Bayesian QSM Reconstruction with Complex Data Fidelity
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current quantitative susceptibility mapping (QSM) techniques face limitations in accurately reconstructing magnetic susceptibility maps due to poor signal-to-noise ratio in regions of high susceptibility, leading to blocky or pixelated images and severe artifacts, especially in areas with low contrast to noise ratio and high susceptibility inhomogeneity.
Innovation Solution
The method involves recasting the data fidelity term in the complex MRI signal domain, using a nonlinear complex term instead of a quadratic one, and incorporating anatomical prior information through joint entropy regularization to improve the correlation with known anatomy, while also employing a Laplace operator for background field removal and iterative weighting to account for spatially varying noise.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional QSM techniques are used to reconstruct magnetic susceptibility maps, then the reconstruction process can be completed, but the signal-to-noise ratio deteriorates in regions of high susceptibility leading to blocky or pixelated images and severe artifacts
Solution Approach 1:
The patent segments the reconstruction process into multiple iterations with different weighting factors. In each iteration, regions with different susceptibility characteristics are treated differently through spatially varying weights, allowing high-susceptibility regions to be processed with appropriate noise suppression while maintaining accuracy in other regions.
Solution Approach 2:
The patent applies local quality by using spatially varying weighting factors that adapt to local susceptibility characteristics. The weighting scheme adjusts the contribution of different regions based on their specific noise levels and susceptibility values, ensuring optimal reconstruction quality for each local area rather than applying a uniform approach.
2Measurement precision
If conventional QSM techniques are used, then susceptibility mapping can be performed, but severe artifacts appear in areas with low contrast to noise ratio and high susceptibility inhomogeneity
Solution Approach 1:
The patent incorporates feedback mechanisms through iterative reconstruction where the weighting factors are updated based on the current reconstruction quality and noise characteristics. The algorithm continuously adjusts the weighting scheme based on feedback from previous iterations, suppressing artifacts in low contrast-to-noise regions while preserving genuine susceptibility variations.
Solution Approach 2:
The patent changes parameters dynamically during reconstruction by adjusting weighting factors based on local susceptibility inhomogeneity and noise characteristics. The weighting scheme adapts parameters such as regularization strength and data fidelity weights to match local conditions, reducing artifacts in challenging regions while maintaining overall reconstruction accuracy.
3Productivity
If standard data fidelity terms are used in QSM reconstruction, then the reconstruction can proceed, but blocky or pixelated images result due to poor signal-to-noise ratio handling
Solution Approach 1:
The patent transforms the static data fidelity term into a dynamic one that evolves through multiple iterations. The weighting factors are updated iteratively based on the current reconstruction and estimated noise levels, allowing the reconstruction process to adapt and improve image quality progressively without sacrificing computational efficiency.
Solution Approach 2:
The patent performs preliminary actions by pre-calculating weighting factors based on initial noise estimates and susceptibility maps before the main reconstruction loop. This preliminary preparation allows the iterative process to focus on refinement rather than starting from scratch, improving both efficiency and final image quality.
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 improved visualization of brain structures, reduced artifacts, and enhanced accuracy in assessing lesions and contrast agent biodistribution, providing a more accurate and detailed magnetic susceptibility map.
Implementation Method 1
Magnetic susceptibility is a physical property intrinsic to a material and measures the amount of magnetization induced in that material when placed in an external magnetic field such as the main magnetic field, B0, of an MRI scanner
Implementation Method 2
measures the amount of magnetization induced in that material when placed in an external magnetic field
Implementation Method 3
Tissue local magnetic field bL relative to, scaled to and along the main magnetic field B0 in image space (referred to as r-space) can be modeled as the convolution of the dipole kernel d(r)=(1⁄4π)(3 cos2 θ−1)/|r|3 with tissue volumetric susceptibility distribution χ(r)
Data Source
AI summary
Exemplary quantitative susceptibility mapping methods, systems and computer-accessible medium can be provided to generate images of tissue magnetism property from complex magnetic resonance imaging data using the Bayesian statistical approach. The likelihood is constructed directly using the complex data. A prior is constructed from matching structures or information content in known morphology. The quantitative susceptibility map can be determined by, e.g., maximizing the posterior. Thus, according to the exemplary embodiment, system, method and computer-accessible medium can be provided for determining information associated with at least one structure. Using such exemplary embodiment, it is possible to receive signals associated with the structure(s), where the signals can include complex data that is in the complex domain.