Resonance Frequency Determination Using Spectrum Matrix Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for determining resonance frequencies in magnetic resonance examinations are computationally intensive due to the need for vector operations and cross-correlation analysis, which requires significant computing time and compromises on quality.
Innovation Solution
The method employs vector-matrix multiplication by transforming the spectrum into a matrix representation and incorporating displacement into the spectrum matrix, reducing unnecessary computations and optimizing memory access, using submatrices to further minimize memory access and enhance CPU cache utilization.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If vector operations and cross-correlation analysis are used to determine resonance frequencies, then measurement precision is improved, but computing time increases significantly
Solution Approach 1:
The patent segments the spectrum into multiple segments and processes each segment separately to determine resonance frequencies. This segmentation allows the system to avoid computationally intensive full-spectrum cross-correlation operations while maintaining accurate frequency determination in each segment, thereby reducing overall computing time without sacrificing measurement precision.
Solution Approach 2:
The patent performs preliminary processing of the spectrum data before resonance frequency determination, including preprocessing steps that prepare the data for more efficient analysis. This preliminary action organizes and conditions the spectral data in advance, enabling faster subsequent processing and reducing the computational burden during the actual frequency determination phase.
2Measurement precision
If multiple model spectra are used for cross-correlation analysis, then measurement precision improves, but device complexity increases
Solution Approach 1:
By dividing the spectrum into segments and applying cross-correlation to each segment with appropriate model spectra, the patent reduces the overall computational complexity. Each segment requires fewer model comparisons than analyzing the entire spectrum at once, thereby lowering device complexity while maintaining precision through comprehensive segment-wise analysis.
Solution Approach 2:
The patent applies cross-correlation analysis selectively to individual spectrum segments rather than performing exhaustive analysis on the entire spectrum with all possible model spectra. This partial action approach uses only the necessary model spectra for each specific segment, reducing computational complexity while maintaining sufficient measurement precision for each localized analysis.
Data Source
AI summary
In a method and a magnetic resonance (MR) system for automated determination of the resonance frequency of a nucleus for magnetic resonance examinations, at least one MR signal is detected, and is Fourier-transformed into a spectrum composed of elements that can be represented as a vector. An analysis of the spectrum is conducted, wherein at least two cross-correlation coefficients of at least one model spectrum are determined by use of the measured spectrum. Prior to the analysis, a spectrum matrix having at least two vectors is determined from the spectrum, with each vector of the spectrum matrix being formed using all or some of the spectrum.