NMR Proton Density Inversion via Global Optimization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current nuclear magnetic resonance (NMR) technologies face challenges in accurately determining proton density distributions from porous media due to noise and non-uniqueness in inversion techniques, leading to local minima issues and ambiguity in solutions.
Innovation Solution
A computer-implemented method using a global optimization algorithm, such as simulated annealing, to invert NMR data and parameterize the proton density distribution with non-linear basis functions, ensuring the determination of absolute minimum solutions and reducing noise impact.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional inversion techniques are used to determine proton density distribution from NMR data, then the process is computationally simpler, but the solution accuracy deteriorates due to noise and local minima issues
Solution Approach 1:
The patent transforms the inversion problem from a direct mathematical inversion to an optimization problem by parameterizing the proton density distribution as a sum of exponential functions with parameters (amplitudes, relaxation times). This parameterization allows the use of global optimization algorithms that can navigate the solution space more effectively, avoiding local minima and improving accuracy despite increased computational complexity.
2Reliability
If global optimization algorithms are used to invert NMR data, then solution accuracy improves by avoiding local minima, but computational time increases
Solution Approach 1:
The patent segments the proton density distribution into multiple exponential components, each characterized by specific parameters. This segmentation allows the global optimization algorithm to independently optimize each component's parameters, making the overall optimization process more manageable and efficient while ensuring a unique, reliable solution that avoids local minima.
3Measurement precision
If noise is present in NMR data, then measurement robustness is maintained, but inversion reliability deteriorates due to ambiguity in solutions
Solution Approach 1:
The patent implements an iterative optimization process where the model predictions are continuously compared with the actual NMR data, and the parameters are adjusted based on the error feedback. This feedback mechanism allows the global optimization algorithm to converge to the optimal solution that best fits the noisy data, reducing ambiguity and improving inversion reliability despite the presence of noise.
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 provides accurate and reliable proton density distributions by overcoming local minima and noise-related ambiguities, offering precise characterization of porous media properties.
Implementation Method 1
Nuclear magnetic resonance systems that manipulate spins of molecules present in a porous media are known. These systems generally perform at least the following functionality to determine information related to the porous median and/or fluids contained therein: polarizing spins through static magnetic fields, manipulating the spins through radio frequency ('RF') pulses, and receiving the response of spins through RF signals emanating from the porous media.
Data Source
AI summary
A computer-implemented method enables a proton density distribution to be obtained. In one embodiment, the method comprises acquiring nuclear magnetic resonance data from porous media; inverting the nuclear magnetic resonance data via a global optimization algorithm to determine a proton density distribution within the porous media; and outputting the determined proton density distribution.