Magnetic Resonance Elastography Stiffness Map Inversion
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Magnetic resonance elastography (MRE) techniques face challenges in accurately estimating tissue stiffness due to noise amplification, particularly around edges, which affects the diagnosis of diseases associated with stiffness changes.
Innovation Solution
An unconstrained optimization method is employed to reduce noise from MRE data while exploring the sparsity of the stiffness map in a sparsifying transform domain, such as the wavelet or Fourier transform domain, using a cost function that minimizes noise and promotes sparsity, thereby improving stiffness estimation accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional inversion algorithms based on the Helmholtz equation are used to obtain stiffness maps from MRE data, then the stiffness estimation can be obtained, but noise amplification occurs due to the Laplacian operation which deteriorates the measurement precision
Solution Approach 1:
The patent changes the mathematical approach from direct Helmholtz inversion to an optimization-based inversion that minimizes a cost function. This parameter change in the inversion methodology allows noise suppression while maintaining stiffness estimation accuracy, directly resolving the contradiction between measurement precision and noise amplification.
Solution Approach 2:
The patent introduces an intermediate optimization process that acts as a mediator between the raw MRE data and the final stiffness map. This intermediate step using cost function minimization with regularization terms filters out noise amplification while preserving the essential stiffness information, thereby resolving the contradiction.
2Ease of manufacture
If Local-frequency estimation (LFE) is used for inversion, then the processing can be simplified, but estimation errors occur around edges which deteriorates the manufacturing precision
Solution Approach 1:
The patent changes the inversion methodology from LFE to an optimization-based approach that minimizes a cost function with appropriate regularization. This parameter change maintains computational tractability while significantly improving edge accuracy in the stiffness map, resolving the contradiction between processing simplicity and edge precision.
3Reliability
If standard MRE inversion techniques are used, then the stiffness map can be generated, but the reliability deteriorates in the presence of severe noise corruption
Solution Approach 1:
The patent introduces an intermediate optimization process with regularization terms that acts as a mediator to filter noise corruption while preserving reliable stiffness information. This intermediate step significantly improves reliability in noisy conditions compared to direct inversion techniques.
Solution Approach 2:
The optimization-based inversion uses iterative feedback mechanisms where the cost function is minimized through repeated adjustments, allowing the system to converge to a reliable stiffness estimate even in the presence of severe noise corruption. This feedback loop enhances reliability by continuously refining the estimate.
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 significantly enhances the accuracy of tissue stiffness estimation in noisy conditions, facilitating more reliable disease diagnosis across various tissues and organs, including challenging cases like the lung, by reducing noise and preserving sharp transitions in stiffness maps.
Implementation Method 1
MRE is a phase-contrast magnetic resonance technique in which the shear stiffness of soft tissues can be estimated
Data Source
AI summary
Systems and methods for magnetic resonance elastography (MRE) are disclosed. In one embodiment, MRE data corresponding to mechanical waves in tissue of interest of a subject is acquired. The MRE data is associated with stiffness of the tissue. The method also includes generating, based on the MRE data, a stiffness map representing stiffness of the tissue. Generating the stiffness map includes performing an unconstrained optimization cost function that is configured to reduce noise in the acquired MRE data and achieve inversion of the reduced-noise data.