METHOD AND DEVICE FOR CONTROLLING A MAGNETIC RESONANCE IMAGE SYSTEM IN AN ECHOPLANAR IMAGE SYSTEM

Description

[0001] Regardless of the grammatical gender of a particular term, persons with male, female or other gender identities are included.

[0002] The invention relates to a method for controlling a magnetic resonance imaging system for generating magnetic resonance image data of an examination subject, in which k-space data is acquired during accelerated echo-planar imaging with subsampling according to a pulse sequence scheme. The invention also relates to a control device for a magnetic resonance imaging system. Furthermore, the invention relates to a magnetic resonance imaging system.

[0003] Imaging systems based on magnetic resonance imaging, particularly of nuclear spins, so-called magnetic resonance tomographs, have successfully established and proven themselves in a wide variety of applications. In this type of image acquisition, a static basic magnetic field B 0 , which serves to initially align and homogenize the magnetic dipoles under investigation, is usually superimposed with a rapidly switched magnetic field, the so-called gradient field, to spatially resolve the imaging signal. To determine the material properties of an object to be imaged, the dephasing or relaxation time is determined after a deflection of the magnetization from the initial alignment, so that various material-specific relaxation mechanisms and relaxation times can be identified.The deflection is usually achieved by a number of RF pulses (the abbreviation RF stands for high frequency), also known as excitation pulses, and the spatial resolution is based on a time-defined manipulation of the deflected magnetization with the help of the gradient field in a so-called measurement sequence or control sequence, which defines a precise temporal sequence of RF pulses, the change of the gradient field (by emitting a switching sequence of gradient pulses) and the acquisition of measured values.

[0004] Typically, an intermediate step is used to map the measured magnetization—from which the aforementioned material properties can be derived—to a spatial coordinate of the measured magnetization in the spatial space in which the object under investigation is located. In this intermediate step, acquired raw magnetic resonance data, also referred to as k-space data, are arranged at readout points in so-called "k-space," with the k-space coordinates encoded as a function of the gradient field. The magnitude of the magnetization (in particular, the transverse magnetization in a plane perpendicular to the previously described basic magnetic field) at a specific location on the object under investigation can be determined from the data of the readout point using a Fourier transformation, which calculates a signal strength of the signal in spatial space from a signal strength (magnitude of magnetization) associated with a specific frequency (the spatial frequency) or phase position.

[0005] Magnetic resonance imaging is a relatively slow imaging technique because data is acquired sequentially along lines in Fourier space or k-space, and a certain time cannot be exceeded for the spin relaxation of the excited spins. The process of acquiring images in two-dimensional slices is significantly less error-prone than in three dimensions because the number of encoding steps is fewer than in a three-dimensional technique. Therefore, many applications use image volumes with stacks of two-dimensional slices instead of a single three-dimensional image. However, the image acquisition times are very long due to the long relaxation times of the spins, which means a reduction in comfort for patients, for example.Patients cannot even briefly leave the MRI scanner or change their position during the scan, as this would disrupt the image acquisition process and require the entire process to start over. Therefore, an important goal is to accelerate the acquisition of two-dimensional slice stacks.

[0006] For example, parallel image acquisition techniques are used to accelerate image acquisition. Some image acquisition techniques can introduce artifacts due to undersampling. These artifacts can be eliminated by applying reconstruction algorithms.

[0007] Echo-planar imaging (EPI or EPI pulse sequence) is often used to accelerate the acquisition of raw magnetic resonance data. In this technique, a large number of echoes, also known as an echo train, are acquired per repetition by switching frequency gradients, also known as readout gradients. This allows an image to be recorded with a small number of excitations or repetitions.

[0008] Echo-planar imaging sequences are often applied in the form of multi-slice imaging sequences to accelerate imaging and, for example, enable images of moving objects. Despite the described echo train, echo-planar imaging sequences typically comprise a large number of repetitions. A repetition is defined as a pulse sequence segment between two excitation RF pulses. Images with at least 60 repetitions are not uncommon, particularly in so-called BOLD imaging (BOLD = blood oxygenation level dependent; this type of imaging illustrates the oxygen content in the red blood cells, which can be used, for example, to depict a patient's brain activity) and extended diffusion protocols such as MDDW (MDDW = multi-directional diffusion weighting).

[0009] Typically, three lines are acquired for each repetition immediately after the excitation pulse as part of a so-called navigator acquisition to correct Nyquist ghosting and / or B0 drift. This results in a very high temporal resolution for correcting these artifacts, given the high number of repetitions mentioned.

[0010] As already mentioned, EPI acquisitions are performed using parallel imaging techniques such as GRAPPA (Generalized Autocalibrating Partially Parallel Acquisitions) or SMS (Simultaneous Multi-Slice Imaging) to simultaneously image multiple slices.

[0011] There are also advanced correction methods, such as "dual polarity" (see Hoge et al., "Dual-Polarity GRAPPA for Simultaneous Reconstruction and Ghost Correction of Echo Planar Imaging Data," Magnetic Resonance in Medicine 2016 Jul; 76(1):32-44. Doi: 10.1002 / mrm.25839. Epub 2015 Jul 24) for particularly precise correction of ghost artifacts, or spatially resolved eddy current correction methods (see Chen et al., "Single-Shot and Segmented EPI Ghost Artifacts Removal with Two-Dimensional Phase Correction," Proc. Intl. Soc. Magn. Reson. Med. 8 (2000)). These special methods require navigator data for the aforementioned correction methods and additional reference data for supplementing undersampled k-space data. This increases the number of required repetitions of the pulse sequence and thus also the measurement time or the time for such an EPI image acquisition procedure.

[0012] Chi et al., "Field-Mapping-Embedded EPI for Geometric Distortion Correction," Proc. Intl. Soc. Mag. Reson. Med. 30 (2022) 1099, describes a method in which the existing three lines of the navigator image are modified to acquire Nyquist ghosting correction data and are continued throughout the entire acquisition so that a different k-space section to be acquired is encoded in each repetition. The acquired k-space sections are combined into suitable k-spaces, and B0 field maps for correcting geometric distortions can be obtained from them over the duration of the acquisition. The polarity of the readout gradients for acquiring the correction data is mirrored in the kx direction (readout direction), so that k-space data is acquired in three unipolar k-spaces for the subsequent image-based calculation of B0 field maps. To generate the B 0 field maps, additional reference scans are required for the application of parallel imaging techniques.The procedure is in . FIG 1 and FIG 2 shown.

[0013] US 10 162 037 B2 and US 11 280 870 B2 describe methods in which an acquisition scheme of the three lines of the navigator acquisition is modified in such a way that layer-specific ghost correction data are obtained for SMS imaging.

[0014] It is therefore an object of the present invention to provide a time-saving accelerated echo-planar imaging method with a correction of the so-called Nyquist ghost effects and with a reduced acquisition time compared to conventional methods.

[0015] This object is achieved by a method for controlling a magnetic resonance imaging system for generating magnetic resonance image data of an examination subject, in which k-space data are acquired as part of accelerated echo-planar imaging with subsampling according to a pulse sequence scheme, according to patent claim 1, by a control device for a magnetic resonance imaging system according to patent claim 12 and by a magnetic resonance imaging system according to patent claim 13.

[0016] The method according to the invention comprises scanning k-space according to a pulse sequence scheme. This pulse sequence scheme has a plurality of repetitions, each comprising a first sampling scheme for acquiring k-space data for Nyquist ghosting correction and / or for generating magnetic field maps for magnetic field correction, preferably for B0 magnetic field correction or for generating so-called B0 field maps, a subsequent second sampling scheme for an accelerated echo-planar acquisition, and an excitation scheme common to both acquisitions. Magnetic field maps indicate the deviation of the field strength of magnetic fields generated during magnetic resonance imaging from nominal values. These deviations lead to phase deviations of the measurement signals, which must be taken into account during image reconstruction.A sampling pattern is the specific way in which k-space is sampled using gradient circuits, with high-frequency signals being received as echo signals or measurement signals. In particular, such a sampling pattern corresponds to a specific k-space trajectory and a specific sequence of gradient pulses, or a specific gradient pulse sequence or gradient pulse subsequence. k-space data refers to the measurement signals assigned to specific points in k-space. The unmodified first sampling pattern is uncoded, i.e., sampling occurs without phase coding, and thus the "zeroth" k-space line is preferably sampled or acquired multiple times, particularly preferably three times, in alternating directions. An accelerated echo-planar acquisition is an echo-planar acquisition in which time is saved in the acquisition of k-space data through subsampling.

[0017] The first sampling scheme is modified in a true subset of the plurality of repetitions such that, based on the k-space data acquired with the modified first sampling scheme, the signal information missing as a result of the undersampling can be supplemented. In particular, the k-space data acquired with the modified first sampling scheme should enable the supplementation of k-space data missing as a result of the undersampling and / or the correction in image space of artifacts occurring due to the undersampling.

[0018] A true subset is a subset of the majority of repetitions that lacks at least one of the repetitions. Because the modification does not occur in all repetitions, there is also a number, or more precisely at least one repetition of the majority of repetitions, in which the modification does not occur and which is / are therefore used to acquire the k-space data for Nyquist ghosting correction. The modified sampling scheme is designed in such a way that additional k-space data is acquired, which forms a basis for supplementing the signal information missing due to the accelerated imaging. As already mentioned, this missing signal information can, in particular, include missing k-space data in undersampled k-space regions. However, the missing signal information can also include the artifacts in the image space caused by the missing k-space data.For the acquisition of a part of k-space, the modified first sampling scheme is therefore preferably encoded in contrast to the unmodified first sampling scheme, ie, phase encoding gradients are switched to acquire phase encoding lines with different phases.

[0019] As explained in more detail later, to calculate the missing k-space data, information about the measurement behavior of the magnetic resonance system is obtained based on the additional k-space data, which can be used to reconstruct the missing k-space data. In particular, in so-called multi-coil systems, which are typically used for accelerated imaging procedures, a so-called autocalibration can be used to obtain information about the weighting(s) that must be used to add signals from different coils from a point in k-space in order to calculate signals from neighboring, possibly unsampled points in k-space.

[0020] Based on these weights and the subsampled k-space data, conclusions can be drawn about the signal intensity in the unsampled regions of k-space. For this purpose, so-called kernels are preferably calculated based on the information obtained during autocalibration. These kernels include the aforementioned weights. Applying these kernels to the subsampled k-space data allows the signal information missing due to undersampling to be determined.

[0021] Alternatively, the missing signal information can also be supplemented in the image space, whereby coil sensitivity values ​​are determined that enable the calculation of artifact-free image data based on artifact-affected image data obtained with different coils by subsampling.

[0022] When reconstructing the missing signal information in image space, as is done, for example, with the SENSE method, the k-space data acquired with the modified first sampling scheme is used to determine coil sensitivity values. Furthermore, artifact-prone image data with a restricted field of view (FOV) are first reconstructed for different coils based on the subsampled k-space data. This image data exhibits a typical folding of protruding image components. Finally, artifact-free image data of the full field of view (FOV) can be calculated based on the coil sensitivity values ​​and the artifact-prone image data with a restricted field of view.

[0023] While the navigator data or k-space data for the Nyquist ghosting correction or B0 magnetic field correction are acquired uncoded or without phase encoding, encoding, preferably phase encoding, occurs when k-space data is acquired according to the modified first sampling scheme. This phase encoding enables the aforementioned autocalibration, which in particular comprises sampling a sub-region of k-space with increased resolution compared to subsampling, preferably with a full scan. Based on this higher-resolution sampled k-space data, missing k-space data or missing signal information can be determined. The reconstruction of the missing signal information can occur both in k-space, as with the GRAPPA method, for example, or by a calculation in image space, as with the SENSE method, for example. Image space data and k-space data can be converted into one another using a Fourier transform.

[0024] During the joint excitation process, a first slice selection gradient pulse is preferably generated in the slice selection direction. In addition, an excitation RF pulse is generated, which comprises N excitation frequencies for exciting one or more, preferably N, slices to be excited simultaneously, where N is an integer. As explained in more detail later, N is usually 2 or a maximum of 8.

[0025] As already mentioned, the readout process for acquiring the k-space data usually comprises the alternating switching of rephasing gradient pulses in the readout direction, whereby additional gradient pulses, for example in the phase direction and / or in the slice selection direction, are generated and RF signals are received for the acquisition of raw magnetic resonance data or the k-space data.

[0026] The method according to the invention utilizes the realization that the temporal resolution for acquiring so-called navigator data for Nyquist ghosting correction or B0 field correction does not need to be particularly high, since the temporal development of interference influences generally occurs slowly. Therefore, some of the pulse sequence sections conventionally used to determine correction data for Nyquist ghosting or B0 field correction can be converted to acquire k-space data as reference data. This reference data is used to correct artifacts resulting from undersampling.

[0027] If compensation for undersampling in k-space occurs, as in the GRAPPA method, the reference data is used to calculate missing k-space data for the actual imaging. Since the acceleration of an echo-planar acquisition is based on undersampling of k-space, the k-space data missing due to undersampling for complete image information must be estimated or approximately calculated based on reference data.

[0028] Advantageously, unlike a conventional approach, no separate acquisition process is required to acquire the reference data, thus reducing the overall imaging time. Furthermore, the duration of a repetition or echo train does not need to be increased, since the acquisition of the reference data in some repetitions simply replaces the acquisition of the navigator data. Thus, the echo time and repetition time of the individual repetitions preferably remain the same or can remain unchanged, regardless of whether navigator data, data for Nyquist ghost correction, or reference data are acquired with the first scanning scheme.

[0029] Compared to a conventional EPI pulse sequence, in which reference data is acquired separately, the image acquisition time or the total time for acquiring k-space data is advantageously reduced.

[0030] The invention also relates to a control sequence for controlling a magnetic resonance imaging system. The control sequence comprises control signals according to a pulse sequence scheme with a plurality of pulse sequence sections or repetitions, each with an excitation section and a readout section.

[0031] The repetitions comprise, as a first pulse sequence subsection, the aforementioned common excitation scheme, as a second pulse sequence subsection, the likewise already mentioned first sampling scheme for acquiring k-space data for Nyquist ghosting correction or for generating magnetic field maps for magnetic field correction, in particular for B0 field correction, and, as a third pulse sequence subsection, the aforementioned subsequent second sampling scheme for an accelerated echo planar acquisition. The first sampling scheme of a true subset of the plurality of repetitions is modified in such a way that, on the basis of the k-space data acquired with the modified sampling scheme, it is possible to supplement k-space data missing as a result of the undersampling and / or to correct artifacts in the image space resulting from the undersampling.The control sequence according to the invention shares the advantages of the method according to the invention for controlling a magnetic resonance imaging system for generating magnetic resonance image data of an examination subject, in which k-space data are acquired in the context of accelerated echo-planar imaging with subsampling.

[0032] The control device according to the invention for a magnetic resonance imaging system is designed to control the magnetic resonance imaging system using the method according to the invention for controlling a magnetic resonance imaging system to generate magnetic resonance image data of an examination subject, in which k-space data is acquired as part of accelerated echo-planar imaging with subsampling according to a pulse sequence scheme.

[0033] The magnetic resonance imaging system according to the invention comprises the control device according to the invention. The magnetic resonance imaging system according to the invention shares the advantages of the method according to the invention for controlling a magnetic resonance imaging system for generating magnetic resonance image data of an examination subject, in which k-space data is acquired as part of accelerated echo-planar imaging with subsampling according to a pulse sequence scheme.

[0034] A majority of the aforementioned components of the method according to the invention for controlling a magnetic resonance imaging system can be implemented entirely or partially in the form of software modules in a processor of a corresponding computing system, e.g., of a control device of a magnetic resonance imaging system or a computer used to control such a system. A largely software-based implementation has the advantage that even previously used computing systems can be easily retrofitted with a software update to operate in the manner according to the invention. In this respect, the object is also achieved by a corresponding computer program product with a computer program that can be loaded directly into a computing system, with program sections for executing the steps of the method according to the invention for controlling a magnetic resonance imaging system when the program is executed in the computing system.Such a computer program product may, in addition to the computer program, include additional components such as documentation and / or additional components, including hardware components such as hardware keys (dongles, etc.) for using the software.

[0035] A computer-readable medium, e.g., a memory stick, a hard disk, or another portable or permanently installed data storage device, can be used for transport to the computer system or control device and / or for storage on or in the computer system or control device, on which the program sections of the computer program that can be read and executed by a computer system are stored. For this purpose, the computer system can, for example, have one or more cooperating microprocessors or the like.

[0036] The dependent claims and the following description each contain particularly advantageous embodiments and developments of the invention. In particular, the claims of one claim category can also be developed analogously to the dependent claims of another claim category. Furthermore, within the scope of the invention, the various features of different embodiments and claims can also be combined to form new embodiments.

[0037] In a preferred embodiment of the method according to the invention for controlling a magnetic resonance imaging system, the duration of the repetition affected by the modification and preferably also the duration of the second pulse sequence subsection, which reproduces the first scanning pattern, remain unchanged when the first scanning pattern is modified. Advantageously, a steady state of the spins deflected with respect to their magnetization, also referred to as a "steady state," is maintained. Such a "steady state" is particularly required for pulse sequences in which the repetition time or the echo time is shorter than the relaxation time of the excited spins. Image artifacts that would be caused by an impairment of the steady state are advantageously avoided.

[0038] If, in a variant of the method according to the invention for controlling a magnetic resonance imaging system, a change in the duration of the affected repetitions is caused by the modification of the first scanning scheme, it may be necessary or advantageous to compensate for this change in the duration by so-called dummy pulses or preparation pulses in order to enable correct acquisition of k-space data or RF resonance signals.

[0039] For example, in such a modification of the first sampling scheme with a changed time duration, the number of k-space lines sampled per repetition is increased in the modified first sampling scheme compared to the unmodified first sampling scheme in order to be able to sample an enlarged k-space area for obtaining the reference data.

[0040] Particularly preferably, all pulse sequence sections or repetitions of the plurality of repetitions of the EPI pulse sequence used for imaging comprise either the first sampling pattern or the modified first sampling pattern. Advantageously, a maximum amount of k-space data is acquired for correcting the artifacts and supplementing the undersampled k-space data.

[0041] It is particularly preferred if only exactly one of these repetitions, preferably the first repetition, has the first sampling pattern and the remaining or all remaining repetitions have the modified first sampling pattern in order to maximize the amount of k-space data used to supplement the undersampled k-space data. In this variant, the resulting image quality is further improved due to the broader database for determining the k-space data missing due to the undersampling. In this variant, only the first repetition is used to acquire uncoded k-space data, preferably navigator data or corresponding k-space data for correcting Nyquist ghost corrections and / or B 0 field corrections.

[0042] However, it can also be very advantageous to use multiple repetitions distributed across the entire pulse sequence scheme to acquire uncoded k-space data, preferably navigator data for correcting Nyquist ghost corrections and / or B0 field corrections, in order to update correction factors for such corrections. The correction factors can advantageously be adapted to dynamic changes that influence the occurrence of artifacts, such as Nyquist ghosting structures.

[0043] Preferably, the first sampling pattern in an EPI pulse sequence is replaced by differently modified sampling patterns in different repetitions.

[0044] In different repetitions, differently modified first scanning schemes are used to scan the k-space. For example, individual gradients must vary to sample different k-space regions. Furthermore, as will be explained in more detail later, it may also be necessary to use multiple differently structured kernels acquired with different sampling schemes to calculate the missing k-space data, for example, if different acquisition techniques are combined or if certain artifacts need to be corrected with particular precision.

[0045] In a preferred variant of the method according to the invention for controlling a magnetic resonance imaging system, the modified first sampling scheme and the number of repetitions comprised by the true subset are selected such that a complete sampling of a k-space sub-area, which is required at least for a complete supplementation of the missing k-space data, is carried out. For the determination of the k-space data initially still missing during subsampling, so-called kernels based on k- Spatial data is calculated from a completely sampled region of k-space. In this case, this is the region of k-space sampled with the modified first sampling scheme.

[0046] Typical values ​​for generating a kernel for supplementing k-space data in a subsampled image acquisition method are 24 phase-space coding lines, for which 8 repetitions are required when choosing a GRAPPA sampling scheme or 16 repetitions when choosing an SMS sampling scheme, if the first sampling scheme is designed to sample 3 phase-space coding lines and this number is not changed during the modification. When selecting a sampling of M phase-space coding lines for Nyquist ghost correction (M is a natural number), 24 / M modified repetitions are required for the GRAPPA sampling scheme and 48 / M modified repetitions are required for the SMS sampling scheme.

[0047] Thus, in the method according to the invention, the modified first sampling scheme preferably comprises a sampling of at least three k-space lines in the readout direction with different phase encoding. Advantageously, three k-space lines are sufficient for a correction in three spatial directions with the N / 2 Nyquist ghost correction. When modifying the first sampling scheme, the number of three gradient pulses in the readout direction does not need to be changed in this advantageous variant. It should be mentioned at this point that the first sampling scheme can also comprise only the sampling of one or two k-space lines or more than three k-space lines in the readout direction. A sampling of k-space lines according to the first unmodified sampling scheme is intended in particular to expressly include the variant in which the same k-space line is possibly scanned multiple times in both directions and is particularly preferably scanned uncoded.Should the number of sampled k-space lines of the modified first sampling scheme differ from the number of sampled . k- If the spatial lines of the unmodified first sampling scheme differ, additional modifications as already mentioned may be useful as compensation in order not to change the echo time and the repetition time of the modified repetitions in comparison to the unmodified repetitions, or to change them too much.

[0048] Particularly preferably, the modified first scanning scheme of the method according to the invention comprises a scanning of preferably at least three k-space lines in the readout direction with at least partially different slice coding.

[0049] This different slice coding is particularly effective in improving image quality in imaging techniques using simultaneous slice imaging, especially SMS, GRAPPA + SMS, and Dual Polarity GRAPPA, for the reasons mentioned above.

[0050] Most preferably, an analogously structured sampling scheme—in other words, a sampling scheme of the same type—is used for the modified first sampling scheme and the second sampling scheme. If a sampling scheme of the same type is used for generating the reference data, i.e., the k-space data on the basis of which subsampled k-space data are to be supplemented, and for sampling the subsampled k-space data itself, interference effects and geometric distortions that would occur with different sampling schemes for acquiring the reference data and the subsampled k-space data can be avoided or reduced.An analog sampling scheme is defined by the fact that for two k-space lines sampled consecutively in time, the distance in k-space between the acquired reference data in one repetition and the subsampled data is the same. In other words, the reference data in one repetition has the same subsampling as the subsampled (imaging) data. However, the reference data combined from all repetitions then naturally results in a fully sampled k-space.

[0051] In a particularly effective embodiment of the method according to the invention for controlling a magnetic resonance imaging system, the k-space data or reference data acquired with the modified first sampling pattern are updated at different times during the accelerated echo-planar imaging. Advantageously, the database for supplementing the subsampled k-space data can be adapted to dynamic effects, in particular interference such as phase changes or patient movements. This adaptation advantageously allows for consistent image quality throughout the entire imaging process.

[0052] For this purpose, a complete sampling of the reference data up to a first point in time can be performed, and this complete sampling can be repeated one or more times at a later point in time. Figuratively speaking, a moving average of reference data can be formed based on the majority of reference data sets, on which the k-space data to be supplemented is determined. Alternatively, "stable" or constant reference data can be determined after multiple acquisitions of the reference data, if an averaging across multiple reference data sets is performed only after all reference data sets have been acquired, or if the reconstruction of the missing k-space data is performed only after all reference data sets have been acquired.

[0053] If kernels are generated based on the reference data to determine the missing k-space data for accelerated imaging, several such updated kernels can be generated using a moving average during a single acquisition or imaging session. This has the advantage of consistently high image quality when calculating the missing k-space data, even when the aforementioned dynamic interference effects occur.

[0054] Alternatively, a single stable kernel can be obtained at the end of the acquisition by sampling or acquiring a part of the k-space required for the reference data, for example the k-space center, or a complete set of the reference data several times during the measurement and then calculating the kernel based on the averaged reference data.

[0055] In a particularly practical embodiment of the method according to the invention, the modified first sampling scheme comprises at least four consecutive phase-encoding gradients and at least three consecutive readout gradients with alternating polarity, which alternate temporally with the phase-encoding gradients. Advantageously, at least three different k-space lines are acquired per repetition for calculating k-space data to compensate for undersampling, thereby accelerating image acquisition.

[0056] Preferably, the second to fourth phase-encoding gradients have a positive polarity. In this variant, the scanning direction is maintained in the phase-encoding direction in one repetition.

[0057] If k-space data is to be acquired simultaneously in multiple slices, slice selection gradients are preferably switched in synchronization with the second to fourth phase-encoding gradients. Simultaneous acquisition can advantageously reduce the acquisition time of magnetic resonance imaging.

[0058] Most preferably, the slice selection gradients have alternating polarity. This scanning scheme is particularly suitable for SMS, GRAPPA + SMS, or dual polarity GRAPPA imaging techniques.

[0059] If an imaging method with subsampling in the phase-encoding direction is selected, for example, a GRAPPA scheme or a combination of a GRAPPA scheme and an SMS scheme, the amplitudes of the phase-encoding gradients of the first modified sampling scheme are preferably selected depending on a subsampling factor of the second sampling scheme. The amplitudes are preferably selected such that, with a subsampling factor of the value F (F is a natural number), F-1 lines are skipped.

[0060] In a preferred embodiment of the method according to the invention for controlling a magnetic resonance imaging system, one of the following specific imaging techniques is used for the accelerated echo planar acquisition: GRAPPA, SMS, SENSE, a combination of GRAPPA and SMS, a dual polarity acquisition, a compressed sensing or SENSE-based imaging technique with a deep learning reconstruction technique, which also requires coil sensitivity information for execution.

[0061] In Kerstin Hammernik et al., Z-net: Systematic Evaluation of Iterative Deep Neural Networks for Fast Parallel MR Image Reconstruction, https: / / arxiv.org / abs / 1912.09278, a Compressed SENSING or SENSE-based imaging method with a deep learning reconstruction method is described.

[0062] When imaging with a GRAPPA method, it is advantageous to first acquire Nyquist ghosting correction lines using the first sampling scheme. By applying the first modified sampling scheme, reference lines can also be acquired for GRAPPA calibration, i.e., for calculating kernels to determine the k-space data that supplements subsampling. By modifying the phase-encoding gradients (Gy gradients), a subsection of the k-space to be acquired is selected, and additional phase blips or smaller phase-encoding gradients are used to navigate further within the navigator formed by the first modified sampling scheme. With this method, a k-space with 24 phase-encoding lines that is well suited for GRAPPA calibration can be acquired with 8 repetitions and an acquisition of 3 phase-encoding lines per repetition using the modified first sampling scheme.In addition to calibrating GRAPPA kernels, the acquired k-space data can also be used to calibrate SENSE methods and, in particular, to obtain coil sensitivity maps for deep learning reconstruction methods.

[0063] When imaging with an SMS method, the first sampling pattern can advantageously be used to obtain Nyquist ghosting correction lines, while the first modified sampling pattern can be used to obtain reference lines for SMS calibration, i.e., for calculating kernels to determine the k-space data that supplements subsampling. This advantageously allows for the acquisition of reference lines for SMS kernel calibration.

[0064] In a specific embodiment of such a method, which is based on the SMS sampling technique, reference data is determined for the Slice GRAPPA kernel calculation. Slice GRAPPA is the common reconstruction method for SMS recordings, see z.B. https: / / onlinelibrary.wiley.com / doi / full / 10.1002 / mrm.23097 The data points to be recorded are selected and arranged such that a slice blip can be treated as a shift of the data in the slice encoding direction (kz direction). This procedure is described in more detail in US patent US 10557903 B2.

[0065] If neighboring k-space points in the ky and kz directions are designated by consecutive integers 1, 2, 3, etc., then when acquiring three phase encoding lines with the first sampling pattern, the data points kz=1 / ky=1, kz=2 / ky=2, kz=1 / ky=3 are acquired in the first repetition with the modified first sampling pattern. This is achieved by alternating blips in the kz direction and unipolar blips in the ky direction. Blips here are gradient pulses that represent a short trajectory section. In the following repetition, the complementary k-space section kz=2 / ky=1, kz=1 / ky=2, kz=2 / ky=3 is acquired. The polarity of the kz blips is reversed accordingly. Subsequently, further sections are acquired. Alternatively, the recording can also be performed "row by row" in the kz direction, ie a single shift along the kz axis is made before the first echo and then ky=1,2,3 is recorded with fixed kz values.As described in US patent US 10557903 B2, slice-specific calibration data can then be obtained by Fourier transformation along the kz axis. To obtain a kernel with 24 k-space lines, 16 repetitions are required with the first sampling scheme, acquiring three phase encoding lines per repetition.

[0066] Advantageously, in imaging with a combined GRAPPA + SMS method, Nyquist ghosting correction lines can be obtained by the first sampling scheme and reference lines for GRAPPA + SMS calibration, i.e. for calculating kernels to determine the k-space data that supplements subsampling, can be obtained by the first modified sampling scheme.

[0067] Reference lines are advantageously acquired for combined GRAPPA + SMS calibration. In this way, reference data can be acquired for both GRAPPA and SMS. Compared to a variant aimed at a pure SMS imaging method, this variant is modified in that the spacing of the phase-encoding lines in the ky direction, i.e., in the phase-encoding direction, corresponds to the subsampling factor, for example, 2. To calibrate the so-called slice GRAPPA kernels, the acquired k-space data are then rearranged according to DE 10 2017 209 988 B3 so that the subsampled acquisition pattern is retained. To calibrate the conventional GRAPPA kernels, however, the phase-encoding lines are adopted as a block, as described above.

[0068] With the different subsampling methods mentioned above, other arrangements of the k-space points acquired in the repetitions are also conceivable, as long as a fully sampled k-space region is achieved at the end.

[0069] For a GRAPPA + SMS calibration, 32 repetitions are already required for two kernels with 24 k-space lines each, with an acquisition of 3 phase encoding lines per repetition with the first sampling scheme.

[0070] If the second sampling scheme comprises a dual-polarity scan, for example, in a dual-polarity GRAPPA imaging method, the first modified sampling scheme comprises at least two sampling schemes with mutually inverted readout polarities for each sampled phase. For this purpose, two consecutive repetitions with differently modified first sampling schemes with mutually inverted readout polarities of the phase-encoding gradients and otherwise identical gradient schemes can preferably be used. Compared to a standard GRAPPA method, the number of repetitions required to determine the reference data is thus doubled with this first sub-variant.

[0071] In a second sub-variant of the dual-polarity GRAPPA imaging technique, the k-space data is acquired using the modified first sampling scheme with mutually inverted readout polarity, by scanning each phase with alternating polarity multiple times, preferably three times. In this sub-variant, only one phase encoding line is acquired per repetition when acquiring the reference data. Thus, with an unmodified first sampling scheme with three phase encoding lines per repetition, the number of repetitions required for acquiring the reference data is tripled for acquiring Nyquist ghost correction data.

[0072] If, in the method according to the invention, an imaging method for simultaneously reading out several layers is used as an accelerated EPI imaging method, such as the aforementioned imaging methods, then in a particularly effectively realizable embodiment of the method, the number N of layers to be read out simultaneously is equal to 2.

[0073] Alternatively, the number N of slices to be read simultaneously can be 3. In principle, it would be optimal to read as many slices as possible simultaneously. However, the number of slices that can be read simultaneously is limited by the following circumstances: The energy delivered to the patient per pulse is proportional to the number of slices N. However, the energy delivery per time is limited. Furthermore, the more slices are read simultaneously, the more difficult separation in image reconstruction becomes, as the noise in the separated images increases. This increase is nonlinear. Therefore, in practice, it is only possible to read a few slices simultaneously.

[0074] The invention is explained in more detail below with reference to exemplary embodiments in the accompanying figures. They show: FIG 1 a diagram illustrating a conventional procedure for a Nyquist ghost correction, FIG 2 a diagram illustrating a conventional method for determining B 0 field maps, in which a different k-space section to be acquired is encoded in each repetition using phase encoding gradients, which are then combined to form matching k-spaces, FIG 3 a diagram showing a pulse sequence for implementing the FIG 2 illustrated imaging procedure shows FIG 4 a scheme for obtaining reference lines for GRAPPA calibration in a method for controlling a magnetic resonance imaging system for generating magnetic resonance image data of an examination subject, in which k-space data are acquired in the context of accelerated echo-planar imaging with subsampling, according to an embodiment of the invention, FIG 5 a diagram showing a pulse sequence for implementing the FIG 4 illustrated scanning schemes in an echo-planar imaging procedure, FIG 6 an alternative scheme for obtaining reference lines for GRAPPA calibration in a method for controlling a magnetic resonance imaging system for generating magnetic resonance image data of an examination subject, in which k-space data are acquired in the context of accelerated echo-planar imaging with subsampling, according to an embodiment of the invention, FIG 7 a diagram showing a pulse sequence for implementing the FIG 6 illustrated scanning schemes in an echo-planar imaging procedure, FIG 8 a scheme for obtaining reference lines for SMS calibration in a method for controlling a magnetic resonance imaging system for generating magnetic resonance image data of an examination subject, in which k-space data are acquired in the context of accelerated echo-planar imaging with subsampling, according to an embodiment of the invention, FIG 9 a diagram showing a pulse sequence for implementing the FIG 8 illustrated scanning schemes in an echo-planar imaging procedure, FIG 10 a scheme for obtaining reference lines for combined GRAPPA + SMS calibration in a method for controlling a magnetic resonance imaging system for generating magnetic resonance image data of an examination subject, in which k-space data are acquired in the context of accelerated echo-planar imaging with subsampling, according to an embodiment of the invention, FIG 11 a diagram showing a pulse sequence for implementing the FIG 10 illustrated scanning schemes in an echo-planar imaging procedure, FIG 12 a scheme for obtaining reference lines of a first kernel for a dual-polarity GRAPPA calibration in a method for controlling a magnetic resonance imaging system for generating magnetic resonance image data of an examination subject, in which k-space data are acquired in the context of accelerated echo-planar imaging with subsampling, according to an embodiment of the invention, FIG 13 a diagram showing a pulse sequence for implementing the FIG 12 illustrated sampling scheme shows FIG 14 a scheme for obtaining reference lines of a second kernel for a dual-polarity GRAPPA calibration in a method for controlling a magnetic resonance imaging system for generating magnetic resonance image data of an examination subject, in which k-space data are acquired in the context of accelerated echo-planar imaging with subsampling, according to an embodiment of the invention, FIG 15 a diagram showing a pulse sequence for implementing the FIG 14 illustrated sampling schemes, and Figur 16 a magnetic resonance imaging system according to an embodiment of the invention.

[0075] In FIG 1 Figure 10 shows a flowchart illustrating a conventional procedure for implementing Nyquist ghost correction in a magnetic resonance imaging method. In step 1.I, three phase encoding lines are acquired in the k-space center during the first three repetitions REP1, REP2, and REP3 as part of a phase correction acquisition. The unencoded phase encoding lines acquired in the k-space center during the first three repetitions REP1, REP2, and REP3 are subsequently averaged in step 1.II to obtain k-space data for a navigator acquisition NA. Based on this averaged k-space data, Nyquist ghost effects are corrected or eliminated from image data in step 1.III. This step 1.III is also symbolized by a graphic representation below the flowchart, with the artifact-affected image of a brain shown on the left and the corrected image on the right. In this conventional variant, the Nyquist ghost correction lines are only sampled in the first three repetitions, and the resulting correction factors are then applied to all other repetitions of the actual imaging procedure. The values ​​shown in . FIG 1 The method for correcting Nyquist ghost effects illustrated is explained in detail in US 604 365 1 B2.

[0076] It should be mentioned here that it is also conventional practice to sample three uncoded k-space lines with alternating polarity as navigator data in each repetition before sampling the k-space data for imaging. FIG 1 The variant described is based on the FIG 2 The acquisition of phase-encoded k-space lines described in

[19] was adapted to obtain B 0 field maps. FIG 2 illustrated repetitions REP4, ..., In this particular case, REPM follow the FIG 1 illustrated the first three repetitions REP1, REP2, REP3.

[0077] In FIG 2 Figure 20 shows a conventional method for image-based field map determination. FIG 2 The procedure illustrated corresponds to the approach in Chi et al., "Field-Mapping-Embedded EPI for Geometric Distortion Correction," Proc. Intl. Soc. Mag. Reson. Med. 30 (2022) 1099.

[0078] In this method, a different k-space section to be acquired is encoded in each repetition using phase encoding gradients, which are then combined to form matching k-spaces (see steps 2.I, 2.II). In step 2.I, a series of repetitions REP4, REP5,..., REPM, in which k-space data or raw data with different phase encoding are acquired slice by slice from an object being examined as part of an echo planar imaging. Unlike the FIG 1 In the method illustrated, in addition to subsampled k-space data for imaging, a phase encoding line is also acquired at each repetition for the creation of a subsequent field map, with a different phase being encoded for each repetition. In step 2.II, the subsampled k-space lines acquired in step 2.I are rearranged. In addition, in step 2.II, additional reference data REF is acquired, which is later used to calibrate kernels to determine k-space data missing from the acquisition of the k-space data in step 2.I. It should be expressly pointed out at this point that the reference data is acquired in a separate acquisition. Therefore, two separate acquisitions are performed: one acquisition for the subsampled k-space data and the phase encoding lines, and a separate acquisition of the reference data REF.

[0079] Subsequently, in step 2.III, the missing k-space data for the respective layers are determined based on the kernels determined in step 2.II and the subsampled k-space data.

[0080] In step 2.IV, layer-by-layer image reconstructions are performed using a two-dimensional fast Fourier transform (2D-FFT) based on the completed k-space data.

[0081] In step 2.V, a phase difference (for example the phase development between the first and third echo Δϕ 31 , which are recorded with the same polarity) and a total phase offset ϕ 0 are calculated on the basis of the plurality of reconstructed slice images.

[0082] Subsequently, in step 2.VI, the image data from the individual channels are combined, and in step 2.VII, phases are unwrapped to create the field maps. This eliminates phase jumps in the image. Because the phase is periodic at 360 degrees, ambiguities can arise between pixels (i.e., one phase difference becomes 0, the neighboring one becomes 360°).

[0083] In FIG 3 is a diagram 30 showing a pulse sequence sequence for implementing the FIG 2 The diagram illustrates a method for image-based determination of field maps. The diagram comprises four sub-diagrams shown one below the other, with the first sub-diagram showing an excitation radio-frequency pulse RF, the second sub-diagram a series of readout gradients Gx, the third sub-diagram a series of phase encoding gradients Gy, and the fourth sub-diagram a series of slice selection gradients Gz. It should be noted at this point that the received radio-frequency pulses, which represent the received measurement signals, are not illustrated in this representation.

[0084] On the left side of the diagram, an excitation scheme A-RF with an RF excitation pulse RF and a slice selection gradient Gz played simultaneously is shown.

[0085] In the center of the diagram, between two vertical dashed lines, a sampling scheme PAS is shown for sampling phase-shifted phase-encoding lines for a navigator acquisition NA for acquiring k-space data for Nyquist ghosting correction. To position a desired starting position in k-space, a frequency encoding gradient Gx, a slice selection gradient Gz, and a phase encoding gradient Gy are first switched on simultaneously. Subsequently, three frequency encoding gradients Gx are switched on, alternating in polarity. During the generation of the phase correction data, the polarity of the readout gradients is mirrored in the kx direction by changing the polarity of the readout gradients on the axis in the direction of the readout gradient Gx. This is necessary for obtaining three unipolar k-spaces for the subsequent image-based calculation of the field maps.

[0086] At the end of the middle scanning scheme PAS, a phase encoding gradient Gy and a slice encoding gradient Gz are switched.

[0087] To the right of this, an EPI pulse sequence EPI with a slice GRAPPA sampling scheme is shown.

[0088] In FIG 4 a scheme 40 for obtaining reference lines for GRAPPA calibration in a method for controlling a magnetic resonance imaging system for generating magnetic resonance image data of an examination subject, in which k-space data are acquired in the context of accelerated echo-planar imaging with GRAPPA subsampling, is shown according to an embodiment of the invention.

[0089] As already explained in more detail in the general section, within the scope of the method according to the invention, a first scanning scheme (not shown) is used for the uncoded recording of three phase encoding lines for a correction of Nyquist ghost effects, and a corresponding plus sequence is played out in a first repetition as part of an imaging process. For the following repetitions, the scanning scheme is now used as in FIG 4 presented, modified.

[0090] For this purpose, the first acquisition scheme for obtaining reference lines for kernels, which are required to detect missing data due to undersampling in the GRAPPA imaging method, is modified by adding additional so-called phase blips or phase encoding gradients. FIG 3 The phase encoding gradients or Gy gradients shown are used to select a section REP1, REP2, REP3 of the k-space to be acquired and by means of additional phase blips, which are shown in FIG 5 As illustrated, the scanning process continues between the individual phase-encoding lines of a repetition. With this method, a k-space with 24 phase-encoding lines, well-suited for GRAPPA calibration, can be obtained with just eight repetitions.

[0091] In FIG 5 A diagram 50 is shown which shows a pulse sequence sequence for implementing the FIG 4 illustrated scanning schemes in an echo-planar imaging technique. First, in the FIG 5 Figure 50 illustrates what has already been FIG 3 The following modified first scanning scheme AS1, arranged between dashed vertical lines, reflects the excitation scheme A-RF shown in FIG 4 The second repetition REP2 shown here is used to acquire reference data. For positioning at the starting point of the FIG 4 In the trajectory of the second repetition REP2 shown, a frequency encoding gradient Gx, a slice selection gradient Gz, and a phase encoding gradient Gy are initially played out simultaneously. The phase encoding gradients Gy, drawn one above the other at the same location with different amplitudes (once thick and once dashed), are assigned to different repetitions. Depending on the amplitude strength, a different phase in k-space is controlled for different repetitions. Subsequently, similar to the FIG 3 The pulse sequence shown comprises three frequency coding gradients Gx with alternating polarity, each followed by a small second or third phase coding gradient Gy, referred to as a phase blip. Finally, the modified first excitation pattern AS1 is completed by an additional frequency coding gradient Gx and a simultaneous fourth phase coding gradient Gy to arrive at the starting point of the trajectory forming the echo train of the actual EPI pulse sequence EPI for echo-planar imaging A-EPI. Here, too, the phase coding gradient Gy concluding the modified first scanning pattern is again drawn with a plurality of superimposed phase coding gradients Gy, each of which has different amplitudes and is assigned different repetitions.The EPI pulse sequence EPI represents a second scanning scheme AS2 and, similar to the first scanning scheme AS1 modified by the increased phase encoding gradients Gy, has uniform, but negatively polarized phase encoding gradients Gy.

[0092] It should be noted that in the FIG 5 illustrated embodiment which is also in FIG 4 The repetitions REP1, ..., REP3 shown are preceded by an unmodified repetition (not shown) with which three uncoded phase space lines with alternating polarity are recorded as navigator data for a Nyquist ghost correction. In this preceding repetition, the FIG 5 The phase encoding gradients drawn in the first scanning scheme, especially the second and third phase encoding gradients, also called phase blips, are eliminated because the navigator data is recorded uncoded. In contrast to the FIG 2 In contrast to the conventional methods illustrated, the acquisition of the navigator data, the reference data, and the k-space data for echo-planar imaging is carried out with a single pulse sequence and a single acquisition. As already mentioned above, integrating the acquisition of the reference data into the echo-planar imaging acquisition process can reduce the overall acquisition time.

[0093] In FIG 6 An alternative scheme 60 for obtaining reference lines for GRAPPA calibration is shown in a method for controlling a magnetic resonance imaging system for generating magnetic resonance image data of an examination subject, in which k-space data are acquired as part of an accelerated echo-planar imaging with subsampling, according to an embodiment of the invention. Unlike in the FIG 4 In Scheme 40 shown, the successive phase coding lines of a repetition REP1, REP2, REP3 are each arranged offset by three phase coding lines.

[0094] In FIG 7 is a diagram 70 with a pulse sequence for implementing the FIG 6 illustrated scanning schemes in an echo-planar imaging procedure.

[0095] In FIG 7 First of all, the already FIG 3 and FIG 5 The following modified first scanning scheme AS1 reflects the excitation scheme A-RF shown in FIG 6 The first repetition REP1 shown in the figure is reflected in the figure. For positioning at the starting point of the FIG 6 In the trajectory of the first repetition REP1 shown, a frequency encoding gradient Gx, a slice selection gradient Gz and a phase encoding gradient Gy are simultaneously played out. This is followed by the following steps, similar to the FIG 5 The pulse sequence shown in Figure 1 consists of three frequency coding gradients Gx with alternating polarity, with the first two frequency coding gradients Gx being followed by a phase coding gradient Gy, referred to as a phase blip. This phase coding gradient has a FIG 5 The phase-encoding gradients Gy shown occupy an area three times larger, as they are used to skip two adjacent phase-encoding lines. Finally, the modified first excitation pattern AS1 is terminated by an additional phase-encoding gradient to reach the starting point of the trajectory forming the echo train of the actual EPI pulse sequence EPI. The EPI pulse sequence EPI represents a second scanning pattern AS2 and, similar to the first scanning pattern AS1 modified by the increased phase-encoding gradients Gy, has uniform, albeit negatively polarized, phase-encoding gradients.

[0096] In FIG 8 a scheme 80 for obtaining reference lines for SMS calibration in a method for controlling a magnetic resonance imaging system for generating magnetic resonance image data of an examination subject, in which k-space data are acquired in the context of accelerated echo-planar imaging with SMS subsampling, is shown according to an embodiment of the invention.

[0097] Reference data for the slice GRAPPA kernel calculation in SMS imaging is shown. The data points to be acquired are presented here in such a way that a slice blip Gz can be treated as a shift of the data in the kz direction (cf. US 10557903 B2). In the first repetition REP1, the data points kz=1 / ky=1, kz=2 / ky=2, kz=1 / ky=3 are acquired. The order of sampling of the different k-space lines within a repetition is symbolized by the numbers 1, 2, 3 assigned to these k-space lines. This is done in FIG 9 by alternating blips Gz in kz direction and unipolar blips Gy in ky direction.

[0098] In the second repetition, REP2, the complementary k-space section kz=2 / ky=1, kz=1 / ky=2, kz=2 / ky=3 is acquired. The polarity of the kz blips Gz is reversed accordingly. Subsequently, further sections of k-space are acquired with the third repetition, REP3, and the fourth repetition, REP4.

[0099] Alternatively, the acquisition can also be performed "row-wise" in the kz direction, i.e., a single shift along kz is performed before the first echo, and then ky=1, 2, 3 is acquired with fixed kz. As described in US 10557903 B2, slice-specific calibration data can then be obtained by Fourier transformation along kz. To obtain a kernel with 24 k-space lines, 16 repetitions are required.

[0100] In FIG 9 is a diagram 90 with a pulse sequence sequence for implementing the FIG 8 illustrated scanning schemes in an echo-planar imaging procedure.

[0101] In FIG 9 First of all, the already FIG 3 , FIG 5 and FIG 7 The following modified first scanning scheme AS1 reflects the excitation pattern A-RF shown in FIG 8 The first repetition REP1 shown in the figure is reflected in the figure. For positioning at the starting point of the FIG 8 In the trajectory of the first repetition REP1 shown, a frequency encoding gradient Gx, a slice selection gradient Gz and a phase encoding gradient Gy are simultaneously played out. This is followed by the following steps, similar to the FIG 7 The pulse sequence shown in Figure 1 consists of three frequency coding gradients Gx with alternating polarity, with the first two frequency coding gradients Gx each being followed by a positively polarized phase coding gradient Gy, referred to as a phase blip. This phase coding gradient Gy has a FIG 7 The phase encoding gradients Gy shown here are only one-third the size, as they are used to control neighboring phase encoding lines. Alternatingly polarized slice selection gradients Gz are switched simultaneously with the phase encoding gradients Gy, referred to as phase blips.

[0102] Finally, the modified first excitation scheme AS1 is terminated by an additional phase encoding gradient Gy, which simultaneously switches a final slice selection gradient Gz, in order to arrive at the starting point of the trajectory forming the echo train of the actual EPI pulse sequence EPI. The EPI pulse sequence EPI represents a second Ab scanning scheme AS2 and, similar to the modified first scanning scheme AS1, has a modified scanning scheme characterized by uniform, but negatively polarized phase encoding gradients Gy and their polarization changing slice selection gradients Gz.

[0103] In FIG 10 a scheme 100 for obtaining reference lines for combined GRAPPA + SMS calibration in a method for controlling a magnetic resonance imaging system for generating magnetic resonance image data of an examination subject, in which k-space data are acquired in the context of accelerated echo-planar imaging with subsampling, according to an embodiment of the invention is shown.

[0104] With the FIG 10 The procedure illustrated can be used to obtain reference data for both GRAPPA and SMS. Compared to the FIG 8 In the sampling scheme 80 illustrated, the method is modified such that the distance in the ky direction corresponds to the subsampling factor (here 2). The pattern of the lines (dotted, thin, medium, thick) of the brackets assigned to the individual repetitions REP1, ..., REP4 corresponds to the pattern of the circumference lines of the k-space lines assigned to the respective repetitions. To calibrate the slice GRAPPA kernels, the acquisition data are then rearranged according to DE 10 2017 209 988 B3 so that the subsampled acquisition scheme is retained. To calibrate the conventional GRAPPA kernels, the lines are adopted as a block. As with the FIG 8 In addition to the method illustrated, other arrangements of the k-space points acquired in repetitions REP1, REP2, REP3, and REP4 are also conceivable, as long as a fully sampled area is achieved at the end. Here, 32 repetitions are required for two kernels with 24 k-space lines each.

[0105] In FIG 11 is a diagram 110 showing a pulse sequence sequence for implementing the FIG 10 illustrated imaging procedure.

[0106] In FIG 11 First of all, the already FIG 3 , FIG 5 , FIG 7 and FIG 9 The previously illustrated excitation scheme A-RF is shown. The following modified first scanning scheme AS1 reflects the FIG 10 The first repetition REP1 shown in the figure is reflected in the figure. For positioning at the starting point of the FIG 10 In the trajectory of the first repetition REP1 shown, a frequency encoding gradient Gx, a slice selection gradient Gz and a phase encoding gradient Gy are simultaneously played out. This is followed by the following steps, similar to the FIG 9 The pulse sequence shown in Figure 1 consists of three frequency coding gradients Gx with alternating polarity, with the first two frequency coding gradients Gx being followed by a positively polarized phase coding gradient Gy, referred to as a phase blip. This phase coding gradient has a FIG 9 The phase encoding gradients Gy shown in the figure double the area, since they are used to skip one phase encoding line in the ky direction and to control the next phase encoding line. Simultaneously with the phase encoding gradients Gy, referred to as phase blips, the phase encoding gradients Gy are applied as shown in FIG 9 alternately polarized slice selection gradients Gz switched.

[0107] Finally, the modified first excitation pattern AS1 is terminated by an additional phase-encoding gradient Gy, which simultaneously switches a final slice selection gradient Gz, to arrive at the starting point of the trajectory forming the echo train of the actual EPI pulse sequence EPI. The EPI pulse sequence EPI represents a second scanning pattern AS2 and, similar to the modified first scanning pattern AS1, features a modified scanning pattern characterized by uniformly, albeit negatively polarized, phase-encoding gradients Gy and slice selection gradients Gz that vary in polarization.

[0108] In FIG 12 A scheme 120 for obtaining reference lines of a first kernel for a dual-polarity GRAPPA calibration in a method for controlling a magnetic resonance imaging system for generating magnetic resonance image data of an examination subject, in which k-space data is acquired as part of accelerated echo-planar imaging with subsampling, is shown according to an embodiment of the invention. The sampling of k-space for the first kernel is very similar to that in FIG 4 shown scanning scheme 40 and is therefore not explained in detail again.

[0109] In FIG 13 is a diagram 130 showing a pulse sequence sequence for implementing the FIG 12 illustrated scanning scheme 120. The FIG 13 The pulse sequence shown is analogous or identical to that shown in FIG 5 shown pulse sequence 50 and is therefore not explained in detail again.

[0110] In FIG 14 A scheme 140 for obtaining reference lines of a second kernel for a dual polarity GRAPPA calibration in a method for controlling a magnetic resonance imaging system for generating magnetic resonance image data of an examination subject, in which k-space data are acquired in the context of accelerated echo-planar imaging with subsampling, is shown according to an embodiment of the invention. The scanning direction of the phase encoding lines is in FIG 14 just the opposite of that in FIG 12 shown sampling scheme 120.

[0111] In FIG 15 is a diagram 150 showing a pulse sequence of the FIG 14 The scanning scheme shown in FIG 15 Pulse sequence shown differs from that in FIG 13 shown pulse sequence only by the fact that the readout gradients Gx of the modified scanning scheme AS1 have a FIG 13 have reversed polarity and a fourth readout gradient Gx is omitted in the modified first scanning scheme AS1.

[0112] As described by Chi et al., "Field-Mapping-Embedded EPI for Geometric Distortion Correction", Proc. Intl. Soc. Mag. Reson. Med. 30 (2022) 1099, with the illustrated acquisition schemes, it is preferable to first acquire at least one uncoded navigator acquisition or one acquisition for Nyquist ghost correction with a first unmodified sampling scheme to further enable ghosting correction. A combination with the previously discussed ghosting correction methods for SMS imaging is also suitable for this purpose (see US 10162037 B2 or US 11280870 B2). If only one such acquisition is performed for ghosting correction, the correction is static, as with the conventional "external phase correction scan," meaning dynamic changes cannot be incorporated. Since drift effects usually develop slowly over time, interspersing these acquisitions, e.g., with a phase correction scan, is not recommended. B. every eight repetitions and an update of the ghosting correction factors e.g.B. using a moving average, is effective in compensating for these slow drift effects.

[0113] As already explained, it is also conceivable to obtain multiple GRAPPA and / or SMS kernels for very long image acquisitions, e.g., using moving averages. This has the advantage of consistently high image quality even in the presence of interference such as phase changes or patient movement.

[0114] Alternatively, a stable kernel can be obtained by sampling a part (e.g., the k-space center) or the complete calibration k-space several times during the measurement and performing the calibration on all averaged data.

[0115] The acquisition schemes outlined in the figure description were described based on a navigator acquisition consisting of three lines. This has the advantage that the modification does not result in a change in the repetition time TR and echo time TE. However, a minimum number of repetitions is required to acquire complete data sets. Of course, it is also conceivable to increase the number of navigator lines, at the expense of a slight increase in the echo time TE and potentially the repetition time TR, in order to acquire more reference lines per repetition.

[0116] In FIG 16 A magnetic resonance system 160 according to the invention (hereinafter referred to as "MR system") is shown in a roughly schematic manner. It comprises, on the one hand, the actual magnetic resonance scanner 102 with an examination room 103 or patient tunnel, into which a patient or examination subject O, in whose body, for example, a specific organ to be imaged is located, can be moved on a couch 108.

[0117] The magnetic resonance scanner 102 is equipped in the usual way with a basic field magnet system 104, a gradient system 106 as well as an RF transmit antenna system 105 and an RF receive antenna system 107. In the illustrated embodiment, the RF transmit antenna system 105 is a whole-body coil permanently installed in the magnetic resonance scanner 102, whereas the RF receive antenna system 107 consists of local coils to be arranged on the patient or subject (in FIG 16 symbolized only by a single local coil). In principle, however, the whole-body coil of the RF transmit antenna system 105 can also be used as an RF receive antenna system and the local coils 107 as an RF transmit antenna system, provided that these coils can each be switched to different operating modes.

[0118] The MR system 160 further comprises a central control device 113, which is used to control the MR system 160. This central control device 113 includes a sequence control unit 114 for pulse sequence control. This unit controls the temporal sequence of radio-frequency pulses (RF pulses) and gradient pulses as a function of a selected imaging sequence PS according to a pulse sequence scheme PSS. Such an imaging sequence PS or the pulse sequence scheme PSS underlying the imaging sequence PS can be specified, for example, within a measurement or control protocol P. Typically, various control protocols P for different measurements are stored in a memory 119 and can be selected by an operator (and modified if necessary) and then used to perform the measurement. The sequence control unit 114 also includes a control sequence determination device 91 according to the invention.The control sequence determining means 91 generates control data SD which contains the control sequences shown in . FIG 5 , FIG 7 , FIG 9 , FIG 11 , FIG 13 or FIG 15 shown sequence of excitation modules and readout modules, which the sequence control unit 114 outputs for pulse sequence control.

[0119] To output the individual RF pulses, the central control device 113 has a radio-frequency transmission device 115, which generates the RF pulses, amplifies them, and feeds them into the RF transmission antenna system 105 via a suitable interface (not shown in detail). To control the gradient coils of the gradient system 106, the control device 113 has a gradient system interface 116. The sequence control unit 114 communicates in a suitable manner, e.g., by transmitting sequence control data SD, with the radio-frequency transmission device 115 and the gradient system interface 116 to transmit the pulse sequences PS in the sequence generated by the control sequence determination device 91. The control device 113 also has a radio frequency receiving device 117 (which also communicates with the sequence control unit 114 in a suitable manner) in order to acquire magnetic resonance signals received by the RF transmit antenna system 107 in a coordinated manner.A reconstruction unit 118 receives the acquired data after demodulation and digitization as raw data or k-space data RD and reconstructs the MR image data from it. This image data BD can then be stored, for example, in a memory 119.

[0120] The central control unit 113 can be operated via a terminal with an input unit 111 and a display unit 109, via which the entire MR system 160 can thus also be operated by an operator. MR images can also be displayed on the display unit 109, and using the input unit 111, possibly in combination with the display unit 109, measurements can be planned and started, and in particular, suitable control protocols with suitable measurement sequences can be selected and modified if necessary, as explained above.

[0121] The MR system 160 according to the invention and in particular the control device 113 can also have a large number of further components, not shown in detail here, but usually present on such devices, such as a network interface in order to connect the entire system to a network and to be able to exchange raw data RD and / or image data BD or parameter maps, but also other data, such as patient-relevant data or control protocols.

[0122] How suitable raw data RD can be acquired and MR images BD reconstructed from them by irradiating RF pulses and generating gradient fields is generally known to the person skilled in the art and will not be explained in more detail here.

[0123] From the above description, it is clear that the invention effectively provides possibilities for improving a method for controlling a magnetic resonance imaging system for generating magnetic resonance image data with regard to the required time duration.

[0124] It should be noted that the features of all embodiments or developments disclosed in figures can be used in any combination.

[0125] Finally, it is emphasized once again that the detailed methods and structures described above are exemplary embodiments and that the basic principle can be varied widely by a person skilled in the art without departing from the scope of the invention, as long as it is defined by the claims. For the sake of completeness, it is also emphasized that the use of the indefinite articles "ein" or "eine" does not exclude the possibility that the features in question may be present multiple times. Likewise, the term "unit" does not exclude the possibility that it consists of multiple components, which may also be spatially distributed. Regardless of the grammatical gender of a particular term, this includes persons with male, female, or other gender identities.

Claims

1. Method for actuating a magnetic resonance imaging system for generating magnetic resonance image data (BD) of an examination object (O), in which k-space data (RD) is acquired as part of an accelerated echoplanar imaging (A-EPI) with undersampling according to a pulse sequence diagram (PSS), which pulse sequence diagram (PSS) has a plurality of repetitions, each of which comprising - a first sampling diagram (AS1) for a recording (NA) of k-space data (RD) for Nyquist ghosting correction or for generating magnetic field maps, - a subsequent second sampling diagram (AS2) for an accelerated echoplanar recording (A-EPI) and - an excitation diagram (A-RF) shared by both recordings (NA, A-EPI), wherein the first sampling diagram (AS1) of a real subset (REP1, REP2, REP3, REP4) of the plurality of repetitions is modified in that, on the basis of the k-space data (RD) acquired using the modified first sampling diagram (AS1), it is made possible to expand k-space data (RD) that is missing as a consequence of the undersampling and / or to correct artifacts occurring due to the undersampling in the image space.

2. Method according to claim 1, wherein, during the modification of the first sampling diagram (AS1), the duration of the repetition (REP1, REP2, REP3, REP4) affected by the modification is maintained in an unaltered manner.

3. Method according to claim 1 or 2, wherein the modified first sampling diagram (AS1) and the number of the repetitions (REP1, REP2, REP3, REP4) comprised by the real subset are chosen such that a k-space subregion, which is required at least for a full expansion of the missing k-space data (RD), is sampled in full.

4. Method according to one of the preceding claims, wherein the modified first sampling diagram (AS1) comprises a sampling of k-space lines in the readout direction (kx) with different phase encoding (ky).

5. Method according to one of the preceding claims, wherein the modified first sampling diagram (AS1) comprises a sampling of k-space lines in the readout direction (kx) with at least partially different slice encoding (kz).

6. Method according to one of the preceding claims, wherein the same type of sampling diagram (AS) is used for the modified first sampling diagram (AS1) and the second sampling diagram (AS2).

7. Method according to one of the preceding claims, wherein at different points in time of the accelerated echoplanar imaging (A-EPI), the k-space data (RD) acquired using the modified first sampling diagram (AS1) is updated.

8. Method according to one of the preceding claims, wherein the modified first sampling diagram (AS1) comprises: - at least four successive phase encoding gradients (Gy) and - at least three successive readout gradients (Gx) with alternating polarity, which alternate with the phase encoding gradients (Gy) over time.

9. Method according to claim 8, wherein slice selection gradients (Gz) are switched in a manner synchronised with the second to fourth phase encoding gradients (Gy) over time.

10. Method according to claim 8 or 9, wherein the amplitudes of the phase encoding gradients (Gy) of the modified first sampling diagram (AS1) are chosen as a function of an undersampling factor of the second sampling diagram (AS2).

11. Method according to one of the preceding claims, wherein one of the following specific imaging techniques is used for the accelerated echoplanar imaging: - GRAPPA, - SMS / Slice-GRAPPA, - SENSE, - a combination of GRAPPA and SMS, - a dual-polarity acceleration technique.

12. Control facility (113) for a magnetic resonance imaging system (160), which is embodied for controlling the magnetic resonance imaging system (160) by using a method according to one of claims 1 to 11.

13. Magnetic resonance imaging system (160), having a control facility (113) according to claim 12.

14. Computer program product with a computer program, which can be loaded directly into a memory unit in a memory of a control facility (113) of a magnetic resonance imaging system (160), with program code sections for carrying out all steps of a method according to one of claims 1 to 11 when the program is executed in the control facility (113).

15. Computer-readable medium, on which program sections that can be executed by a computer unit are stored, for carrying out the steps of a method according to one of claims 1 to 11 when the program sections are executed by the computer unit.