Analysis of nuclear magnetic resonance spectra with non-stationary peaks

• US 10,139,461 B2
• Filed: 03/21/2017
• Issued: 11/27/2018
• Est. Priority Date: 06/14/2016
• Status: Active Grant

• Current Assignee: Aspect Imaging Ltd.
• Original Assignee: Aspect Imaging Ltd.
• Inventors: Cohen, Tal, Rapoport, Uri
• Primary Examiner(s): Fuller, Rodney E
• Application Number:
  US15/465,317
• Publication Number:
  US 20170356974A1
• Time in Patent Office: 616 Days
• Field of Search: 324307
• CPC Class Codes: G01N 24/08 by using nuclear magnetic resonance G01N24/12 takes precedence
• CPC Class Codes: G01R 33/0064 comprising means for performing simulations, e.g. of the magnetic variable to be measured
• CPC Class Codes: G01R 33/12 Measuring magnetic properties of articles or specimens of solids or fluids involving magnetic resonance G01R33/20
• CPC Class Codes: G01R 33/46 NMR spectroscopy
• CPC Class Codes: G01R 33/4625 Processing of acquired signals, e.g. elimination of phase errors, baseline fitting, chemometric analysis
• CPC Class Codes: G01R 33/485 based on chemical shift information [CSI] or spectroscopic imaging, e.g. to acquire the spatial distributions of metabolites
• CPC Class Codes: G06F 17/10 Complex mathematical operations function generation by table look-up G06F1/03; evaluation of elementary functions by calculation G06F7/544

Abstract

A method of determining a NMR prediction result of a sample is provided. The method can include receiving a NMR spectrum of the sample and/or identifying a section of a ppm range in the NMR spectrum having a non-stationary peak. The method can include determining a modified data point for the NMR spectrum based on data points in the identified section. The modified data point can be determined such that the modified data point is a weighted average value of the data points in the identified section in the NMR spectrum. The method can include replacing the identified section in the NMR spectrum with the modified data point for the NMR spectrum to determine a modified NMR spectrum. The method can include determining the NMR prediction result of the sample based on the modified NMR spectrum and a calibration vector (e.g., using a partial least square (PLS) analysis).

Claims

1. A method of determining a NMR prediction result of a sample, the method comprising:

receiving a plurality of NMR spectra and a corresponding plurality of predetermined measured parameters, wherein each of the plurality of NMR spectra corresponds to one of the plurality of predetermined measured parameters;

identifying a section of a ppm range in each of the plurality of NMR spectra having a non-stationary peak;

determining a modified data point for each of the plurality of NMR spectra based on its respective data points in the identified section of the ppm range such that the modified data point being a weighted average value of the respective data points in the identified section;

replacing the identified section of the ppm range in each of the plurality of NMR spectra with its respective determined modified data point to generate a corresponding plurality of modified NMR spectra;

determining a modified calibration vector based on the plurality of modified NMR spectra and the plurality of predetermined measured parameters;

receiving a measured NMR spectrum of the sample;

identifying a section of a ppm range in the measured NMR spectrum having a non-stationary peak;

determining a modified data point for the measured NMR spectrum based on data points in the identified section in the measured NMR spectrum such that the modified data point for the measured NMR spectrum being a weighted average value of the data points in the identified section in the measured NMR spectrum;

replacing the identified section in the measured NMR spectrum with the modified data point for the measured NMR spectrum to determine a modified measured NMR spectrum; and

determining the NMR prediction result of the sample based on the modified measured NMR spectrum and the modified calibration vector.

2. The method of claim 1, wherein the modified calibration vector and the NMR prediction result of the sample are further determined by applying at least one of:

partial least square (PLS) analysis, principal component regression (PCR) analysis, neutral network analysis, or any combination thereof to the plurality pf modified NMR spectra and the modified measured NMR spectrum, respectively.

3. The method of claim 1, further comprising adding to each of the plurality of modified NMR spectra at least one additional data point that represents additional data and adding to the modified measured NMR spectrum corresponding at least one additional data point.

4. The method of claim 3, wherein the additional data point comprises at least one of:

additional measurements, measurement conditions and external data.

5. A system for determining a NMR prediction result of a sample, the system comprising:

an nuclear magnetic resonance (NMR) device to generate at least one NMR spectrum of the sample in a fixed measurement volume; and

at least one processing unit to;

receive a plurality of NMR spectra and a corresponding plurality of predetermined measured parameters, wherein each of the plurality of NMR spectra corresponds to one of the plurality of predetermined measured parameters;

determine a modified data point for each of the plurality of NMR spectra based on data points in an identified section of a ppm range in respective NMR spectrum, identified as having a non-stationary peak, such that the modified data point being a weighted average value of the respective data points in the identified section;

replace the identified section of the ppm range in each of the plurality of NMR spectra with its respective determined modified data point to generate the corresponding plurality of modified NMR spectra;

determine a modified calibration vector based on the plurality of modified NMR spectra and the plurality of predetermined measured parameters;

receive a measured NMR spectrum of the sample;

determine a modified data point for the measured NMR spectrum based on data points in an identified section of a ppm range in the measured NMR spectrum, identified as having a non-stationary peak, such that the modified data point for the measured NMR spectrum being a weighted average value of the data points in the identified section in the measured NMR spectrum;

replace the identified section in the measured NMR spectrum with the modified data point for the measured NMR spectrum to determine a modified measured NMR spectrum; and

determine the NMR prediction result of the sample based on the modified measured NMR spectrum and the modified calibration vector.

6. The system of claim 5, wherein the at least one processing unit is further to identify the section of the ppm range in each of the plurality of NMR spectra and in the measured NMR spectrum having the non-stationary peak.

7. The system of claim 5, wherein the modified calibration vector and the NMR prediction result of the sample are further determined by applying at least one of:

partial least square (PLS) analysis, principal component regression (PCR) analysis, neutral network analysis, or any combination thereof to the plurality pf modified NMR spectra and the modified measured NMR spectrum, respectively.

8. The system of claim 5, wherein each of the plurality of modified NMR spectra comprises at least one additional data point that represents additional data and the modified measured NMR spectrum comprises corresponding at least one additional data point.

9. The system of claim 8, wherein the additional data point comprises at least one of:

additional measurements, measurement conditions and external data.