Systems and methods for acceleration of dictionary generation and matching in perfusion analysis

Patent metadata

Field	Value
Publication number	US-10859654-B2
Application number	US-201916351404-A
Country	US
Kind code	B2
Filing date	Mar 12, 2019
Priority date	Mar 13, 2018
Publication date	Dec 8, 2020
Grant date	Dec 8, 2020

Abstract

A method for determining quantitative parameters for dynamic contrast-enhanced MR data includes acquiring a set of contrast-enhanced MR data for a region of interest using a T1-weighted pulse sequence, generating at least one contrast concentration curve based on the set of contrast-enhanced MR data, accessing a comprehensive dictionary of contrast concentration curves and generating a grouped dictionary that has a plurality of groups based on the comprehensive dictionary. Each group includes a plurality of correlated contrast concentration curves and a group representative signal for the group. The method also includes comparing a contrast concentration curve with the group representative signal of each group to select a group, comparing the contrast concentration curve to the plurality of correlated contrast concentration curves in the selected group to identify a set of quantitative parameters for the concentration curve and generating a report including the set of quantitative parameter.

First claim

The invention claimed is:

1. A method for determining quantitative parameters for dynamic contrast-enhanced magnetic resonance (MR) data, the method comprising:
acquiring a set of contrast-enhanced MR data for a region of interest using a T1-weighted pulse sequence;
generating at least one contrast concentration curve based on the set of contrast-enhanced MR data;
accessing a comprehensive dictionary of contrast concentration curves;
generating a grouped dictionary based on the comprehensive dictionary, the grouped dictionary including a plurality of groups, wherein each group includes a plurality of correlated contrast concentration curves and a group representative signal that represents the group;
comparing at least one contrast concentration curve with the group representative signal of each group to select a group;
comparing the at least one contrast concentration curve to the plurality of correlated contrast concentration curves in the selected group to identify at least one quantitative parameter for the at least one concentration curve; and
generating a report including the at least one quantitative parameter.

2. The method according to claim 1 , wherein the set of contrast-enhanced MR data includes a signal intensity and wherein the at least one contrast concentration curve is generated based on the signal intensity of the set of contrast enhanced MR data.

3. The method according to claim 1 , wherein the T1-weighted pulse sequence is a fast spoiled gradient echo pulse sequence.

4. The method according to claim 1 , wherein the at least one quantitative parameter includes permeability, perfusion and blood flow.

5. The method according to claim 1 , wherein the at least one quantitative parameter is a set of quantitative parameters.

6. A method for determining quantitative parameters for dynamic contrast-enhanced magnetic resonance (MR) data, the method comprising:
acquiring a set of contrast-enhanced MR data for a region of interest using a T1-weighted pulse sequence;
generating at least one contrast concentration curve based on the set of contrast-enhanced MR data;
accessing a compressed dictionary, the compressed dictionary generated by performing a randomized singular value decomposition on a dictionary of concentration curves, the compressed dictionary comprising a series of singular values from the dictionary of concentration curves;
comparing at least one contrast concentration curve with the compressed dictionary to identify at least one quantitative parameter for the at least one contrast concentration curve; and
generating a report including the at least one quantitative parameter.

7. The method according to claim 6 , wherein the series of singular values are low rank approximations of true values of the concentration curves in the dictionary.

8. The method according to claim 6 , wherein the T1-weighted pulse sequence is a fast spoiled gradient echo pulse sequence.

9. The method according to claim 6 , wherein the at least one quantitative parameter includes permeability, perfusion and blood flow.

10. The method according to claim 6 , wherein the at least one quantitative parameter is a set of quantitative parameters.

11. A method for determining quantitative parameters for dynamic contrast-enhanced magnetic resonance (MR) data, the method comprising:
acquiring a set of contrast-enhanced MR data for a region of interest using a T1-weighted pulse sequence;
generating at least one measured contrast concentration curve based on the set of contrast-enhanced MR data;
providing at least one measured concentration curve to a neural network, the neural network trained using a dictionary of contrast concentration curves as a training data set for a dictionary matching process, each entry in the dictionary of contrast concentration curves having an associated set of quantitative parameters used to generate the dictionary entry, wherein the dictionary of contrast concentration curves is used to train the neural network to generate at least one quantitative parameter associated with the at least one measured concentration curve;
generating the at least one quantitative parameter for the at least one measured contrast concentration curve using the neural network; and
generating a report including the at least one quantitative parameter.

12. The method according to claim 11 , wherein the neural network generates the at least one quantitative parameter using a supervised learning model.

13. The method according to claim 11 , wherein the neural network generated the at least one quantitative parameter using an unsupervised learning model.

14. The method according to claim 11 , wherein the T1-weighted pulse sequence is a fast spoiled gradient echo pulse sequence.

15. The method according to claim 11 , wherein the at least one quantitative parameter includes permeability, perfusion and blood flow.

16. The method according to claim 11 , wherein the at least one quantitative parameter is a set of quantitative parameters.

17. A magnetic resonance imaging system comprising:
a magnet system configured to generate a polarizing magnetic field about at least a portion of a subject;
a magnetic gradient system including a plurality of magnetic gradient coils configured to apply at least one magnetic gradient field to the polarizing magnetic field;
a radio frequency (RF) system configured to apply an RF field to the subject and to receive magnetic resonance signals from the subject using a coil array; and
a computer system comprising a graphics processing unit, the computer system programmed to:
access a dictionary of concentration curves;
acquire a set of contrast-enhanced MR data for a region of interest using a T1-weighted pulse sequence;
generate at least one contrast concentration curve based on the set of contrast-enhanced MR data;
using the graphics processing unit, compare at least one contrast concentration curve with the dictionary to identify at least one quantitative parameter for the at least one contrast concentration curve, wherein the comparison is performed as a plurality of operations, each operation performed independently and simultaneously by the graphics processing unit; and
generate a report including the at least one quantitative parameter.

18. The system according to claim 17 , wherein the graphics processing unit is further configured to generate the dictionary of contrast concentration curves by simulating a plurality of contrast concentration curves.

19. The system according to claim 18 , wherein each contrast concentration curve in the dictionary is independent of the other contrast concentration curves in the dictionary and the contrast concentration curves in the dictionary are generated simultaneously using the graphics processing unit.

20. The system according to claim 18 , wherein the plurality of contrast concentration curves are generated using a pharmacokinetic model.

21. The method according to claim 17 , wherein the at least one quantitative parameter is a set of quantitative parameters.

Assignees

• Univ Case Western Reserve

Inventors

• Gulani Vikas
• Ghodasara Satyam
• Wright Katherine
• Seiberlich Nicole
• Griswold Mark A

Classifications

• A61B5/055

  involving electronic [EMR] or nuclear [NMR] magnetic resonance, e.g. magnetic resonance imaging · CPC title

• G01R33/5601 Primary

  involving use of a contrast agent for contrast manipulation, e.g. a paramagnetic, super-paramagnetic, ferromagnetic or hyperpolarised contrast agent · CPC title

• G01R33/56308

  Characterization of motion or flow; Dynamic imaging · CPC title

• G01R33/281

  Means for the use of in vitro contrast agents (G01R33/282 takes precedence; involving use of a contrast agent in MR imaging G01R33/5601; in vivo contrast agents A61K49/0002) · CPC title

• G01R33/5602 Primary

  by filtering or weighting based on different relaxation times within the sample, e.g. T1 weighting using an inversion pulse · CPC title