Real-Time Diffusional Kurtosis Imaging via Linear Constrained Estimation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional diffusional kurtosis imaging (DKI) methods require substantial time for post-processing, are prone to errors due to noise and imaging artifacts, and struggle to provide physically and biologically plausible tensor estimates, especially in regions with complex fiber configurations.
Innovation Solution
A method and system for real-time DKI that uses a closed-form solution procedure to determine diffusional kurtosis, imposing constraints on diffusivity and kurtosis parameters through linear least squares, allowing for efficient estimation of orientation distribution functions and resolving fiber crossings, using a combination of diffusion and kurtosis tensors.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional DKI post-processing methods are used, then accurate tensor estimation can be achieved, but processing time becomes excessively long (1 hour or more)
Solution Approach 1:
The patent transforms the traditional nonlinear optimization problem into a linear system by changing the mathematical parameters and formulation approach. This allows the use of efficient linear algebra algorithms (such as singular value decomposition) instead of iterative nonlinear optimization, dramatically reducing computation time while maintaining accuracy in tensor and ODF estimation.
Solution Approach 2:
The patent replaces the computational mechanism of iterative nonlinear optimization with a direct linear algebraic solution. By substituting the computational approach from iterative numerical methods to closed-form linear solutions, the processing time is reduced from over an hour to near real-time while preserving measurement precision.
2Measurement precision
If q-space imaging techniques are used to obtain complete diffusion displacement probability, then accurate ODF can be achieved, but the number of required measurements and processing complexity increase substantially
Solution Approach 1:
The patent extracts and utilizes only the essential information needed for ODF estimation from the diffusion-weighted images, rather than requiring complete q-space sampling. By selecting specific gradient directions and b-values, the method obtains sufficient data for accurate ODF calculation while avoiding the excessive measurement requirements of full q-space imaging.
Solution Approach 2:
The patent creates a unified mathematical framework that handles both simple and complex fiber configurations using the same linear system approach. This universal method works across different tissue types and fiber architectures without requiring separate specialized procedures, reducing overall system complexity while maintaining accuracy.
3Measurement precision
If conventional DKI methods are used, then non-Gaussian diffusion can be characterized, but the results are prone to errors from noise and imaging artifacts
Solution Approach 1:
The patent performs preliminary corrections for noise and imaging artifacts before the main tensor and ODF estimation process. By addressing these confounding factors in advance through the linear system formulation, the method prevents them from propagating errors through the calculation, thereby improving reliability while maintaining accurate diffusion characterization.
Data Source
AI summary
Exemplary method, system, and computer-accessible medium can be provided for determining a measure of diffusional kurtosis by receiving data relating to at least one diffusion weighted image, and determining a measure of a diffusional kurtosis as a function of the received data using a closed form solution procedure. In accordance with certain exemplary embodiments of the present disclosure, provided herein are computer-accessible medium, systems and methods for, e.g., imaging in an MRI system, and, more particularly for facilitating estimation of tensors and tensor-derived measures in diffusional kurtosis imaging (DKI). For example, DKI can facilitate a characterization of non-Gaussian diffusion of water molecules in biological tissues. The diffusion and kurtosis tensors parameterizing the DKI model can typically be estimated via unconstrained least squares (LS) methods. In the presence of noise, motion, and imaging artifacts, these methods can be prone to producing physically and/or biologically implausible tensor estimates. The exemplary embodiments of the present disclosure can address at least this deficiency by formulating an exemplary estimation problem, e.g., as linearly constrained linear LS, where the constraints can ensure acceptable tensor estimates.