Pareto-optimal magnetic resonance data acquisition

• US 7,078,899 B2
• Filed: 05/17/2004
• Issued: 07/18/2006
• Est. Priority Date: 05/15/2003
• Status: Active Grant

• Current Assignee: Case Western Reserve University
• Original Assignee: Case Western Reserve University
• Inventors: Lewin, Jonathan S., Duerk, Jeffrey L., Dale, Brian M.
• Primary Examiner(s): Gutierrez, Diego
• Assistant Examiner(s): Fetzner, Tiffany A.
• Application Number:
  US10/847,224
• Publication Number:
  US 20050057249A1
• Time in Patent Office: 792 Days
• Field of Search: None
• US Class Current: 324/314
• CPC Class Codes: G01R 33/54 Signal processing systems, e.g. using pulse sequences ; Generation or control of pulse sequences; Operator console
• CPC Class Codes: G01R 33/543 Control of the operation of the MR system, e.g. setting of acquisition parameters prior to or during MR data acquisition, dynamic shimming, use of one or more scout images for scan plane prescription G01R33/546 takes precedence

Abstract

A magnetic resonance data acquisition method includes designating a plurality of parameters that are representative of conditions for acquiring data from a magnetic resonance apparatus, at least one of the parameters being variable; designating at least one objective function measuring the quality of the acquired magnetic resonance data; optimizing the at least one objective function using an optimization algorithm to find at least one set of optimum values for the parameters characterizing data acquisition; configuring the magnetic resonance apparatus with the parameters determined above, configuring by one set of the optimum values of the parameters, and instructing a magnetic resonance imaging apparatus to apply the field to the target of the data acquisition to acquire the data.

Claims

1. A magnetic resonance data acquisition method, comprising:

designating a set of MRI data acquisition parameters, wherein at least one of said designated parameters is variable such that multiple solution values to the set of parameters are possible;

designating at least two objective measures where each objective measure is representative of a quality of magnetic resonance data to be acquired, and wherein values of said objective measures are determined, at least in part, by one or more of said multiple solution values of said designated set of MRI data acquisition parameters;

applying a multi-objective optimization algorithm to said designated set of MRI data acquisition parameters in order to cause said designated set of MRI data acquisition parameters to optimize to at least one single solution set of values for said set of parameters, from a plurality of possible solution set values, such that said single solution set of values is Pareto-optimal with respect to all said designated objective measures; and

acquiring magnetic resonance data by configuring a magnetic resonance data acquisition apparatus to acquire the magnetic resonance data, based on, at least in part, said Pareto-optimal single solution set of values for said set of designated MRI data acquisition parameters.

2. The method of claim 1 wherein the MRI data acquisition parameters are selected from the group consisting of:

pulse sequence design parameters;

representation of a locus of a plurality of points to be sampled in k-space;

a repetition time (TR);

an echo time (TE);

a number of interleaves, views or repetitions;

a flip angle;

a coil diameter;

a fat suppression compensation;

a flow compensation;

a type of coil used; and

a combination thereof.

3. The method of claim 2, wherein the locus of the plurality of points is selected from the group consisting of:

a spiral trajectory, a radial trajectory, a rosette trajectory, a non-rectilinear trajectory, and a combination thereof.

4. The method of claim 3, wherein the spiral trajectory includes a substantially uniform-density spiral trajectory.

5. The method of claim 3, wherein the spiral trajectory includes a variable-density spiral (VDS) trajectory.

6. The method of claim 5, wherein the density of the variable-density spiral trajectory varies over a range comprising:

a first radial region;

a second radial region outside the first radial region and substantially concentric with the first radial region; and

a third radial region outside the second radial region and substantially concentric with each of the first and second radial regions.

7. The method of claim 6, wherein the density of the variable-density spiral in the second radial region is substantially lower than the density of the spiral in each of the first radial region and the third radial region.

8. The method of claim 3, wherein the spiral trajectory is represented by an integer parameter denoting the number of interleaves, and at least one real-number parameter denoting the radial density of the spiral as a function of the radial distance from the center of the spiral.

9. The method of claim 2, wherein a type of coil having said coil diameter is selected from the group consisting of:

a saddle coil, an opposed solenoid, and a combination thereof.

10. The method of claim 1, wherein at least one of the objective measures is selected from the group consisting of:

image acquisition time;

image acquisition speed;

aliasing energy;

off-resonance blurring;

flow-sensitivity;

contrast-to-noise ratio (CNR);

perceptual difference;

point-spread function main-lobe width;

engineering cost;

quantitative imaging precision;

an image quality measure;

a data acquisition process quality measure; and

a combination thereof.

11. The method of claim 1, further comprising designating at least one constraint to be imposed on the MRI data acquisition parameters.

12. The method of claim 11, wherein each of the at least one constraint is selected from the group consisting of:

a safety constraint, a hardware constraint, and a combination thereof.

13. The method of claim 12, wherein the safety constraint is selected from the group consisting of:

a gradient stimulation limit, a specific absorption rate (SAR) limit, and a combination thereof.

14. The method of claim 12, wherein the hardware constraint is selected from the group consisting of:

a maximum slew rate, a gradient amplitude, and a combination thereof.

15. The method of claim 1, wherein the multi-objective optimization algorithm is an evolutionary algorithm.

16. The method of claim 15, wherein the evolutionary algorithm is a genetic algorithm.

17. The method of claim 16, wherein the genetic algorithm selects an initial population randomly.

18. The method of claim 16, wherein the genetic algorithm implements a reproduction operator according to one of a group consisting of:

a tournament selection, a proportionate selection, a ranking selection, and a combination thereof.

19. The method of claim 18, wherein the proportionate selection is performed according to one of another group consisting of:

a roulette wheel selection (RWS), a stochastic roulette-wheel selection (SRWS), and a combination thereof.

20. The method of claim 16, wherein the genetic algorithm implements a crossover operator.

21. The method of claim 20, wherein the crossover operator, includes a single-site crossover.

22. The method of claim 20, wherein the crossover operator includes a multiple-site crossover.

23. The method of claim 20, wherein the crossover operator is selected from the group consisting of:

a linear crossover;

a bland crossover;

a simulated binary crossover;

a fuzzy recombination crossover;

a uni-modal normally-distributed crossover;

a simplex crossover;

a fuzzy connectives-based crossover;

an unfair average crossover; and

a combination thereof.

24. The method of claim 16, wherein the genetic algorithm implements a mutation operator.

25. The method of claim 24, wherein the mutation operator is selected from the group consisting of:

a random mutation;

a non-uniform mutation;

a normally-distributed mutation;

a polynomial mutation; and

a combination thereof.

26. The method of claim 16, wherein the genetic algorithm implements elitism.

27. The method of claim 26, wherein the elitism includes local elitism.

28. The method of claim 26, wherein the elitism includes global elitism.

29. The method of claim 26 wherein, in at least one of the generations, the genetic algorithm preserves a predetermined percentage of the population as elites.

30. The method of claim 29, wherein the percentage is approximately 10%.

31. A method of designing a k-space trajectory for magnetic resonance data acquisition, comprising:

a. designating a set of k-space trajectory parameters, wherein said set of k-space trajectory parameters are to be manipulated in order to generate a plurality of solution set values of said set of k-space trajectory parameters starting from an initial solution set of values;

b. designating at least two objective measures that are representative of a quality of magnetic resonance data to be acquired, wherein values of said objective measures are determined, at least in part, by one or more of said plurality of solution set values of said designated set of k-space trajectory parameters;

c. defining at least one termination condition for each objective measure;

d. applying a genetic algorithm to the designated set of k-space trajectory parameters and to, only for a first iteration, said initial solution set of values, in order to produce at least one genetically modified solution set of values of said set of k-space trajectory parameters;

e. evaluating said objective measures, representative of a quality of magnetic resonance data to be acquired, based on, at least in part, said genetically modified solution set of values; and

f. determining when said evaluated objective measures associated with said genetically modified solution set of values meets at least one of said defined termination conditions of at least one of the objective measures; and

g. repeating steps d–

f until said at least one of the termination conditions of at least one of the objective measures is met,

whereby the final result of step f, is a final set of genetically-modified parameter values of said set of k-space trajectory parameters and is Pareto-optimal with respect to all of said at least two objective measures which are representative of the quality of magnetic resonance data to be acquired.

32. The method of claim 31, wherein the k-space trajectory comprises a non-rectilinear trajectory.

33. The method of claim 31, wherein the k-space trajectory comprises a spiral trajectory.

34. The method of claim 33, wherein the k-space trajectory comprises a spiral trajectory with a substantially spiraling section and a substantially linear section.

35. The method of claim 33, wherein the spiral trajectory comprises interleaved spiral trajectories.

36. The method of claim 33, wherein the spiral trajectory comprises a variable-density spiral trajectory.

37. The method of claim 36, wherein the variable-density spiral trajectory has a lower density in at least one of a center section within k-space, an intermediate section within k-space, and an outer section within k-space.

38. The method of claim 37, wherein a density of the intermediate section within k-space is lower than a density of the center section within k-space and a density of the outer section within k-space.

39. The method of claim 36, wherein the center section within k-space, the intermediate section within k-space, and the outer section within k-space are arranged substantially concentrically around an origin of the k-space.

40. A method of designing a magnetic resonance pulse sequence, comprising:

a. designating a set of magnetic resonance pulse sequence parameters, wherein said set of magnetic resonance pulse sequence parameters are to be manipulated in order to generate a plurality of solution set values of said set of magnetic resonance pulse sequence parameters starting from an initial solution set of values;

b. designating at least two objective measures representative of a quality of magnetic resonance data to be acquired, wherein values of said objective measures are determined, at least in part, by one or more of said plurality of solution set values of said designated set of magnetic resonance pulse sequence parameters;

c. defining at least one termination condition for each objective measure;

d. applying a genetic algorithm to the designated set of magnetic resonance pulse sequence parameters and to, only for a first iteration, said initial solution set of values, in order to produce at least one genetically modified solution set of values of said set of magnetic resonance pulse sequence parameters;

e. evaluating said objective measures representative of a quality of magnetic resonance data to be acquired based on, at least in part, said genetically modified solution set of values; and

f. determining when said evaluated objective measures associated with said genetically modified solution set of values meets at least one of said defined termination conditions of at least one of the objective measures; and

g. repeating steps d-f, until at least one of the termination conditions of at least one of the objective measures is met,

whereby a final result of step f, is a final set of genetically-modified parameter values of said set of magnetic resonance pulse sequence parameters and is Pareto-optimal with respect to all of said at least two objective measures which are representative of the quality of magnetic resonance data to be acquired.

41. A system for generating a magnetic resonance pulse sequence, comprising:

an apparatus configured for;

designating a set of magnetic resonance pulse sequence parameters, wherein at least one of said designated parameters is variable such that multiple solution values to the set of parameters are possible,

designating at least two objective measures where each objective measure is representative of a quality of magnetic resonance data to be acquired, and wherein values of said objective measures are determined, at least in part, by one or more of said multiple solution values of said designated set of magnetic resonance pulse sequence parameters, and

applying a multi-objective optimization algorithm to said designated set of magnetic resonance pulse sequence parameters in order to cause said designated set of magnetic resonance pulse sequence parameters to optimize to at least one single solution set of values for said set of parameters, from a plurality of possible solution set values, such that said single solution set of values is Pareto-optimal with respect to all said designated objective measures; and

a magnetic resonance pulse sequence generator receiving the Pareto-optimal single solution set of parameter values for said set of magnetic resonance pulse sequence parameters from said apparatus, and generating the pulse sequence specified by the single set of parameter values, in a magnetic resonance data acquisition.