Method and magnetic resonance apparatus for accelerate acquisition of magnetic resonance data of an examination object by undersampling the acquired magnetic resonance data

1. A method for acquiring magnetic resonance (MR) data from an examination object comprising:

providing control signals that comprise an acquisition pulse sequence to an MR data acquisition unit from a computer, and operating said MR data acquisition unit, comprising a gradient system, while an examination object is situated in said MR data acquisition unit, to acquire MR data from the examination object using the control signals comprising the acquisition pulse sequence, including operating said gradient system with said control signals so as to generate a readout gradient in a readout direction and a plurality of phase coding gradient pulses in a phase coding direction that is perpendicular to said readout direction;

via said computer, entering the acquired MR data into an electronic memory organized as two dimensional k-space comprising a plurality of k-space lines proceeding in a k-space direction corresponding to said readout direction, by entering said MR data along a k-space trajectory that proceeds through respective data entry points in said k-space lines by undersampling at least some of said lines by entering said MR data into fewer than all of the data entry points respectively therein, and, in an acquisition section proceeding through an entirety of k-space in said k-space direction corresponding to said readout direction, implementing a plurality of jumps between respective k-space lines, in a k-space direction corresponding to said phase coding direction, said jumps being produced by said phase coding gradient pulses, said k-space direction corresponding to said readout direction being perpendicular to said k-space direction corresponding to said phase coding direction, thereby producing an undersampled data set in said memory organized as k-space; and

making said undersampled data set available in electronic form as a data file from said electronic memory organized as k-space.

2. A method as claimed in claim 1 comprising activating said phase coding gradient pulses to cause said jumps to occur between adjacent k-space lines.

3. A method as claimed in claim 1 comprising selecting said k-space trajectory to cause, at least in a localized neighborhood of data entry points in k-space, a distribution of said data entry points to exceed a predetermined uniform distribution level.

4. A method as claimed in claim 1 comprising generating said k-space trajectory in said computer as a random k-space trajectory that causes said jumps to occur randomly in said random acquisition trajectory and, in said computer, automatically testing said random k-space trajectory against at least one distribution criterion, and using said random k-space trajectory to acquire said MR data from said examination object only when said testing shows that said random k-space trajectory fulfills said at least one distribution criterion.

5. A method as claimed in claim 4 comprising using a homogeneity criterion as said at least one distribution criterion, and, with said homogeneity criterion, testing whether said random k-space trajectory produces a distribution of data entry points in k-space into which said MR data are actually entered, dependent on a degree of said undersampling.

6. A method as claimed in claim 5 comprising selecting said homogeneity criterion to require a density distribution in k-space selected from the group consisting of a first density distribution comprising a density reducing radially from a center of k-space, and a second density distribution comprising a density that decreases along a line in said k-space direction corresponding to said phase coding direction, that includes a center of k-space.

7. A method as claimed in claim 5 wherein said first density distribution reduces radially from said center of k-space according to one divided a polynomial of a distance from the center of k-space, and wherein said second density distribution decreases along said line according to one divided by polynomial of a distance from the center of k-space along said k-space direction corresponding to the phase coding direction.

8. A method as claimed in claim 7 wherein said polynomial for each of said first and second density distributions is r n , wherein r is a radial distance from the center of k-space or a distance along said line from the center of k-space in the k-space direction corresponding to the phase coding direction, and wherein n is in a range between one and three.

9. A method as claimed in claim 5 comprising testing the density distribution in k-space against said homogeneity criterion by dividing k-space into a plurality of segments, each of said segments having a density therein of data entry points in which said MR data are entered, and comparing the respective densities in said segments each with at least one reference value that is determined from said density distribution, and wherein said homogeneity criterion is considered as being fulfilled when the density in a respective segment does not deviate from said reference value by more than a predetermined deviation amount.

10. A method as claimed in claim 4 comprising using, as said distribution criterion, a neighborhood criterion that compares a number of data entry points respectively filled with said MR data in adjacent lines in k-space.

11. A method as claimed in claim 10 comprising requiring, for fulfillment of said neighborhood criteria, said number of data entry points in said adjacent lines that are filled with said MR data to deviate from each other by no more than a predetermined proportion.

12. A method as claimed in claim 11 comprising selecting said proportion from the group consisting of a range between 10% and 30%, and a proportion that is dependent on a degree of said undersampling.

13. A method as claimed in claim 4 comprising determining said random acquisition trajectory in said computer using a random walk algorithm.

14. A method as claimed in claim 1 comprising activating said phase coding gradient pulses to produce at least four of said jumps in each acquisition section.

15. A method as claimed in claim 1 comprising employing, as said acquisition trajectory, an acquisition trajectory that causes a line containing a center of k-space to be completely filled with said MR data at all data entry points of said line, or that causes a line closest to a center of k-space to be completely filled with said MR data at all data entry points of said line closest to said center.

16. A method as claimed in claim 1 comprising operating said MR data acquisition unit to acquire said undersampled data set in one of a plurality of chronologically successive acquisitions of MR data from said examination object and, in acquisitions of MR data other than said one of said acquisitions, acquiring said MR data from said examination object with a further undersampling by leaving random lines in k-space completely empty of MR data, each of said acquisitions with said further undersampling producing a respective further undersampled MR data set and providing said undersampled data set and said further undersampled data sets to an image reconstruction computer and, in said image reconstruction computer, reconstructing an image of said examination object from said undersampled data set and said further undersampled data sets, using time as a third dimension in addition to said readout direction and said phase coding direction.

17. A method as claimed in claim 1 comprising selecting said undersampling to produce an acceleration factor associated with the acquisition of said MR data in a range between 2 and 4.

18. A method as claimed in claim 1 comprising operating said MR data acquisition unit to activate, as said readout gradient, a readout gradient pulse during said acquisition section configured to reduce a bandwidth in a region of k-space closest to a center of k-space.

19. A method as claimed in claim 18 comprising reducing said bandwidth by reducing a gradient field strength of said readout gradient pulse.

20. A method as claimed in claim 1 comprising back-calculating MR data that is acquired during said jumps.

21. A method as claimed in claim 1 comprising calculating a phase difference for said MR data during jumps of a repeatedly acquired data entry point in k-space, and using the calculating of phase difference as an indicator of movement of said examination object.

22. A magnetic resonance (MR) apparatus comprising:

an MR data acquisition unit comprising a gradient system;

an electronic memory organized as two-dimensional k-space comprising a plurality of k-space lines proceeding in a k-space direction corresponding to said readout direction;

a control computer configured to provide control signals that comprise an acquisition pulse sequence to said MR data acquisition unit from a computer, and to operate the MR data acquisition unit, while an examination object is situated in said MR data acquisition unit, to acquire MR data from the examination object using the control signals comprising the acquisition pulse sequence operating said gradient system with said control signals so as to generate a readout gradient in a readout direction and a plurality of phase coding gradient pulses in a phase coding direction that is perpendicular to said readout direction;

said control computer being configured to enter the acquired MR data into said electronic memory by entering said MR data along a k-space trajectory that proceeds through respective data entry points in said k-space lines by undersampling at least some of said lines by entering said MR data into fewer than all of the data entry points respectively therein, and, in an acquisition section proceeding through an entirety of k-space in said k-space direction corresponding to said readout direction, to implement a plurality of jumps between respective k-space lines, in a k-space direction corresponding to said phase coding direction, said jumps being produced by said phase coding gradient pulses, said k-space direction corresponding to said readout direction being perpendicular to said k-space direction corresponding to said phase coding direction, thereby producing an undersampled data set in said memory organized as k-space; and

said control computer being configured to make said undersampled data set available in electronic form as a data file from said electronic memory organized as k-space.