Method for determining the spatial distribution of magnetic resonance signals in subvolumes of an object under examination

Abstract

A method for determining the spatial distribution of magnetic resonance signals from at least one of N subvolumes predefines a reception encoding scheme and determines unique spatial encoding for at least one of the subvolumes but not for the entire volume under examination (UV). A transmission encoding scheme is also defined, wherein encoding is effected via the amplitude and/or phase of the transverse magnetization. The temporal amplitude and phase profile of the RF pulses is then calculated and each reception encoding step is carried out I times with variations according to the I transmission encoding steps in the transmission encoding scheme. The method makes it possible to largely restrict the spatially resolving MR signal encoding and image reconstruction to subvolumes of the object under examination without the achievable image quality sensitively depending on imperfections in the MR apparatus.

Claims

1. A method for determining a spatial distribution of magnetic resonance signals from at least one of N non-overlapping subvolumes of an object under examination in a measurement volume of a magnetic resonance apparatus, wherein N2, the method comprising the steps of: a) executing a preparation step, the preparation step comprising: a1) selecting a measurement sequence with encoding steps, wherein each encoding step includes irradiation of one or more spatially selective RF pulses to effect one magnetization change in each encoding step; a2) selecting the N subvolumes such that, taken together, those subvolumes completely cover at least one volume under examination, the volume under examination corresponding to a part of the object under examination in which nuclear spins excited during execution of the selected measurement sequence contribute to at least one of acquired MR signals; a3) selecting a reception encoding scheme with K reception encoding steps, wherein K1, the reception encoding scheme defining unique spatial encoding in at least one spatial dimension for at least one of the subvolumes, wherein that spatial encoding is not unique for an entire volume under examination in at least one spatial dimension; a4) defining a transmission encoding scheme with I transmission encoding steps, wherein IN2, wherein encoding is effected via an amplitude and/or phase of transverse magnetization defined spatially dependently by means of a magnetization change, wherein, for each of the I transmission encoding steps, the magnetization change is defined such that, at no position within each subvolume, a same encoding is defined as at another position within another subvolume, with excited nuclear spins contributing to an acquired magnetic resonance signal in at least one transmission encoding step and in at least two of the subvolumes; and a5) calculating a temporal amplitude and phase profile of the spatially selective RF pulses to be irradiated to effect the magnetization changes; b) carrying out an execution step for all encoding steps, each reception encoding step that is defined according to the reception encoding scheme being executed I times with variations according to the I transmission encoding steps of the transmission encoding scheme, wherein, in each encoding step, all RF pulses calculated for each transmission encoding step of the transmission encoding scheme are applied by means of at least one transmission element and, without overlapping in time with those RF pulses, a spatial encoding is effected according to a reception encoding scheme, with magnetic resonance signals being acquired by means of at least one reception element; c) executing a reconstruction step based on the transmission encoding scheme, wherein components of the acquired magnetic resonance signals are assigned to the N subvolumes and, for at least one of the subvolumes spatially encoded according to the reception encoding scheme, one or more spatial distributions of the magnetic resonance signals is reconstructed from the acquired magnetic resonance signals and/or variables derived therefrom are calculated, wherein this or these subvolumes are designated mapping volumes; and d) executing a result step in which results of the reconstruction step are stored and/or displayed.

2. The method of claim 1, wherein the transmission encoding scheme only defines the amplitude of the transverse magnetization to be set spatially dependently by means of the magnetization change across the I transmission encoding steps.

3. The method of claim 1, the transmission encoding scheme only defines the phase of the transverse magnetization to be set spatially dependently by means of the magnetization change across the I transmission encoding steps.

4. The method of claim 1, wherein an entirety of mapping volumes is a non-contiguous region.

5. The method of claim 1, wherein at least one mapping volume is restricted to a size that is essential for a measurement task.

6. The method of claim 1, wherein only one mapping volume is selected.

7. The method of claim 1, wherein at least two mapping volumes are selected and each of these mapping volumes is uniquely spatially encoded with the reception encoding scheme, wherein a union of the mapping volumes is thereby not uniquely spatially encoded.

8. The method of claim 1, wherein reception of the magnetic resonance signals is performed by means of at least two reception elements.

9. The method of claim 1, wherein the magnetization changes are effected by means of at least two transmission elements.

10. The method of claim 1, wherein in step a), temporally and spatially varying additional magnetic fields, which are produced with a gradient system and act during irradiation of the RF pulse or RF pulses to be irradiated to effect the magnetization change, are defined and, for these additional magnetic fields for each of the I transmission coding steps of the transmission coding scheme, a temporal amplitude and phase profile of the RF pulses to be irradiated to effect the magnetization change is calculated, wherein, in step b), those RF pulses are applied during action of the additional magnetic fields.

11. The method of claim 1, wherein at least one subvolume is adapted to anatomical, morphological or functional characteristics of the object under examination.

12. The method of claim 1, wherein calculation of assignment of magnetic resonance components of the mapping volumes is effected by means of Fourier transformation, Hadamard transformation or wavelet transformation.

13. The method of claim 1, wherein, in all encoding steps and for at least one subvolume, a same flip angle is set everywhere to effect the change in magnetization.

14. The method of claim 1, wherein, in all encoding steps and by means of the spatially selective RF pulses effecting the magnetic change, different characteristics of the MR signal are defined in at least two subvolumes in addition to the different encoding in accordance with the transmission encoding scheme.

Description

BRIEF DESCRIPTION OF THE DRAWING

(1) FIG. 1 a schematic representation of a magnetic resonance apparatus according to prior art suitable for performing the inventive method;

(2) FIG. 2 a schematic representation of especially preferred measurement topologies;

(3) FIG. 3 a flow chart of a possible sequence of the inventive method;

(4) FIG. 4 an MR overview image with a water bottle as the object under examination and with subvolumes drawn in;

(5) FIG. 5 a suitable amplitude and phase distribution of the transverse magnetization for transmission encoding steps 1 and 2 according to the transmission encoding scheme for spatial encoding;

(6) FIG. 6 a schematic representation of the sequence of gradients and RF pulses in the encoding steps during the execution step;

(7) FIG. 7 the experimentally determined amplitude and phase distribution of transverse magnetization realized with RF pulses according to the transmission encoding scheme;

(8) FIG. 8 the separate representation of signals from two subvolumes; and

(9) FIG. 9 the MR image of a segment of a tangerine implemented with a reduced reception encoding scheme and excitation encoding according to the transmission encoding scheme.

DESCRIPTION OF THE PREFERRED EMBODIMENT

(10) FIG. 1 schematically shows a magnetic resonance apparatus that is suitable for performing the inventive method. The apparatus contains a main magnet M, with which the essentially homogeneous and static basic magnetic field is produced in a measurement volume V. The part of the object under examination that is contained in the measurement volume will subsequently be referred to as the object under examination or simply the object O. Surrounding the measurement volume V, a gradient system is put into the bore of the main magnet M with which different variants of additional magnetic fields can be implemented by connecting coils, usually a plurality of coils, to form coil combinations G1, G2, G3, . . . . FIG. 1 shows an example of three such coil combinations, G1, G2, and G3. With the gradient system, additional magnetic fields of controllable duration and strength can be superimposed on the basic field. With gradient amplifiers A1, A2, A3, that are controlled by a sequence control unit SEQ to produce gradient pulses at the right instant, the gradient coils sets G1, G2, and G3 are supplied with electric power to produce the additional fields.

(11) Within the gradient field system, there are multiple transmission elements, TA1 to TAn, which are together termed the transmission antenna device. They surround an object under examination O and are powered from multiple independent RF power transmitters TX1 . . . TXn. The RF waveforms produced by these RF power transmitters TX1 . . . TXn are determined by the sequence control unit SEQ and triggered at the correct time. With the transmission elements TA1 to TAn, RF waveforms are irradiated onto the object under examination O in the volume under examination V, where they excite nuclear spins. The magnetic resonance signals caused by this are converted into electrical voltage signals with one or more RF reception elements RA1, . . . , RAm and are then fed into a corresponding number of reception units RX1, . . . , RXm. The reception elements RA1, . . . , RAm are together termed the reception antenna equipment consisting of m reception elements RA1, . . . , RAm. They are also located within the gradient coils G1, G2, G3, and surround the object under examination O.

(12) To reduce the complexity of the apparatus, the transmission and reception antenna devices can be designed and connected in such a way that one or more of the transmission elements TA1 to TAn are also used to receive the magnetic resonance signals. In such a case, which is not shown in FIG. 1, switchover between transmission and reception modes is assured by one or more of the electronic transmission-reception switches controlled by the sequence control unit SEQ, that is, that during the RF transmission phases of the executed pulse sequence, this antenna or these antennas are connected with the corresponding RF power transmitter or transmitters and disconnected from the allocated reception channel or channels, while, for the reception phases, transmitter disconnection and reception channel connection is performed.

(13) With the reception units RX1 to RXm shown in FIG. 1, the signals received are amplified and converted to digital signals using known signal processing methods and passed on to an electronic computer system COMP. In addition to reconstruction of images and spectrums and derived quantities from the measured data received, the control computer system COMP is used to operate the entire MR measurement apparatus and to initiate execution of the pulse sequences by appropriate communication with the sequence control unit SEQ. User-guided or automatic execution of programs for adjusting the measurement apparatus characteristics and/or for generating magnetic resonance images is also performed by this control computer system COMP, as are visualization of the reconstructed images and storage and administration of the measurement and image data and control programs. For these tasks, this computer system is equipped with at least one processor, a working memory, a computer keyboard KB, a pointing device PNTR, for example, a computer mouse, a monitor MON, and an external digital storage unit DSK.

(14) An explanation of how the inventive method can be performed with such an MR measurement apparatus is given below based on specific examples. In the presentation of these examples, eight transmission elements were used, which are simultaneously used as reception elements.

(15) FIG. 2 schematically illustrates two especially preferred measurement topologies for the inventive method, wherein the figures can be interpreted as direct representations of a two-dimensional reception encoding scheme or as cross-sectional representations of a three-dimensional reception encoding scheme. FIG. 2a illustrates the mapping of a single subvolume SV1, i.e. SV1 represents a mapping volume and the inner volume, while another subvolume SV2 corresponds to the outer volume. SV1 and SV2 completely cover the volume under examination UV. The reception encoding scheme is designed in such a way that it spatially encodes uniquely in volume EV. In FIG. 2b, SV1, SV2, SV3, and SV4 represent the mapping volumes that together define the inner volume, while SV5 represents the outer volume, which is not to be mapped. In this case, the reception encoding scheme in volumes EV1, EV2, EV3, and EV4 permit unique spatial encoding, wherein, in the present example, these volumes are identical in size.

(16) The examples described below have the measurement topology shown in FIG. 2a.

(17) Firstly, execution of a transmission encoding, implemented specifically here as excitation encoding, is described for two subvolumes by means of an imaging experiment, in which a water-filled bottle was used as the object under examination (FIG. 4). Then the inventive method is applied to map a single segment of a tangerine (FIG. 9) using the same procedure, wherein this segment was covered by a subvolume and the remainder of the tangerine was acquired in an additional subvolume that defines the outer volume. The experiments described are three-dimensional imaging experiments, wherein only individual slices of the reconstructed data sets are represented in the illustrations.

(18) FIG. 3 schematically shows the sequence in which the examples are performed.

(19) In both examples, the sequence starts by choosing two subvolumes, which together entirely cover the volume under examination. These were defined in both cases based on a previously acquired overview image acquired in an MR experiment, from which the delimitation of the volume under examination can be seen. It is important to ensure that the subvolumes do not overlap and are preferably slightly larger than the volume under examination, to avoid the precondition of full coverage of the volume under examination being violated by imprecisions in the geometrical delimitation.

(20) FIG. 4 shows the selection of a cuboid inner subvolume SV1 for the first example and an outer subvolume SV2, which fully surrounds the inner subvolume SV1. In principle, both subvolumes SV1, SV2, can be any shape, an aspect which is better demonstrated by the example of a tangerine segment. In this example, subvolume SV1 is selected as a mapping volume and therefore an inner volume from which an image is to be generated.

(21) For this, a three-dimensional reception encoding scheme is defined in such a way that using frequency and phase encoding, which is implemented by creating additional magnetic fields, MR signals that are derived from the subvolume SV1 are fully and uniquely three-dimensionally spatially encoded.

(22) For this, the duration and amplitude of the gradient pulses G.sub.x, G.sub.y, G.sub.z, which are generated by gradient coils G1, G2, and G3 are determined and the number of encoding steps in this example is defined as K=6464, so that the desired volume is encoded with the desired resolution.

(23) Because the reception encoding scheme specified in this way is uniquely spatially encoded in SV1 but not for the entire volume under examination, MR signal components that occur outside the reception encoding region create artifacts in the image of subvolume SV1. In the spatial direction, in which frequency encoding is performed, this can be avoided by frequency-selective filtering of the acquired signal. This does not work in the phase encoding directions.

(24) According to the invention, therefore, a second encoding scheme, the transmission encoding scheme is defined to differentiate signals from subvolumes SV1 and SV2, which is achieved by irradiating spatially selective RF pulses.

(25) In order to differentiate between signals from the two subvolumes SV1 and SV2, the simplest form of a Fourier encoding scheme with I=2 transmission encoding steps is used, in which the transverse magnetization phase is varied. FIG. 5 shows that the amplitude of the transverse magnetization in both transmission encoding steps is kept constant by defining a thoroughly homogeneous flip angle of 8 across all transmission encoding steps, while the phase distribution is varied. The transverse magnetization generated in the experiment can have any phase distribution in the first transmission encoding step, which is used as a reference for the second transmission encoding step and is assumed by definition to be homogeneously 0. In the second transmission encoding step, the transverse magnetization phase within SV1 relative to this reference is 0 while, in SV2, it is to assume a value of 180.

(26) In this example, therefore, only the transverse magnetization phase is varied. However, it is not absolutely necessary to define an homogeneous flip angle for the entire examination volume. Rather it is possible to define different homogeneous flip angles within each of the subvolumes SV1 and SV2. For example, it can be advantageous to excite the mapping volume with a flip angle, which achieves (on average) an MR signal maximum (called Ernst angle), while the outer volume is excited with a flip angle, in which imperfections of the excitation only have a slight effect on the resulting MR signal strength. In this way, the intended MR signal separation with respect to subvolumes SV1 and SV2 can be performed with greater precision.

(27) In these examples, the transmission encoding is implemented by irradiation of spatially selective excitation pulses and the flip angle and phase patterns illustrated in FIG. 5 show the requirements for the change in magnetization for each transmission encoding step. Although the requirements geometrically extend beyond the volume under examination, they are not applied outside the volume under examination because, for example, no nuclear spins exist there or they lie outside the region of sensitivity of the transmission antenna equipment.

(28) In the examples, PEX pulses are used as spatially selective RF pulses to achieve the change in magnetization, i.e., multiple channel RF pulses, which in combination with gradient pulses, are irradiated via a corresponding number of transmission elements, eight in our examples. Because spatially linear gradient fields are used, the gradient pulses deployed can be represented as a k-space trajectory, which, in this case, exhibits a progression of spirals stacked one on top of the other. Due to the undersampling of the k-space trajectory resulting from the additional sensitivity encoding of the transmission elements, the length of the excitation pulses could be reduced by a factor of 4 as compared with single-channel transmission.

(29) The phase and amplitude profile of the two required RF pulses for implementing the change in magnetization according to the selected transmission encoding scheme is calculated using a method according to [8], wherein the targeted change in magnetization, the spiral-shaped k-space trajectory, and the transmission profiles of the 8 transmission elements used are included in the calculation.

(30) In the example, an imaging experiment now follows in the form of a so-called gradient echo experiment, whose operating sequence is schematically represented in FIG. 6. The plotted gradient waveforms for spatial RF pulse encoding and the RF waveforms, in particular, are only schematic and do not provide a precise rendition of the waveforms actually used.

(31) The first excitation is performed according to transmission encoding step 1 with the first of the calculated PEX pulses. Then spatial encoding is performed according to reception encoding scheme 1 with phase gradients in the y and z direction. Another component of the reception encoding scheme is encoding in the x direction using a readout gradient, which is applied while the MR signals are acquired. This procedure is repeated for each reception encoding step as a loop S1, wherein the amplitude of the phase gradients is varied. This entire procedure is then repeated again for each transmission encoding step as loop S2, wherein the corresponding RF pulses are applied. The sequence in which the loops are executed is not important. It is merely necessary to ensure that every encoding step combination of reception and transmission encoding scheme is executed. Data reconstruction must be adapted to the sequence in which the encoding steps are executed.

(32) The data acquired in loop S1 can either be reconstructed separately from those in loop S2 as individual images, in this case by means of three-dimensional Fourier transformation, or combined in a signal reconstruction step by means of four-dimensional Fourier transformation.

(33) FIG. 7 shows the reconstruction result of the two data sets acquired in loop S1. Please note that, unlike in the inventive method, a reception encoding scheme, which encodes the volume under examination uniquely and in full was used here in order to visualize the working principle of the transmission encoding scheme for acquiring these data sets. The amplitudes and phases of the transverse magnetization essentially show the profiles as defined in FIG. 5 (weighted with the reception profiles of the reception elements). In the simple case of Fourier encoding used for the transmission encoding scheme, the signals from SV1 and SV2 can now be simply separated by adding and subtracting the complex data sets.

(34) Adding the data sets results in the addition of the signal amplitudes in SV1, and the phase difference during excitation according to the transmission encoding scheme results in the subtraction and cancellation of the signal amplitudes in SV2. Conversely, subtracting the data sets results in cancellation in SV1 and addition of the signal amplitudes in SV2. The signal distribution for SV1 and SV2 is shown in FIG. 8. The MR signals from the inner volume are thus clearly separated from those from the outer volume.

(35) Although the total measurement time increases with each transmission encoding step, the resulting ratio of desired signal to noise per unit time remains unchanged because, when the data are reconstructed, the correlated desired signals from the individual encoding steps increase while uncorrelated noise is reduced by averaging effects.

(36) FIGS. 7 and 8 show reconstructed images of the first example, which, in order to demonstrate the transmission encoding scheme, were acquired with a reception encoding scheme that uniquely encodes the entire volume under examination. In the second example, in which a tangerine was used as the object under examination, in accordance with the inventive method the volume encoded with the reception encoding volume was chosen to be considerably smaller than the volume under examination, as illustrated in FIG. 9.

(37) FIG. 9a shows an overview image of the tangerine, in which the two subvolumes, the mapping volume SV1 and the outer volume SV2, as well as the volume uniquely encoded by the reception encoding scheme are drawn in. FIGS. 9b and c show the images acquired in transmission encoding steps 1 and 2 and reconstructed according to the reception encoding scheme, which clearly show artifacts caused by the non-unique reception encoding in the volume under examination. Only the separation of the signals by MR signals from subvolumes SV1 and SV2 according to the transmission encoding scheme produces an artifact-free image of subvolume SV1, as shown in FIG. 9d.

(38) By reducing the encoding region of the reception encoding scheme while retaining the same number of encoding steps, in this case the resolution in FIG. 9d could be increased over FIG. 9a without having to increase the measurement time.

(39) The number of encoding steps can also be reduced in an analogous manner to acquire the data in a shorter measurement time while retaining the spatial resolution. A combination of increased resolution and shorter measurement time is also possible.

LIST OF REFERENCE SYMBOLS

(40) A1, A2, A3 Gradient amplifier COMP Computer system DSK Storage unit EV, EV1 . . . 4 Volumes in which the reception encoding scheme is uniquely spatially encoded G Housing G1, G2, G2 Gradient coils G.sub.x, G.sub.y, G.sub.z Gradient fields KB Computer keyboard M Main magnet MO Screen O Object under examination PNTR Pointing device RA1 . . . M RF reception elements RX1 . . . M Reception units SEQ Sequence control unit SV1 . . . 5 Subvolume TA1 . . . N RF transmission elements TX1 . . . N RF power transmitter V Measurement volume UV Volume under examination

REFERENCES

(41) [1] Feinberg D A, Hoenninger J C, Crooks L E, Kaufman L, Watts J C, Arakawa M. Inner Volume MR Imaging: Technical Concepts and Their Application. Radiology 156 (1985), pp. 743-747. [2] Glover G H. Phase-offset multiplanar (POMP) volume imaging: a new technique. J Magn Reson Imaging 1 (1991) pp. 457-461 [3] Pauly J, Nishimura D, Macovski, A. A k-space analysis of small-tip-angle excitation. Journal of Magnetic Resonance 81 (1989), pp. 43-56. [4] Ullmann P, Schneider J T, Haas M, Wissmann R, Ruhm W. Single-And Multi-Voxel MR-Spectroscopy Using Parallel Excitation. Proc 17th Scientific Meeting, Intl Soc Magn Reson Med, Honolulu (2009), p. 2602 [5] DE 10 2007 044 463 B4 [6] Katscher U, Brnert P, Leussler C, van den Brink J S. Transmit SENSE. Magnetic Resonance in Medicine 49 (2003), pp. 144-50. [7] Seifert F, Wuebbeler G, Junge S, Rinneberg H. B1(+) steering by an adaptive 4-channel transmit/receive coil array. Proc 12th Scientific Meeting, Intl Soc Magn Reson Med, Kyoto (2004), pp. 1569. [8] Grissom W, Yip C-Y, Zhang Z, Stenger, V A, Fessler J A, Noll D C. Spatial domain method for the design of RF pulses in multicoil parallel excitation. Magnetic Resonance in Medicine 56 (2006), pp. 620-629.