instructions (110) causes a processor (102) to: receive (200) multiple diffusion weighted images (112) of a subject (318), wherein the multiple diffusion weighted images each have an assigned b-value, wherein the multiple diffusion weighted images each have an assigned diffusion weighting direction, wherein for a region of interest (309) there is at least one corresponding voxel (506) in each of the multiple diffusion weighted images; construct (202) a set of equations (114) for each of the at least one corresponding voxel, wherein the set of equations is constructed from an apparent diffusion equation for the assigned diffusion weighting direction of each of the multiple diffusion weighted images; solve (204) the set of equations for each voxel for the b0 value as an optimization; and construct (206) a b0 image using the b0 value for each voxel.