Diffusion-Weighted MRI Reconstruction Using Regularized Inverse Processes

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Conventional diffusion-weighted MRI techniques are sensitive to magnetic field inhomogeneity and motion, leading to geometric distortions and reduced diagnostic accuracy, especially in medical imaging applications like brain and prostate imaging.

Innovation Solution

The method involves acquiring and reconstructing diffusion-weighted MR images using a combined diffusion-weighted spin-echo and single-shot stimulated-echo sequence with non-Cartesian radial k-space trajectories, employing multiple diffusion-encoding gradients of varying strengths and directions, and utilizing regularized nonlinear and linear inverse reconstruction processes to determine coil sensitivities and image content across contiguous cross-sectional slices, thereby reducing sensitivity to magnetic field inhomogeneity and motion.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If conventional diffusion-weighted MRI techniques are used, then diffusion contrast is achieved, but geometric distortions and reduced diagnostic accuracy occur due to sensitivity to magnetic field inhomogeneity and motion

Engineering Contradiction:
Improvediagnostic accuracyVSAvoidsensitivity to magnetic field inhomogeneity and motion
Core Design Contradiction:
Measurement precisionVSObject-affected harmful factors

Solution Approach 1:

The imaging volume is divided into multiple contiguous cross-sectional slices that are sequentially acquired. Each slice is independently imaged and then reconstructed to form a complete volume, allowing motion and field inhomogeneity effects to be localized and corrected on a per-slice basis rather than affecting the entire volume.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

Coil sensitivities are determined in advance through a preliminary reconstruction process using non-diffusion-weighted data. These pre-determined coil sensitivities are then applied during the diffusion-weighted image reconstruction, enabling accurate image formation without requiring real-time field homogeneity adjustments.

Inventive Principle:
Principle #10Preliminary action

2Measurement precision

If multiple diffusion-encoding gradients with varying strengths and directions are applied, then diffusion contrast and tissue characterization are improved, but acquisition time increases

Engineering Contradiction:
Improvediffusion contrastVSAvoidacquisition time
Core Design Contradiction:
Measurement precisionVSLoss of time

Solution Approach 1:

Multiple diffusion-encoding gradients with different strengths and directions are applied in continuous succession without interruption to the imaging sequence. This continuous application of varying gradients maintains diffusion contrast quality while minimizing gaps that would extend total acquisition time.

Inventive Principle:
Principle #20Continuity of useful action

Solution Approach 2:

The diffusion-encoding gradients are applied in a periodic pattern, cycling through different gradient strengths and directions in a systematic sequence. This periodic structure allows for efficient sampling of diffusion information while maintaining a predictable and optimized acquisition timeline.

Inventive Principle:
Principle #19Periodic action

3Area of stationary object

If contiguous cross-sectional slices are acquired to cover a volume, then gap-free volume coverage is achieved, but acquisition complexity and processing time increase

Engineering Contradiction:
Improvevolume coverageVSAvoidprocessing complexity
Core Design Contradiction:
Area of stationary objectVSDevice complexity

Solution Approach 1:

Multiple contiguous cross-sectional slices are acquired and merged during the reconstruction process to form a complete three-dimensional volume. The coil sensitivity maps and image data from all slices are combined using a unified reconstruction algorithm that processes them simultaneously, achieving gap-free coverage while managing complexity through integrated processing.

Inventive Principle:
Principle #5Merging (Combining)

4Measurement precision

If regularized nonlinear and linear inverse reconstruction processes are used, then image quality and spatial fidelity are improved, but computational complexity increases

Engineering Contradiction:
Improvespatial fidelityVSAvoidcomputational complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The reconstruction process is divided into two stages: a preliminary nonlinear inverse reconstruction that determines coil sensitivities and basic image parameters, followed by a linear inverse reconstruction that refines the images with regularization. This staged approach breaks down the computationally intensive task into manageable steps, improving spatial fidelity while controlling overall complexity.

Inventive Principle:
Principle #10Preliminary 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 approach results in high-quality, gap-free volume coverage of MR images with improved signal-to-noise ratio, reduced acquisition time, and increased spatial fidelity, making it suitable for clinical applications without susceptibility artefacts and motion-induced errors.

Implementation Method 1

a self-compensating set of strong magnetic field gradients

Methodology Applied
Scientific EffectMagnetic field gradients: Magnetic Field

Implementation Method 2

diffusion encoding of the MRI signal is commonly accomplished by a DW spin-echo sequence which comprises two radiofrequency pulses and a self-compensating set of strong magnetic field gradients

Methodology Applied
Scientific EffectDiffusion encoding: Diffusion

Implementation Method 3

an initial excitation radiofrequency pulse, a first interval with a duration of half the spin-echo time TE/2, a refocusing radiofrequency pulse

Methodology Applied
Scientific EffectNuclear magnetic resonance: Magnetic Field

Implementation Method 4

In the absence of motion, i.e. without any spatial displacement of the water molecules, the nuclear spin moments of water protons completely refocus in the spin-echo signal

Methodology Applied
Scientific EffectSpin echo: Echo

Data Source

PatentEP4189416B1Method and device for acquiring and reconstructing a sequence of diffusion-weighted magnetic resonance images covering a volume
Publication Date: 2024.12.25 MAX PLANCK GESELLSCHAFT ZUR FOERDERUNG DER WISSENSCHAFTEN EV
  • EP4189416B1 patent drawingFigure 1~2
  • EP4189416B1 patent drawingFigure 3
  • EP4189416B1 patent drawingFigure 4~5

AI summary

A method for creating multiple sequences of diffusion-weighted magnetic resonance (MR) images of an object is described, wherein each of said sequences of MR images represents the same series of contiguous cross-sectional slices covering a volume of the object. The method comprises (a) providing multiple sequences of sets of image raw data being collected with the use of at least one radiofrequency receiver coil of a magnetic resonance imaging device, wherein each set of image raw data includes data samples being generated with a combined diffusion-weighted spin-echo and single-shot stimulated-echo sequence with diffusion-encoding gradients, (b) subjecting a sequence of sets of image raw data generated with diffusion-encoding gradients of zero strength or lower strength to a regularized nonlinear inverse reconstruction process to provide a sequence of coil sensitivities and MR images with no or lower diffusion weighting, and (c) subjecting all se- quences of sets of image raw data with diffusion-encoding gradients of zero or lower strength as well as of higher strength to a regularized linear inverse reconstruction process, to provide a sequence of MR images with no or lower diffusion weighting and a sequence of MR images with higher strength, each of the MR images representing one of the cross-sectional slices and being created by using the sensitivity of the at least one receiver coil determined in step (b) for the same cross-sectional slice and in dependency on a difference between a current image content estimation and an image content estimation of a neighboring cross-sectional slice. Furthermore, an MRI device for creating a sequence of diffusion-weighted MR images of an object is described.