MRI Motion Correction via Neural Network Upsampling

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Magnetic Resonance Imaging (MRI) techniques are susceptible to subject motion, which prolongs the time required to acquire data and affects image quality, as existing motion compensation methods may lack robustness and efficiency.

Innovation Solution

A medical system employing an upsampling neural network to convert a lower resolution preliminary MRI to a higher resolution image, which is then used for motion correction in clinical k-space data, enabling both prospective and retrospective motion correction.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If conventional motion compensation methods are used, then motion correction can be performed, but the process lacks robustness and efficiency

Engineering Contradiction:

Improvemotion correction robustnessVSAvoidmotion correction efficiency

Core Design Contradiction:

ReliabilityVSProductivity

Solution Approach 1:

The patent applies preliminary action by acquiring a low-resolution preliminary MRI image before the main clinical scan. This preliminary image is then used to guide and correct motion during the subsequent high-resolution clinical imaging, improving both robustness and efficiency by preparing correction data in advance

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The patent uses an intermediary approach by introducing a low-resolution preliminary MRI image as a mediator between the scanning process and motion correction. This intermediary image serves as a reference to identify and correct motion artifacts in the final high-resolution image, resolving the contradiction between robustness and efficiency

Inventive Principle:

Principle #24Intermediary (Mediator)

2Measurement precision

If high resolution MRI is acquired, then image quality is improved, but acquisition time is prolonged

Engineering Contradiction:

Improveimage resolutionVSAvoidacquisition time

Core Design Contradiction:

Measurement precisionVSLoss of time

Solution Approach 1:

The patent applies segmentation by dividing the MRI acquisition into two parts: a low-resolution preliminary scan and a high-resolution clinical scan. The preliminary scan is acquired quickly at low resolution, while the clinical scan is acquired at high resolution with motion correction guidance, thus reducing total acquisition time while maintaining high image quality

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The patent uses preliminary action by obtaining a low-resolution preliminary image before the main high-resolution scan. This preliminary image provides motion information that guides the subsequent high-resolution acquisition, allowing faster acquisition by pre-planning the correction strategy

Inventive Principle:

Principle #10Preliminary action

3Loss of time

If low resolution preliminary image is used for motion correction, then acquisition time is reduced, but correction accuracy is compromised

Engineering Contradiction:

Improveacquisition timeVSAvoidmotion correction accuracy

Core Design Contradiction:

Loss of timeVSMeasurement precision

Solution Approach 1:

The patent uses an intermediary approach by introducing a low-resolution preliminary MRI image as a mediator that captures motion information. Although low-resolution, this intermediary image provides sufficient motion data to guide correction algorithms, achieving both time efficiency and adequate correction accuracy

Inventive Principle:

Principle #24Intermediary (Mediator)

Solution Approach 2:

The patent replaces traditional mechanical motion correction methods with a computational approach using neural networks. The neural network processes the low-resolution preliminary image to generate motion correction parameters, substituting complex mechanical correction systems with an intelligent computational model that achieves accuracy despite low input resolution

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

Data Source

PatentUS20240404131A1Motion correction using low resolution magnetic resonance images

Publication Date: 2024.12.05 KONINKLIJKE PHILIPS NV

AI summary

Described herein is a medical system (100, 300) comprising a memory (110) storing machine executable instructions (120) and an upsampling neural network (122). The upsampling neural network is configured to output an upsampled magnetic resonance image (130) with a second resolution in response to receiving a preliminary magnetic resonance image (126) with a first resolution which is lower than the second resolution. The execution of the machine executable instructions causes a computational system (104) to: receive (200) preliminary k-space data (124); reconstruct (202) the preliminary magnetic resonance image from the preliminary k-space data; receive (204) clinical k-space data (204); receive (206) the upsampled magnetic resonance image in response to inputting the preliminary magnetic resonance image into the upsampling neural network; and provide (208) a motion corrected magnetic resonance image (132) using the upsampled magnetic resonance image and the clinical k-space data.