A weighted-based method for t1 noise suppression in magnetic resonance

By using the sliding window method and histogram estimation combined with the nonlinear Logsitic function to generate a weight matrix in the nuclear magnetic resonance spectrum, and multiplying it point by point to suppress noise, the problem of slow t1 noise suppression speed and poor effect in the prior art is solved, and fast and efficient noise suppression and spectrum quality improvement are achieved.

CN117688296BActive Publication Date: 2026-07-24WUHAN ZHONGKE NIUJIN MAGNETIC RESONANCE TECH CO LTD
View PDF 1 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
WUHAN ZHONGKE NIUJIN MAGNETIC RESONANCE TECH CO LTD
Filing Date
2022-09-09
Publication Date
2026-07-24

AI Technical Summary

Technical Problem

Existing techniques for suppressing t1 noise in two-dimensional nuclear magnetic resonance spectra suffer from slow computation speed, difficulty in effectively preserving complex peak shapes and weak peaks, and a tendency to disrupt the relative relationships between spectral peaks, leading to a reduction in spectral quality.

Method used

By calculating the noise level of multidimensional nuclear magnetic resonance spectral data, a sliding window method and histogram estimation are used, combined with a nonlinear Logsitic function to generate a weight matrix, and point-by-point multiplication is performed to suppress noise, ensuring that the noise level reaches the target level while preserving weak peaks and relative relationships.

Benefits of technology

It achieves rapid and effective suppression of T1 noise, adapts to various peak types, improves spectrum quality, preserves weak peaks, does not destroy inter-peak relationships, has low computational cost, fast operation speed, and is suitable for TOCSY and HSQC spectra.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure HDA0003839776670000011
    Figure HDA0003839776670000011
  • Figure HDA0003839776670000012
    Figure HDA0003839776670000012
  • Figure HDA0003839776670000021
    Figure HDA0003839776670000021
Patent Text Reader

Abstract

This invention provides a weighted NMR t1 noise suppression method, comprising the following steps: S101, loading multidimensional NMR spectral data, denoted as matrix S(m,n); where m represents the number of points in the indirect dimension, i.e., the number of rows in matrix S of the multidimensional spectral data; n represents the number of points in the direct dimension, i.e., the number of columns in matrix S; S102, calculating the noise level of each point in matrix S, and taking the lowest value of the noise level in each column of matrix S as the target noise level σ of matrix S; S103, calculating the weight matrix W based on the target noise level σ and the noise level h of each point, multiplying matrix S and weight matrix W point by point to make the noise level of each column of data reach the target noise level σ, and the matrix obtained by multiplying matrix S and weight matrix W point by point is the spectrum after t1 noise suppression. This invention can effectively suppress the t1 noise level of two-dimensional spectra, effectively improve the spectrum quality, and has low computational load, fast operation speed, and also has good effect on spectra with baseline shift, showing strong adaptability.
Need to check novelty before this filing date? Find Prior Art

Citation Information

Patent Citations

  • CN111175335A