System and method for magnetic resonance imaging an object via a stochastic optimization of a sampling function

1. A system for magnetic resonance imaging an object via a stochastic optimization of a K-space sampling function comprising:

a magnet assembly operative to acquire MR data from the object; and

a controller operative to:

acquire a first MR data set using the magnet assembly;

select the K-space sampling function from a plurality of sampling function candidates using the stochastic optimization, wherein the K-space sampling function includes an acquisition trajectory along which the magnet assembly scans through K-space; and

acquire a second MR data set from the object using the magnet assembly by acquiring positions in the K-space according to the acquisition trajectory.

2. The system of claim 1 , wherein the stochastic optimization includes:

determining a K-space energy density distribution based at least in part on the first MR data set; and

generating the plurality of sampling function candidates based at least in part on the K-space energy density distribution.

3. The system of claim 1 , wherein the stochastic optimization further includes:

determining a parallel imaging sensitivity operator based at least in part on the first MR data set;

generating an evaluation function based at least in part on the parallel imaging sensitivity operator; and

comparing two or more of the sampling function candidates of the plurality to each other based at least in part on the evaluation function.

4. The system of claim 3 , wherein generation of the evaluation function is further based at least in part on a PSF Incoherence metric.

5. The system of claim 3 , wherein the controller is further operative to:

determine a reduction factor based at least in part on the evaluation function.

6. The system of claim 1 , wherein the controller is further operative to:

acquire a third MR data set from a different object using the magnet assembly based at least in part on the K-space sampling function.

7. The system of claim 1 , wherein the stochastic optimization is based at least in part on one of simulated annealing and genetic evolution.

8. A method for magnetic resonance imaging an object via a stochastic optimization of a K-space sampling function comprising:

acquiring a first MR data set from the object using a magnet assembly;

selecting the K-space sampling function from a plurality of sampling function candidates using the stochastic optimization via a controller, wherein the K-space sampling function includes an acquisition trajectory along which the magnet assembly scans through K-space; and

acquiring a second MR data set from the object using the magnet assembly by acquiring positions in the K-space according to the acquisition trajectory.

9. The method of claim 8 , wherein the stochastic optimization includes:

determining a K-space energy density distribution based at least in part on the first MR data set; and

generating the plurality of sampling function candidates based at least in part on the K-space energy density distribution.

10. The method of claim 8 , wherein the stochastic optimization further includes:

determining a parallel imaging sensitivity operator based at least in part on the first MR data set;

generating an evaluation function based at least in part on the parallel imaging sensitivity operator; and

comparing two or more of the sampling function candidates of the plurality to each other based at least in part on the evaluation function.

11. The method of claim 10 , wherein the selected K-space sampling function is the sampling function candidate with the best performance as determined by the evaluation function.

12. The method of claim 10 , wherein generation of the evaluation function is further based at least in part on a PSF Incoherence metric.

13. The method of claim 10 further comprising:

determining a reduction factor based at least in part on the evaluation function via the controller.

14. The method of claim 8 , wherein the stochastic optimization is based at least in part on one of simulated annealing and genetic evolution.

15. The method of claim 8 further comprising:

acquiring a third MR data set from a different object using the magnet assembly based at least in part on the K-space sampling function.

16. A non-transitory computer readable medium storing instructions configured to adapt a controller to:

acquire a first MR data set from an object using a magnet assembly of a magnetic resonance imaging system;

select a K-space sampling function from a plurality of sampling function candidates using a stochastic optimization, wherein the K-space sampling function includes an acquisition trajectory along which the magnet assembly scans through K-space; and

acquire a second MR data set from the object using the magnet assembly by acquiring positions in the K-space according to the acquisition trajectory.

17. The non-transitory computer readable medium of claim 16 , wherein the stochastic optimization includes:

determining a K-space energy density distribution based at least in part on the first MR data set; and

generating the plurality of sampling function candidates based at least in part on the K-space energy density distribution.

18. The non-transitory computer readable medium of claim 16 , wherein the stochastic optimization includes:

determining a parallel imaging sensitivity operator based at least in part on the first MR data set;

generating an evaluation function based at least in part on the parallel imaging sensitivity operator; and

comparing two or more of the sampling function candidates of the plurality to each other based at least in part on the evaluation function.

19. The non-transitory computer readable medium of claim 18 , wherein generating an evaluation function is further based at least in part on a PSF Incoherence metric.

20. The non-transitory computer readable medium of claim 16 , wherein the stochastic optimization is based at least in part on one of simulated annealing and genetic evolution.