Local Gradient Non-Linearity Correction in Diffusion MRI

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current magnetic resonance (MR) systems for diffusion imaging assume ideal linear diffusion gradients, leading to systematic errors in the determination of trace-weighted MR images and apparent diffusion coefficient (ADC) maps, especially at distances from the isocenter, due to uncorrected non-linearities in diffusion gradient fields.

Innovation Solution

A method that generates approximate trace-weighted MR images by creating at least three diffusion-weighted MR images with different coding directions, determining spatial inhomogeneities in the gradient fields, and applying optimization-based weighting factors to combine these images, allowing for local correction of gradient non-linearities without requiring explicit ADC correction.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If ideal linear diffusion gradients are assumed in diffusion imaging, then the imaging process is simplified and faster, but systematic errors increase with distance from the isocenter due to uncorrected gradient nonlinearities

Engineering Contradiction:
Improveimaging speedVSAvoidADC map accuracy
Core Design Contradiction:
ProductivityVSMeasurement precision

Solution Approach 1:

The patent applies preliminary action by performing a field map measurement before the actual diffusion imaging to characterize the gradient nonlinearities. This pre-characterization allows the system to store correction factors that are then applied during image reconstruction, enabling both fast imaging and accurate correction without requiring additional scan time during the main imaging sequence.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The patent replaces the mechanical/physical approach of using ideal linear gradients with a computational correction method. Instead of physically adjusting gradient coils to achieve perfect linearity, the system uses field map measurements and computational algorithms to calculate and apply correction factors, substituting a mathematical model for the physical gradient field correction.

Inventive Principle:
Principle #28Mechanics substitution (Replace mechanical system)

2Measurement precision

If gradient non-linearities are corrected using existing methods (Malyarenko or Bammer), then ADC accuracy is improved, but the complexity of the imaging protocol increases significantly

Engineering Contradiction:
ImproveADC map accuracyVSAvoidimaging protocol complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The patent extracts the correction problem from the main imaging sequence by performing field map measurements separately before the actual diffusion imaging. This separation allows the correction factors to be pre-calculated and stored, then applied during image reconstruction without complicating the main imaging protocol. The correction step is taken out as a independent preprocessing operation.

Inventive Principle:
Principle #2Taking out (Extraction)

Solution Approach 2:

The patent introduces an intermediary field map measurement that characterizes the gradient nonlinearities. This field map acts as a mediator between the gradient system and the diffusion imaging sequence, providing the necessary correction information without requiring modification of the main imaging protocol. The field map serves as an intermediate step that decouples the correction complexity from the main imaging sequence.

Inventive Principle:
Principle #24Intermediary (Mediator)

3Measurement precision

If more diffusion directions are used to achieve exact correction (Bammer method), then gradient non-linearity correction is more accurate, but the scanning time and data processing complexity increase

Engineering Contradiction:
Improvegradient correction accuracyVSAvoidscan time
Core Design Contradiction:
Measurement precisionVSLoss of time

Solution Approach 1:

The patent applies partial action by using a simplified correction approach that does not require the full six-direction tensor reconstruction. Instead of performing complete tensor fitting with multiple diffusion directions, the method uses a reduced set of measurements combined with field map-based correction factors, achieving sufficient correction accuracy with less scanning time and computational effort.

Inventive Principle:
Principle #16Partial or excessive action

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

This method provides higher quality and cost-effective diffusion imaging by accurately accounting for non-linear gradient field variations, improving the precision of trace-weighted MR images and ADC maps, even with fewer than six diffusion directions and non-orthogonal coding.

Implementation Method 1

diffusion imaging using an MRI system is an indispensable procedure in medical diagnostics today. One established application is the determination of trace-weighted MRI images, in which direction-dependent diffusion effects are averaged

Methodology Applied
Scientific EffectDiffusion: Diffusion

Implementation Method 2

diffusion imaging using a magnetic resonance (MR) system... each diffusion-weighted MR image being based on a diffusion coding... the diffusion coding can comprise a diffusion coding direction and an associated diffusion gradient field

Methodology Applied
Scientific EffectMagnetic Field: Magnetic Field

Data Source

PatentEP3299836B1Local correction of gradient non-linearities in diffusion weighted MRI
Publication Date: 2024.09.25 SIEMENS HEALTHINEERS AG
  • EP3299836B1 patent drawingFigure 1
  • EP3299836B1 patent drawingFigure 2~3
  • EP3299836B1 patent drawing

AI summary

The present invention relates to a method for generating an approximate track-weighted MRI image of a subject. At least three diffusion-weighted MRI images of the subject are acquired, based on different diffusion coding directions and diffusion gradient fields with substantially the same diffusion coding strengths. Furthermore, based on spatial inhomogeneities of the diffusion gradient fields, a weighting factor is determined for each pixel of each of the at least three diffusion-weighted MRI images using an optimization method. Based on these weighting factors, the at least three diffusion-weighted MRI images are combined to generate the approximate track-weighted MRI image. Finally, an ADC map is determined based on the approximate track-weighted images and the effective diffusion coding strength.In addition, based on a track-weighted image at effective diffusion coding strength, a track-weighted MR image at nominal diffusion coding strength is synthesized.