Method for ascertaining a deviation of at least one gradient field from a reference

Abstract

The disclosure relates to a method for ascertaining a deviation of at least one gradient field of a magnetic resonance system from a reference. The method includes providing at least one first image data set and one second image data set of a phantom with isotropic diffusion properties, recorded with a diffusion-weighted imaging sequence, wherein the first image data set and the second image data set are recorded with different diffusion-weightings along a gradient direction to be tested of the gradient field using the magnetic resonance system. The method further includes ascertaining a map of apparent diffusion coefficients from the image data sets for at least a portion of the image points of the image data sets. The method further includes comparing the apparent diffusion coefficients with the reference.

Claims

1. A method for ascertaining a deviation of at least one gradient field of a magnetic resonance system from a reference, the method comprising: providing a first image data set and a second image data set of a phantom with isotropic diffusion properties, recorded with a diffusion-weighted imaging sequence, wherein the first image data set and the second image data set are recorded with different diffusion-weightings along a gradient direction to be tested of the gradient field using the magnetic resonance system; ascertaining a map of apparent diffusion coefficients from the first image data set and the second image data set for at least a portion of image points of the first image data set and the second image data set; and comparing the apparent diffusion coefficients with reference values.

2. The method of claim 1, wherein the diffusion weightings are obtained by diffusion gradients along the gradient direction to be tested.

3. The method of claim 1, wherein the providing, the ascertaining, and the comparing are repeated for a second gradient direction and/or a third gradient direction.

4. The method of claim 1, wherein a first diffusion-weighted imaging sequence with which the first image data set has been recorded has a b-value in a range of 100 s/mm.sup.2 and 2000 s/mm.sup.2.

5. The method of claim 4, wherein a second diffusion-weighted imaging sequence with which the second image data set has been recorded has a b-value which differs by at least 100 s/mm.sup.2 from the b-value of the first diffusion-weighted imaging sequence.

6. The method of claim 1, wherein a second diffusion-weighted imaging sequence with which the second image data set has been recorded has a b-value which differs by at least 100 s/mm.sup.2 from a b-value of a first diffusion-weighted imaging sequence with which the first image data set has been recorded.

7. The method of claim 1, wherein, in the diffusion-weighted imaging sequence, diffusion gradients are arranged symmetrically round a refocusing pulse.

8. The method of claim 1, wherein diffusion gradients are configured eddy current-compensated.

9. The method of claim 1, further comprising: providing at least one additional image data set, wherein in a portion of the first image data set, the second image data set, and the at least one additional image data set, a polarity of diffusion gradients is swapped.

10. The method of claim 1, wherein the diffusion-weighted imaging sequence for the recording of the first image data set and the second image data set has a refocusing pulse, and wherein a slice thickness of an excited magnetization is greater than in an excitation pulse.

11. The method of claim 1, wherein the reference values are maps of apparent diffusion coefficients recorded at an earlier time point with the magnetic resonance system.

12. The method of claim 11, wherein the maps of the apparent diffusion coefficients are recorded with the diffusion-weighted imaging sequence.

13. The method of claim 1, wherein the maps of the apparent diffusion coefficients are normalized before the comparing of the apparent diffusion coefficients.

14. The method of claim 1, further comprising: providing a third image data set and a fourth image data set, wherein in a portion of the first image data set, the second image data set, the third image data set, and the fourth image data set, directions of phase encoding gradients and readout gradients are swapped.

15. A computer program product for a control apparatus for controlling a computer of a magnetic resonance system, wherein the computer program product, when executed by the control apparatus, is configured to cause the computer to: provide a first image data set and a second image data set of a phantom with isotropic diffusion properties, recorded with a diffusion-weighted imaging sequence, wherein the first image data set and the second image data set are recorded with different diffusion-weightings along a gradient direction to be tested of a gradient field using the magnetic resonance system; ascertain a map of apparent diffusion coefficients from the first image data set and the second image data set for at least a portion of image points of the first image data set and the second image data set; and compare the apparent diffusion coefficients with reference values.

16. The computer program product of claim 15, wherein the computer is an image generating unit or an evaluating unit of the magnetic resonance system.

17. A magnetic resonance system comprising: a control apparatus configured to: provide a first image data set and a second image data set of a phantom with isotropic diffusion properties, recorded with a diffusion-weighted imaging sequence, wherein the first image data set and the second image data set are recorded with different diffusion-weightings along a gradient direction to be tested of a gradient field using the magnetic resonance system; ascertain a map of apparent diffusion coefficients from the first image data set and the second image data set for at least a portion of image points of the first image data set and the second image data set; and compare the apparent diffusion coefficients with reference values.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

(1) Further advantages, features, and peculiarities of the present disclosure are disclosed in the following description of exemplary embodiments.

(2) FIG. 1 depicts an example of a magnetic resonance system.

(3) FIG. 2 depicts a phantom in a first embodiment.

(4) FIG. 3 depicts a phantom in a second embodiment.

(5) FIG. 4 depicts an example of a DW-SE sequence diagram.

(6) FIG. 5 depicts an example of a representation of the non-linearity of a magnetic field gradient.

(7) FIG. 6 depicts an example of a DW-EPI sequence diagram.

(8) FIG. 7 depicts an overview.

(9) FIG. 8 depicts a flow diagram of an example of a method for ascertaining a deviation of a gradient field of a magnetic resonance system from a reference.

DETAILED DESCRIPTION

(10) FIG. 1 depicts a magnetic resonance system 1 with a scanner 2 and a control apparatus 3. Three gradient coils 4, 5, and 6 are provided in the scanner 2. The gradient coil 4 generates, for example, a gradient field in the z-direction shown by the z-axis 7 which lies parallel to the axis of the B.sub.0 field. The second gradient coil 5 generates a gradient field in the direction of the x-axis 8. This direction is also called the x-direction. The third gradient coil 6 generates a gradient field in the y-direction, shown by the y-axis 9.

(11) The intersection of the axes 7 to 9 is called the origin 10 of the gradient system. The axes 7 to 9 thereby generate the device coordinate system 11. This is not identical to the gradient coordinate system of a sequence as described in the introduction, because the alignment or the readout direction is arranged arbitrarily in space.

(12) The designation of the directions is a pure convention, and they are, in principle, arbitrarily identifiable.

(13) The origin 10 may match the isocenter of the magnetic resonance system but does not have to.

(14) The control apparatus 3 of the magnetic resonance system 1 may have a data carrier 12 on which a computer program product 13 for carrying out the described method is stored.

(15) The control apparatus may have a monitor 14 as the output device and a keyboard 15 as the input unit.

(16) Further common components of the magnetic resonance system such as a transmitting coil arrangement, a receiving coil arrangement, a patient support, etc. are not shown for the sake of clarity.

(17) The transmitting coil arrangement may be configured as a body coil. However, it may also be a transmitting coil array.

(18) A coil array may be used as a receiving coil arrangement. With a coil array, the scan time may then be shortened.

(19) FIG. 2 depicts a phantom 16 in a first embodiment in cross-section. It is configured spherical and has a wall 17. Exclusively a liquid 18 is arranged as the material within the wall 17. Where relevant, air may collect at the upper side of the phantom 16, for example, because small quantities of the liquid 18 emerge from the phantom 16. Nevertheless, the liquid 18 is the only liquid and there may also be no solid body situated within the wall 17.

(20) Accordingly, the phantom 16 has isotropic diffusion properties in all directions.

(21) The liquid 18 may be water. It may be mains water or distilled water. The water may have a relaxation agent added. By this, the T1 relaxation time is reduced, whereby the scan time may be shortened.

(22) Alternatively, the liquid may be an oil. This liquid (e.g., oil) may also be mixed with a relaxation agent.

(23) FIG. 3 depicts, in cross-section, an alternative embodiment of a phantom 16. In this case, only the shape of the wall 17 is cuboid. Otherwise, the statements relating to FIG. 2 apply. The overall shape may be cylindrical or cuboid.

(24) FIG. 4 depicts a sequence diagram 19 of a DW-SE scan sequence.

(25) The DW-SE scan sequence has, as a high frequency pulse shown on the acquisition (ACQ) axis, an excitation pulse 20, and a refocusing pulse 21. The excitation pulse may have a flip angle of 90° and a refocusing pulse of 180°. The excitation pulse 20 and the refocusing pulse 21 together generate a spin echo as the echo signal 22.

(26) In the phase encoding direction G.sub.p, the phase encoding gradient 23 is applied.

(27) In the slice selection direction G.sub.s, apart from the slice selection gradients 24 and the slice rephasing gradients 25, there is also a slice selection gradient 26 for the refocusing pulse 21 and crusher gradients 27.

(28) The slice rephasing gradient 25 rephases the magnetization which has been dephased by the slice selection gradients 24. Its gradient moment may be half the size of that of the slice selection gradient 25.

(29) The crusher gradients 27 may also be configured such that they are linked to the slice selection gradients 26 without gradient ramps. This combined gradient has a shape like, for example, the slice selection gradient 24, but is longer. By the crusher gradients 27, all the portions of the magnetization which were not refocused by 180° are destroyed. They may alternatively or additionally also be applied in the reading direction G.sub.R or the phase encoding direction G.sub.P.

(30) In the reading direction G.sub.r, both the reading dephasing gradient 28, the reading gradient 29 and also the diffusion gradient 30 and 31 are applied.

(31) By the diffusion gradients 30 and 31, a b-value of b=g.sup.2G.sup.2d.sup.2(D−d/3) is generated as described above. These are idealized with perpendicular gradient ramps, as assumed in the calculation of the b-value. In reality, the ramps naturally have a slope.

(32) The reading-dephasing gradient 28 and/or the phase encoding gradient 23 may also be arranged directly before the readout gradient 29, but as known, with a changed polarity.

(33) For carrying out the method described, on the z-gradient, the reading direction G.sub.r may be placed in the z-direction as shown in FIG. 1. The phase encoding direction G.sub.p may be placed in the x-direction and the slice selection direction G.sub.s in the y-direction. An arrangement as shown in FIG. 5 is then obtained.

(34) FIG. 5 is an exemplary representation of the non-linearity of a magnetic field gradient. The gradient field 32 to be tested lies in the z-direction. The ordinates of the gradient field represents a magnetic field amplitude. Variations of the gradient field may lie not only in the z-direction, but also along the x-axis (not shown).

(35) The phase encoding direction G.sub.p lies in the x-direction and the slice selection direction G.sub.s lies in the y-direction.

(36) The dashed line shows a linear form, whereas the real gradient field 32 is flattened at the edges of the visual field. The frame surrounding the gradient field 32 shows the field of view of an image data set to be recorded, e.g. the image data set 36.

(37) In order to be able to calculate a map of apparent diffusion coefficients, a second image data set with a different b-value is to be recorded. For the sake of simplicity, the sequence according to FIG. 4 may be run again, whereby the diffusion gradients are set to 0. The echo time and other parameters remain unaffected.

(38) In order to accelerate the image generation, in place of the DW-SE shown in FIG. 4, a sequence diagram 34 of a DW-EPI shown in FIG. 6 may be used. This matches the DW-SE in many parts, so that reference is made to the description relating thereto.

(39) As distinct from DW-SE, however, in place of an individual echo signal 22, a gradient echo train 35 for generating a plurality of echo signals 22 is used. In particular, all the necessary k-space rows may be recorded as single shot-EPIs.

(40) For phase encoding, so-called gradient blips 33 are also used. This involves constant gradient moments which bring about the predetermined phase change.

(41) A further reduction of the scan time may be obtained by using a receiving coil array.

(42) FIG. 7 depicts an overview for carrying out the method described.

(43) In the exemplary method, the image data sets 36, 37, 38, 39, and 40 are recorded on the magnetic resonance system 1. The image data set 36 may be recorded as shown in FIG. 4. When the image data set 37 is recorded, the diffusion gradients are set to 0. During the recording of the image data set 38, the polarity of the gradients is swapped as compared with the image data set 36. The image data sets 39 and 40 correspond to the image data sets 36 and 37, wherein the reading direction G.sub.r and the phase encoding direction G.sub.p were swapped as compared with the image data sets 36 and 37. The diffusion gradients are each applied in the z-direction, as also in the image data sets 36 to 38 of the reading gradient. In the image data sets 39 and 40, the reading gradient lies in the x-direction. The slice selection gradient lies in the y-direction.

(44) The image data sets 41 to 45 are subject to the same scheme, whereby here the gradient field is investigated in the y-direction. Accordingly, the diffusion gradients lie in the y-direction. The reading gradient direction G.sub.r then lies, in the first three image data sets 41, 42 and 43, in the y-direction and, in the image data sets 44 and 45, in the z-direction. The phase encoding direction G.sub.p only then lies in the z-direction and thereafter in the y-direction. The slice selection takes place in the x-direction.

(45) For control of the gradient field in the x-direction, a similar process is used: the diffusion gradients lie in the x-direction. The reading gradient direction G.sub.r then lies, in the first three image data sets 46, 47 and 48, in the x-direction and, in the image data sets 49 and 50, in the y-direction. The phase encoding direction G.sub.p only then lies in the y-direction and thereafter in the x-direction. The slice selection takes place in the z-direction.

(46) The image data sets 36 to 50 may be evaluated on the control apparatus 3 of the magnetic resonance system 1 or on an external computer. They are then transferred to this computer after the recording.

(47) From the image data sets 36 to 40, a first map 51 of apparent diffusion coefficients is created. Thereby, spherical surface functions may be fitted to the diffusion coefficients. As reference values for the apparent diffusion coefficients obtained thereby, a map 52 is used which was generated at an earlier time point, for example, during the commissioning of the magnetic resonance system 1, with the same recording and evaluating parameters. If the deviations exceed a predetermined threshold value, the gradient fields are scanned again. Then, the map 51 may be used as a new reference map.

(48) The same procedure leads, for the image data sets 41 to 45, to a map 53 of apparent diffusion coefficients which are compared with a reference map 54. The image data sets 46 to 50 result in a map 55 of apparent diffusion coefficients which are compared with a reference map 56.

(49) Naturally, the image data sets 36 to 50 are recorded as so-called raw data sets which are to be at least Fourier transformed for generating an image. On which computer the reconstruction of the image data sets takes place is unimportant.

(50) In order to keep the evaluation simple, at least two different diffusion-weighted image data sets are to be created for each gradient direction that is to be monitored. If it is wished to check only the z-direction, then only the provision of the image data sets 36 and 37 is necessary for the execution of the method. The recording of the image data sets itself may be an act preceding the method.

(51) At least for a DW-EPI recording, the image data set recorded with a b-value of 0 may also be repeated with a swapped phase encoding direction. In this way, the influence of undesirable side-effects such as, for example, distortions from B.sub.0-inhomogeneities on the ascertaining of the ADC maps may be reduced.

(52) An exemplary list of combinations for the investigation of the y-gradient field shows the position of the gradient coordinate system with the slice selection direction G.sub.S, the readout direction G.sub.R and the phase encoding direction G.sub.P in respect of the device coordinate system with the directions x, y and z:
G.sub.S=x,G.sub.R=y,G.sub.P=z
G.sub.S=x,G.sub.R=z,G.sub.P=y
G.sub.S=z,G.sub.R=y,G.sub.P=x
G.sub.S=z,G.sub.R=x,G.sub.P=y

(53) Each is recorded with:

(54) b=0

(55) b!=0, first polarity of the y-diffusion gradients

(56) b!=0, second polarity of the y-diffusion gradients wherein the operator != indicates a value not equal to 0.

(57) All in all, a maximum of 12 scans are to be recorded, of which at least two with different b-values are absolutely necessary.

(58) FIG. 8 depicts a flow diagram of a method for ascertaining a deviation of a gradient field of a magnetic resonance system from a reference.

(59) In act S1, a phantom 16 with isotropic diffusion properties is positioned in the scanner 2. The phantom 16 may be, for example, a spherical or cylindrical hollow body filled with water or other liquid with isotropic diffusion properties. The phantom may be positioned so that it includes the spatial region of the gradient field to be tested, in particular, a region to be tested for non-linearity. If a plurality of regions are to be acquired, a plurality of scans may be carried out with different positioning of the phantom.

(60) Thereby, a positioning aid may be used in order to achieve a defined positioning of the phantom.

(61) Through the use of local receiving coils, that is, a receiving coil array, the SNR (signal-to-noise ratio) may thus be increased and/or the scan time may be shortened. In principle, however, the whole-body coil may be used as the receiving coil.

(62) In act S2, a scan is to be carried out with diffusion weighting along the gradient field direction to be tested, with a first b-value, (e.g., one of the image data sets denoted as image data set 36, 41, or 46 in FIG. 7 is to be recorded).

(63) For example, the physical x-, y-, and z-axes of the gradient coils may be scanned one after another. The method may however equally be applied to any desired overlaying of physical directions or to gradient coil arrays, that is, gradient systems with more than three independent channels. For example, b=0 may be selected as the first b-value. However, the method also functions with other selections of the values, provided diffusion coefficients may be ascertained from the scans.

(64) As the scanning method, any diffusion-weighted scan which permits the ascertaining of diffusion coefficients may be used. For example, a diffusion-weighted echoplanar imaging may also be used in clinical routine—as a single shot or a segmented method.

(65) In act S3, a scan is to be carried out with diffusion weighting along the same gradient field direction with a second b-value, (e.g., one of the image data sets denoted as image data set 37, 42, or 47 in FIG. 7 is to be recorded).

(66) The at least one second b-value may differ from the first to the extent that a reliable ascertainment of the apparent spatial variations of the diffusion coefficient is possible. For this, the first and the second b-value differ by at least 10 s/mm.sup.2, or by more than 100 s/mm.sup.2. As the second b-value, for example, a value in the range [e.g., 100 s/mm.sup.2, 2000 s/mm.sup.2] may be selected. Depending upon the liquid used and the main field strength, the selection of the parameters may be different in order to optimize the sensitivity of the method. Of decisive importance ultimately are a) the SNR and b) a sufficient difference of the signal strength brought about by diffusion effects.

(67) Acts S1 to S3 are to be carried out on the magnetic resonance system 1 and acts S4 and S5 may be carried out on any computers.

(68) In act S4, a map of apparent diffusion coefficients is created.

(69) The signal amplitude of a diffusion-weighted scan is described by S(b)=S0 exp(−b ADC). From two scans of the signal S(b) with at least two different and known values of the assumed b-value, an apparent diffusion coefficient ADC may be ascertained for each image element analytically or by regression methods.

(70) Because the liquid has an isotropic diffusion behavior, variations in the ascertained ADC map are attributable to spatial deviations of the actual local b-value. The latter are caused by the non-linearities of the gradient system.

(71) In act S5, the ADC map is compared with reference values. The acts S1 to S4 are repeated at different time points in order to align the ADC maps then obtained with the initially created ADC map. A comparison of the spatial distribution of the ADC with that of the reference scan permits temporal variations of the non-linearity of the tested gradient field to be detected.

(72) For example, the ADC maps ascertained at different time points may be subtracted from one another. Variations in the difference map indicate a change in the gradient field, in particular, non-linearities.

(73) Dependent upon the liquid, the actual diffusion coefficient may dependent upon the temperature. In order to eliminate this influence, the ADC maps may be normalized before or after the subtraction. The normalization may take place at a defined location, for example, the isocenter of the magnet.

(74) Through the setting of limit values, a criterion may be defined for a sufficient stability of the non-linearity. For example, a linearity deviation of +/−1% or +/−5% may be defined as sufficient. Therefrom, for example, for the diffusion encoding with two identical rectangular gradient pulses of amplitude G and duration d at the spacing D, a permissible deviation of the actual local b-value may be calculated according to b=g.sup.2 (D−d/3). Finally, a limit may thereby be ascertained for the spatial variation of the difference ADC maps. Through spatial filtration, for example, with a Gaussian filter or by adaptation of model functions, (e.g., spherical surface functions), where needed, the noise in the ADC maps or in the difference map may be reduced in order to ascertain reliably the variations of non-linearities, e.g., on relatively large length scales.

(75) It is to be understood that the elements and features recited in the appended claims may be combined in different ways to produce new claims that likewise fall within the scope of the present disclosure. Thus, whereas the dependent claims appended below depend from only a single independent or dependent claim, it is to be understood that these dependent claims may, alternatively, be made to depend in the alternative from any preceding or following claim, whether independent or dependent, and that such new combinations are to be understood as forming a part of the present specification.

(76) While the disclosure has been illustrated and described in detail with the help of the disclosed embodiments, the disclosure is not limited to the disclosed examples. Other variations may be deducted by those skilled in the art without leaving the scope of protection of the claimed disclosure.