Determining excitation pulses for the simultaneous acquisition of at least two parallel slices during a magnetic resonance imaging (MRI) scan.

Abstract

The invention relates to a computer-implemented method for determining excitation pulses for the simultaneous acquisition of at least two parallel slices during a magnetic resonance imaging measurement with a magnetic resonance imaging system (4), wherein the method comprises the following steps: - Defining a target position (22) for each of the at least two parallel layers; - Determining theoretical gradient values ​​at the target positions (22) under the assumption of an ideal, in particular linear, local gradient profile (11); - Determining actual gradient values ​​at the target positions (22) based on a predetermined location-dependent real gradient value distribution; - Determining a frequency offset for each of the parallel layers based on a difference between the actual gradient value and the theoretical gradient value; - Determining an applicable pulse frequency of an excitation pulse as a superposition of individual pulses of the parallel layers, wherein the frequency of the individual pulses is based on the frequency offset and the theoretical gradient value of the respective parallel layers.

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 computer-implemented method for determining excitation pulses for the simultaneous acquisition of at least two parallel slices during a magnetic resonance tomography measurement, a method for the simultaneous acquisition of at least two parallel slices in a magnetic resonance tomography measurement, a corresponding computer program product and a magnetic resonance tomography system.

[0003] Magnetic resonance imaging (MR) systems employ a gradient system for spatial encoding of the MR signal. Spatial encoding is typically based on linear gradient fields. However, due to design compromises that are usually necessary or advantageous, gradient systems generally exhibit a degree of nonlinearity, particularly in the peripheral regions of the image area of ​​the respective MR scanner. This nonlinearity of the gradient system results in inaccurate excitation of the slices, meaning that the actual slice position does not exactly match the desired slice position.

[0004] When exciting multiple slices simultaneously, the slices are excited with a single pulse, which is a combination of two or more individual pulses that have a frequency shift relative to each other. This type of MR imaging is also known as SMS (Simultaneous Multi-Slice) imaging. Assuming a linear gradient G z The frequency f(z) of the individual pulses for exciting a layer at position z can be given by f(z) = 2π×γ(B 00 +G z ×z), with the gyromagnetic ratio y, also called Larmor constant, and an offset of the magnetic field density B 00 , can be calculated. In this application, for the sake of simplicity, the offset is sometimes set to zero in examples, so that the frequency is calculated with f(z) = m×G. z×z can be described, where m is a prefactor comprising the gyromagnetic ratio γ (Larmor constant). The individual pulses thus exhibit a shift in their frequency that is proportional to G. z is.

[0005] The actual nonlinearity in G z This leads to the situation that the layers actually recorded do not exactly correspond to the desired layers.

[0006] There are prior art approaches to mathematically correct distortion caused by the nonlinearity of the gradient profile by correcting the distortion based on the actual gradient field. For example, a method for correcting distortion is described in US 8,054,079 B2. While this can correct distortion within a slice, it cannot resolve the problem that the acquired slices were not acquired at the desired positions. This can lead to discrepancies, where the acquired slices do not show the anatomy expected during the planning phase.

[0007] It is therefore an object of the present invention to provide a means by which the aforementioned problems can be at least partially solved. In particular, it would be desirable to provide a means by which parallel layers can be detected with improved accuracy.

[0008] This problem is solved by a method according to claim 1, a method according to claim 8, a computer program product according to claim 9, and a magnetic resonance imaging system according to claim 10. Further features and advantages will become apparent from the dependent claims, the description, and the accompanying figures.

[0009] According to a first aspect of the invention, a computer-implemented method is provided for determining excitation pulses for the simultaneous acquisition of at least two parallel slices during a magnetic resonance imaging (MR) scan using a magnetic resonance imaging (MR) system. The method comprises the following steps: - Defining a target position for each of the at least two parallel layers; - Determining theoretical gradient values ​​at the target positions under the assumption of an ideal, in particular linear, local gradient profile; - Determining actual gradient values ​​at the target positions based on a predetermined location-dependent real gradient value distribution; - Calculating the gradient value differences between the theoretical gradient value and the actual gradient value of each layer; - Determining a frequency offset for each of the parallel layers based on a difference between the actual gradient value and the theoretical gradient value; - Determining an applicable pulse frequency of an excitation pulse as a superposition of individual pulses of the parallel layers, wherein the frequency of the individual pulses is based on the frequency offset and the theoretical gradient value of the respective parallel layers.

[0010] Advantageously, the method according to the invention also allows layers to be placed with improved precision in peripheral regions of the image area of ​​the MR scanner. For example, an entire anatomy, particularly regardless of its location and position, can be acquired with improved accuracy. In particular, displacements due to distortion of the gradient system can be compensated for. The method can be implemented as a computer-based procedure, and in particular can be carried out automatically by a computer unit. For example, the method can be carried out by a control unit of the MR system. For the smooth operation of the MR scanner, it is advantageous if a frequency offset is determined for each of the parallel layers, instead of recalculating the frequency entirely.This allows an existing imaging sequence to be used for the MR measurement, with the excitation pulses also being calculated normally based on the target position of the parallel slices. Before the excitation pulses are played back, they are simply corrected with the frequency offsets calculated according to the invention. The method is designed for a multi-slice (also multiband) excitation pulse with which at least two slices are excited simultaneously. Typically, 2 to 4, preferably 2 slices are excited simultaneously. However, the method can generally also be carried out for single excitation pulses.

[0011] According to the invention, a target position for the at least two parallel slices is first defined. This target position can, for example, be entered by a user. The target position can be defined according to the user input. The target position can, for example, be selected or defined based on a prescan. The target position can be defined as a position on an axis perpendicular to the respective slice and / or as a position on the z-axis of the MRI system. The z-axis of the MRI system is typically parallel to a longitudinal axis of an MRI tunnel. However, it can also have any other orientation; in particular, it can also be non-parallel to an axis of the MRI system. For example, with two slices, the two slice positions A and B can be defined.

[0012] Starting from the defined target positions, theoretical gradient values ​​are determined assuming an ideal, in particular linear, local gradient profile. For example, the theoretical gradient values ​​g can be determined for layer positions A and B. z,t (A), g z,t (B) based on a slope G z of the gradient profile according to equations of the form g z,t (A) = G z × A and g z,t(B) = Gz × B is determined. Additionally, actual gradient values ​​are determined. Actual gradient values ​​are defined, in particular, by the fact that they more closely approximate the real gradient values ​​of the MR system than the theoretical gradient values ​​and / or substantially correspond to the real gradient values ​​of the MR system. The determination of the actual gradient values ​​is based on a predetermined spatially dependent real gradient value distribution. Such a real gradient value distribution is usually already known and stored in MR systems. It is used, for example, for subsequent distortion correction during image reconstruction. Advantageously, an existing gradient value distribution can thus be used. Based on the theoretical and the actual gradient values, a frequency offset is also determined.For example, determining the frequency offset can be based on the difference between the respective theoretical gradient value and the actual gradient value. For example, the frequency offsets df(A) and df(B) can be determined for two layers at positions A and B. Using the frequency offsets, and especially starting from the theoretical gradient value, the individual pulses for the layers can be determined. For example, according to the theoretical gradient values, frequencies of f can be determined for two layers. t (A) = m × G z × A and f t (B)=m ×G z × B and result, where m is a constant prefactor comprising the Lamor constant. The individual pulses can be determined, in particular, by adding or subtracting the respective frequency offset. According to the example of two layers, the frequencies of the individual pulses can be determined, for example, with f E (A) = ft(A) + df(A) and f E (B)= f t(B) + df(B). The excitation pulse is determined as the superposition of the individual pulses of the parallel layers. For example, the excitation pulse can be determined according to f A = f E (A) + f E (B) can be determined. The excitation pulse can therefore be directly derived or composed from the individual pulses.

[0013] According to one embodiment, the frequency offsets are determined by determining a distorted layer position based on the respective actual gradient value and the respective theoretical gradient value of the respective layer, and then determining the respective frequency offset based on the distorted layer position. The distorted layer position is, in particular, the actual layer position resulting from using one of the theoretical gradient values. Due to the non-perfectly linear gradient behavior in reality, assuming theoretical linear gradient values ​​results in layer positions that deviate from the target positions. These are referred to as distorted layer positions within the scope of the invention. For example, the frequency offset can be determined by calculating the distance between the respective target position and the distorted layer position.For example, if the distance is dA for a target position A and dB for a target position B, then the frequency can be expressed as df(A) = m × G. z × dA or df(B) = m × G z The distorted layer position can be calculated in dB. For example, it can be determined by equating a formula for the theoretical gradient value (at the target position) and for the actual gradient value (at the distorted layer position) and solving for the distorted layer position.

[0014] According to one embodiment, the frequency offsets are determined by identifying a distorted layer position based on the inverse application of a distortion correction to the target position and then determining the respective frequency offset based on the distorted layer position. Methods for correcting the distortion are known in the prior art. For example, a method for correcting distortion is described in US 8,054,079 B2. Advantageously, a known method or formula can thus be used to determine the distorted layer position. For example, the target positions can be placed in a virtual image matrix, which lies orthogonally on the stack of parallel layers, and processed by the inverse function of the distortion correction.

[0015] According to one embodiment, the distorted layer positions are considered along a theoretical layer plane, defined based on a center point or center line, or based on a central position of the respective layers distorted according to the actual gradient value. The center point or center line can correspond to or define an averaged position of the respective layer. The central position is, in particular, the point of the layer that lies centrally or most centrally within the scan area. In addition to a displacement of the layers along the z-axis, distortions of the layers themselves can also occur. This embodiment offers a solution to address this. Typically, the most important data of an image are located centrally in the real image space.By concentrating on a central position, it can be ensured that the most important data are captured in the best possible positional accuracy. Determining an average position for the layers can be another way to counteract distortions within the layers themselves. In particular, this can ensure that all data for each layer are recorded with relatively accurate positional accuracy.

[0016] According to one embodiment, the frequency offsets are determined by calculating the gradient value difference between the theoretical gradient value and the actual gradient value of each layer and determining the respective frequency offset based on the respective gradient value difference of the respective layer. For example, with a difference Δg z = g z,p (A) - g z,t (A) of the respective theoretical gradient g z,t (A) and actual gradient g z,p(A) at the target position A the frequency offset df(A) = m × Δg z × A can be determined. From this, the respective individual pulse can in turn be determined, e.g. in the form f E (A) = m × (g z,t (A) × A + Δg z × A). This embodiment can represent a particularly simple way to determine the frequency offsets or the individual pulses.

[0017] According to one embodiment, the frequency of the individual pulses is determined by adding a frequency based on the respective theoretical gradient value and the frequency offset. For example, the individual pulses f E (A) and f E (B) for two target positions A and B of two layers with frequency offsets df(A) and df(B) are determined in the form f E (A) = m × g z,t (A) × A + df(A) or f E (B) = m × g z,t(B) × B + df(B). This embodiment can represent a particularly effective way to determine the frequency of the individual pulses.

[0018] According to one embodiment, the actual gradient values ​​are taken from a gradient field map or a function, in particular comprising a spherical harmonic that describes the gradient field map. The gradient field map can, in particular, be a map or list that specifies real gradient values ​​as a function of location. The function or spherical harmonic can, in particular, be a parameterized function. Since the gradient values ​​are generally not discontinuous with respect to location, the real gradient values ​​can also be well described by a function. A spherical harmonic can be a particularly efficient way to describe the gradient values ​​as a function of location. In particular, real gradient values ​​can be approximated very accurately with a spherical harmonic. Since typical gradient systems are generally rotationally symmetric, they can be represented particularly well by a spherical harmonic.For example, the gradient field at a point P(x, y, z) can be determined using a function. Alternatively, other types of functions can also be used. The gradient field map can, for instance, be determined by measuring the actual gradient field strength as a function of location. The gradient field map can be approximated by parameterizing a function, in particular a spherical harmonic. Advantageously, with a parameterized function, not all explicit values ​​of the gradient field map need to be stored. The spherical harmonic can, for example, include normalized coefficients. For simplification, small coefficients that do not make a significant contribution can be ignored. A normalization could, for example, take the form... norm(n,m)=(−1)m(2n+1)(n−m)!2(m+2) exhibit, whereby norm(n,0)=1 The field, or its z-component, can be derived from the spherical harmonic, for example, using equations of the form Bz(r,θ,ϕ)=Gxr0∑n∑mA(n,m)norm(n,m)(rr0)ncos(mϕ)Pnm(cos(θ)) for the X-gradient, Bz(r,θ,ϕ)=Gyr0∑n∑mB(n,m)norm(n,m)(rr0)nsin(mϕ)Pnm(cos(θ)) for the Y-gradient and Bz(r,θ,ϕ)=Gzr0∑nA(n,0)(rr0)nsin(mϕ)Pn0(cos(θ)) to be calculated for the Z-gradient, whereby Pnm(cos(θ)) a Legendre polynomial, which is considered Pnm(x)=(−1)m(1−x2)m / 2dm[Pn(x)] / dxm

[0019] is defined as follows: the Legendre polynomial is the differential equation (1−x2)y"−2xy'+n(n+1)y=0 The condition is met. The expressions for X and Z can be used, in particular, for Z-gradients without rotational symmetry.

[0020] Another aspect of the invention is a method for simultaneously acquiring at least two parallel slices in a magnetic resonance tomography measurement with a magnetic resonance tomography system, wherein the method comprises the following steps: - Applying a method for determining excitation pulses for the simultaneous acquisition of at least two parallel slices during a magnetic resonance imaging measurement as described herein; - Performing the magnetic resonance imaging measurement with the specified excitation pulses.

[0021] All the advantages and features of the method for determining excitation pulses for the simultaneous acquisition of at least two parallel slices during a magnetic resonance imaging (MRI) scan can be applied analogously to the method for simultaneously acquiring at least two parallel slices in a MRI scan, and vice versa. In particular, the MRI scan can be based on the acquisition of at least two parallel slices. Specifically, the MRI scan can be based on a simultaneous multi-slice (SMS) technique.

[0022] Another aspect of the invention is a computer program product or computer-readable storage medium comprising instructions which, when executed by a computer, in particular a control device of a magnetic resonance imaging system, cause the computer to perform the steps of one of the methods described herein. All advantages and features of the methods described herein can be transferred analogously to the computer program product and vice versa. The computer program product can, for example, be stored on a computer-readable storage medium, in particular a non-volatile storage medium. The storage medium can, for example, be a hard drive, an SSD, flash memory, an online server, etc.

[0023] Another aspect of the invention is a magnetic resonance imaging (MRI) system designed to simultaneously acquire at least two parallel slices during an MRI scan, comprising a control device with a computer program as described herein. The control device is specifically designed to control the measurement operation of the MRI system. All advantages and features of the methods and computer program described herein can be transferred analogously to the MRI system and vice versa.

[0024] All embodiments described herein can be combined with one another, unless explicitly stated otherwise.

[0025] The following describes embodiments with reference to the attached figures. Fig. Figure 1 shows a flowchart of a computer-implemented method for determining excitation pulses for the simultaneous acquisition of at least two parallel slices during a magnetic resonance imaging measurement with a magnetic resonance imaging system according to an embodiment of the invention; Fig. Figure 2 shows a flowchart of a method for simultaneously acquiring at least two parallel slices in a magnetic resonance tomography measurement with a magnetic resonance tomography system according to an embodiment of the invention; Fig. Figure 3 shows an ideal gradient curve that is linear, compared to an example of a real gradient curve that deviates from a linear curve; Fig. Figure 4 shows a cross-sectional image of a body part with target positions for several layers; Fig. Figure 5 shows a cross-sectional image of a body part with distorted layer positions for multiple layers; Fig. Figure 6 schematically indicates a target position and a distorted layer position for each of two layers; Fig. Figure 7 schematically indicates a target position and a distorted layer position for each of two layers, with a center line of a distorted layer position drawn and Fig. Figure 8 shows a partial view of a magnetic resonance imaging system according to an embodiment of the invention.

[0026] Fig. Figure 1 shows a flowchart of a computer-implemented method for determining excitation pulses for the simultaneous acquisition of at least two parallel slices during a magnetic resonance imaging (MRI) scan with a MRI system 4 according to an embodiment of the invention. In a first step 101, a target position 22 of the at least two parallel slices is defined. For two parallel slices, two target positions 22 can be defined accordingly. If several scans of parallel slices are to be acquired sequentially, several sets of target positions 22 can be defined, and the method can be carried out for all sets of target positions 22. In a further step 102, theoretical gradient values ​​are determined at each of the target positions 22, assuming an ideal, in particular, linear gradient profile 11.In a further step 103, actual gradient values ​​are determined at the target positions 22. Determining the actual gradient values ​​in this step 103 is based on a predetermined real gradient value distribution. For example, the actual gradient values ​​can be taken from a gradient field map or from a function that describes the gradient field map. The function can, for example, be a spherical harmonic or encompass a spherical harmonic.

[0027] Fig. Figure 3 shows an exemplary ideal gradient profile 11, which is linear, compared to an (exemplary) real gradient profile 12, which deviates from a linear profile, particularly locally, here on the right side of its path. Such a gradient profile 12 influences the actual position of the layers when applying the theoretical gradient value determined on the basis of an ideal gradient profile 11. Fig. Figure 4 shows a cross-sectional image of a body part 20, in particular the elbow, with intended target positions 22 for several layers. These are the positions that are actually to be measured. Fig. Figure 5, in contrast, also shows the image of body part 20, where the positions are set by applying a theoretical gradient value. The results are as follows: Fig. 4 distorted layer positions 21. These are shown here with exaggerated distortion for clarity. It can be seen that in this example the distorted layer positions 21 are shifted to the right compared to the target positions 22, especially in the right area of ​​the image.

[0028] Such a distortion is to be prevented by the method according to the invention. Referring again to the in Fig. In the method shown, a frequency offset is determined for each of the parallel layers in a further step 104. The frequency offsets are each based on the difference between the actual gradient value 12 and the theoretical gradient value 11. For example, the frequency offsets can be determined based on determined distorted layer positions 21. In particular, a difference between the respective target position 22 and the respective distorted layer position 21 can be calculated. According to a formula of the form f(z) = m × G z × z can thus be calculated based on the slope G z The ideal gradient profile 11 at a position z of the frequency offset df(z) can be determined. The respective distorted layer position 21 can be determined, for example, based on the respective actual gradient value 12 and the respective theoretical gradient value 11 of the respective layer. In particular, the difference Δg can be used for this purpose.z These two gradient values ​​11, 12 are formed so that the frequency offset df(z) is determined by the formula f(z) = m × G z × z the variable G z by Δg z is replaced so that df(z) = m × Δg z × z. Alternatively, a distorted layer position can be determined by applying the inverse of a (known) distortion correction to the target position 22. From the difference dz between the distorted layer position 21 and the target position 22, the frequency offset df(z) = m × G can then be calculated. z × dz can be determined.

[0029] In addition to a distortion of the layer positions in the z-direction, the layers themselves can also become distorted. This is exemplified in Fig. Figure 6 indicates the target position 22 and the distorted layer position 21 for each of two layers. The path of the layer itself in a direction perpendicular to the z-direction is also shown. This can lead to curved layers, particularly in the edge regions of the scan area, as indicated here by the distorted layer on the right. To define the distorted layer position 21 in such a case, it may be possible to define a center point or center line of the respective layer, which serves as the basis for defining the distorted layer position 21. Alternatively, a central position of the layers may be used to define the distorted layer position 21.

[0030] Referring again to the in Fig. In the method shown in Figure 1, a suitable pulse frequency for an excitation pulse is determined in a further step (105) as a superposition of individual pulses from the parallel layers. The frequency of each individual pulse is based on the frequency offset and the theoretical gradient value of the respective parallel layers. In particular, the frequency offset can be added to the frequency calculated using the theoretical gradient values ​​to determine the individual pulses. This addition can also be a subtraction, i.e., an addition with a negative sign.

[0031] Fig. Figure 2 shows a flowchart of a method for simultaneously acquiring at least two parallel slices in a magnetic resonance imaging (MRI) scan using a MRI system 4 according to an embodiment of the invention. In a first step 201, a target position 22 of each of the at least two parallel slices is defined. In a further step 202, theoretical gradient values ​​at the target positions 22 are determined, assuming an ideal, in particular linear, local gradient profile 11. In a further step 203, actual gradient values ​​at the target positions 22 are determined based on a predetermined, spatially dependent real gradient value distribution. In a further step 204, a frequency offset for each of the parallel slices is determined based on the difference between the actual gradient value and the theoretical gradient value.In a further step 205, a pulse frequency of an excitation pulse is determined as a superposition of individual pulses from the parallel layers. The frequency of the individual pulses is based on the frequency offset and the theoretical gradient value of the respective parallel layers. These steps 201-205 can be compared in particular to steps 101-105 of the procedure with reference to... Fig. The procedure described in section 1 is followed. In a further step 206, a magnetic resonance imaging (MRI) measurement is performed using the specified excitation pulses. This can, in particular, be a MRI measurement based on a simultaneous multi-slice (SMS) technique.

[0032] Fig. Figure 8 shows a section of a magnetic resonance imaging system 4 according to an embodiment of the invention. The magnetic resonance imaging system 4 is configured to simultaneously acquire at least two parallel slices during a magnetic resonance imaging measurement. The magnetic resonance imaging system 4 comprises a control device 5, which is configured to perform a method as described in relation to Fig. 1 or Fig. 2 described, to execute or to cause the magnetic resonance imaging system 4 to execute such a procedure. For this purpose, a corresponding computer program product can be installed on the control device 5. QUOTES INCLUDED IN THE DESCRIPTION

[0000] This list of documents cited by the applicant was automatically generated and is included solely for the reader's convenience. The list is not part of the German patent or utility model application. The DPMA accepts no liability for any errors or omissions. Cited patent literature

[0000] US 8 054 079 B2 [0006, 0014]

Claims

[1] Computer-implemented method for determining excitation pulses for the simultaneous acquisition of at least two parallel slices during a magnetic resonance imaging measurement with a magnetic resonance imaging system (4), wherein the method comprises the following steps: - Defining a target position (22) for each of the at least two parallel layers; - Determining theoretical gradient values ​​at the target positions (22) under the assumption of an ideal, in particular linear, local gradient profile (11); - Determining actual gradient values ​​at the target positions (22) based on a predetermined location-dependent real gradient value distribution; - Determining a frequency offset for each of the parallel layers based on a difference between the actual gradient value and the theoretical gradient value; - Determining an applicable pulse frequency of an excitation pulse as a superposition of individual pulses of the parallel layers, wherein the frequency of the individual pulses is based on the frequency offset and the theoretical gradient value of the respective parallel layers. [2] Method according to claim 1, where the frequency offsets are determined by Determine each distorted layer position based on the respective actual gradient value and the respective theoretical gradient value of the respective layer and Determining the respective frequency offset based on the distorted layer position. [3] Method according to any one of the preceding claims, where the frequency offsets are determined by Determining each distorted layer position (21) based on the inverse application of a distortion correction to the target position (22) and Determining the respective frequency offset based on the distorted layer position (21). [4] Method according to claim 2 or 3, wherein the distorted layer positions (21) are considered along a theoretical layer plane, based on a center point or center line (25) or based on a central position of the respective layers distorted according to the actual gradient value. [5] Method according to claim 1, where the frequency offsets are determined by Calculating the gradient value difference between the theoretical gradient value and the actual gradient value of each layer and Determining the respective frequency offset based on the respective gradient value difference of the respective layer. [6] Method according to one of the preceding claims, wherein the frequency of the individual pulses is determined by adding a frequency based on the respective theoretical gradient value and the frequency offset. [7] Method according to one of the preceding claims, wherein the actual gradient values ​​are taken from a gradient field map or a function, in particular a spherical harmonic, that describes the gradient field map. [8] Method for simultaneously acquiring at least two parallel slices in a magnetic resonance tomography measurement using a magnetic resonance tomography system (4), wherein the method comprises the following steps: - Applying a method according to one of the preceding claims; - Performing the magnetic resonance imaging measurement with the specified excitation pulses. [9] Computer program product or computer-readable storage medium comprising instructions which, when executed by a computer, in particular a control device (5) of a magnetic resonance imaging system (4), cause it to perform the steps of the method according to one of the preceding claims. [10] Magnetic resonance imaging system (4) configured to simultaneously acquire at least two parallel slices during a magnetic resonance imaging measurement, comprising a control device (5) with a computer program product according to claim 9.

Images