Method for diffusion imaging by means of a magnetic resonance device, magnetic resonance device, computer program product and electronically readable data carrier

Through the plane diffusion encoding method, the trajectory weighted image data set is calculated using geometric mean values, which solves the problems of frequent measurement times and low efficiency in the prior art, and realizes efficient calculation and efficiency improvement of the trajectory weighted data set.

CN113625211BActive Publication Date: 2024-11-15SIEMENS HEALTHINEERS AG
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202110483125.4
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Priority Date
2020-05-06
Filing Date
2021-04-30
Publication Date
2024-11-15
Estimated Expiration
2041-04-30

AI Technical Summary

Technical Problem

When obtaining trajectory-weighted image data sets, the prior art measures a lot and is greatly affected by microscopic anisotropy and mesoscopic orientation dispersion, and the efficiency is low.

Method used

The plane diffusion encoding method is used to specify the diffusion gradient pulse sequence to make its b matrix have two non-zero eigenvalues, and the trajectory-weighted image data set is calculated using the geometric mean value to reduce the number of measurements and reduce the influence of mesoscopic orientation diffusion.

Benefits of technology

Efficient calculation of trajectory weighted image dataset is realized, reducing the number of measurements, improving efficiency, and reducing the impact on microscopic anisotropy and mesoscopic orientation dispersion.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN113625211B_ABST
    Figure CN113625211B_ABST
Patent Text Reader

Abstract

The invention relates to a method for diffusion imaging of an examination region of a patient by means of a magnetic resonance device (8), the method comprising the following steps: - specifying a number of diffusion gradient pulse sequences for recording diffusion data sets by means of the magnetic resonance device (8) such that a b-matrix describing a planar diffusion coding is associated with the diffusion gradient pulse sequence, the number being at least two, the b-matrix having exactly two non-zero eigenvalues; - recording the diffusion data sets by means of the specified diffusion gradient sequence; determining a trajectory-weighted image data set by geometrically averaging at least two diffusion data sets; - wherein the diffusion gradient pulse sequence is determined such that the sum of all b matrices results in a unit matrix multiplied by a factor characterizing the diffusion weighting, in particular a predetermined b value and the number divided by three.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The invention relates to a method for diffusion imaging of an examination region of a patient by means of a magnetic resonance apparatus, wherein trajectory-weighted image data sets are ascertained. The invention also relates to a magnetic resonance apparatus, a computer program and an electronically readable data carrier. Background Art

[0002] In addition to other imaging techniques in magnetic resonance imaging, diffusion imaging has become significantly more relevant in recent years. The development of diffusion imaging (diffusion-weighted magnetic resonance imaging, "Diffusion-Weighted Imaging" - DWI) is based on the pulse gradient spin echo technique (PGSE) for measuring diffusion-weighted echo signals, which was first proposed by Edward Stejskal and John Tanner in the mid-1960s. The pulse gradient spin echo technique has been further developed in various ways and methods during this period. Diffusion imaging is characterized in that the diffusion of water molecules is used to generate contrast in the magnetic resonance image data set. In this way, the diffusion process of molecules in biological tissue can be mapped. In order to be able to measure the diffusion process, in addition to the imaging gradient pulses, additional diffusion gradient pulses are used in the magnetic resonance sequence, which shape the magnetic resonance sequence sensitively to motion. Here, the diffusion in biological tissue should generally be referred to as anisotropy. Applications in which the anisotropy is the object of examination and the individual components of the diffusion tensor are measured are known, whereas in other applications it may be that the anisotropy is strong and masks the underlying changes in the local surface diffusion coefficient (“Apparent Diffusion Coefficient” - ADC). In this and other cases it is preferred to determine the trajectory of the diffusion tensor after it has been made orientation-independent. The corresponding imaging approach is also referred to as trajectory-weighted diffusion imaging or tracer-weighted diffusion imaging.

[0003] In clinical magnetic resonance diffusion imaging, trajectory-weighted image data sets and parameter maps derived therefrom, such as parameter maps of the surface diffusion coefficient ADC, are in particular the basis for detecting microstructural changes in tissue, making them useful, for example, in the diagnosis of stroke or dynamic tumor processes in humans.

[0004] In conventional diffusion coding via corresponding diffusion gradient pulse sequences, the microscopic movement of the signal-emitting nuclear spins along a certain spatial direction is examined with each acquisition of a diffusion data set, whereby this type of diffusion coding is also referred to as linear encoding. From at least three acquisitions, thus three diffusion data sets with different encoding directions, a trajectory-weighted image data set can then be calculated.

[0005] As a measure of how sensitive a cell is to diffusional motion, the so-called b-value is often used in the literature. In the spatial description, the b-matrix indicates how the sensitivity is distributed in space. In this case, in clinical applications, for example, one can aim for a value between 500 s / mm 2 Up to 2000s / mm 2 b-values ​​in the range of , wherein approaches are also known in which a trajectory-weighted image data set is to be determined for a plurality of b-values.

[0006] Diffusion gradient pulse sequences are used for the already mentioned linear diffusion coding, for which the b matrix for each single measurement, ie each acquisition of the diffusion data set, has exactly one non-zero eigenvalue. In order to be able to determine the image data set with trajectory weightings, the b matrices b for at least N=3 encoding directions are n The following conditions must be met:

[0007] N / 3 b 1 = Σ n=1…N b n , (1)

[0008] where 1 is the identity matrix and b is the b value. The diffusion dataset can then be derived from each diffusion weight

[0009] S n (r) = S0(r) exp( -Tr(D(r) b n )) (2)

[0010] Based on

[0011] S Trace (r) = Π n=1…N S n (r) 1 / N

[0012] =S0(r)exp(-1 / N Tr(D(r)Σ n=1…N b n ))

[0013] = S0(r) exp( -b / 3 Tr(D(r)) ) (3) Compute trajectory weights on image datasets.

[0014] Here, D(r) represents the symmetric 3×3 diffusion tensor at position r, Tr represents the trajectory forming operation (sum of diagonal elements), and S0(r) represents the magnetic resonance signal received without diffusion weighting. In this context, the so-called surface diffusion coefficient ADC is defined as:

[0015] ADC = 1 / 3 Tr(D(r)). (4)

[0016] The advantage of linear diffusion coding is that it is extremely efficient, so high b-values ​​can be achieved with the help of the short duration of the corresponding diffusion gradient pulse sequence. In addition, for each single b-value, an independent calculation of the trajectory-weighted image data set can be achieved. However, it is disadvantageous that three measurements are required in order to obtain the trajectory-weighted image data set. In addition, trajectory weighting is affected by microscopic anisotropy and mesoscopic orientation diffusion, as shown, for example, in the article "Linear, planar and spherical tensor-valued diffusionMRI data by free waveform encoding in healthy brain, water, oil and liquid crystals" by Filip Szczepankiewicz et al., Data in Brief 25 (2019) 104028.

[0017] So-called spherical diffusion coding has been proposed as another way to determine trajectory-weighted image data sets in diffusion imaging, for this see, for example, the article by Eric C. Wong et al., “Optimized isotropic diffusion weighting”, Magnetic Resonance in Medicine 34 (1995), pages 139-143. In this case, only a single diffusion data set must be recorded, which directly corresponds to the trajectory-weighted image data set. In this case, the following diffusion gradient pulse sequence is used, the b matrix of which has three almost identical, non-zero eigenvalues ​​b1=b2=b3=b / 3. Therefore, the b matrix can be expressed as b=b / 3 1. Therefore, the diffusion data set of a single spherical diffusion code already repeats the trajectory weighting,

[0018] S(r) = S0(r) exp( -Tr(D(r) b)) = S0(r) exp( -b / 3 Tr(D(r)) ) (5)

[0019] Spherical diffusion coding advantageously requires only one measurement, allowing for the recording of trajectory-weighted image data sets independently for each single b-value, and whose trajectory weightings are only affected by microscopic anisotropy. However, spherical diffusion coding is disadvantageously very inefficient, meaning that a long period of time is required to reach the desired b-value.

[0020] As a further alternative approach, it has been proposed to calculate the trajectory weighting or specifically the trajectory-weighted image data set based on the tensor estimation, for which see the article by Peter J. Basser et al. "Estimation of the effective self-diffusion tensor from the NMR spin echo", Journal of Magnetic Resonance, Series B 103 (1994), pages 247-254. Here, the complete diffusion tensor D(r) and the magnetic resonance signal S0(r) without diffusion weighting are estimated from all recorded diffusion data sets by means of known methods. From this, the trajectory-weighted image data set can be calculated for a given b value as:

[0021] S0(r)exp(-b / 3Tr(D(r))). (6)

[0022] The described approach has the advantage that a complete diffusion tensor is obtained from which other parameter maps can be determined, and that an extremely efficient coding is obtained, in particular a linear diffusion coding can be used. Disadvantageously, however, the described approach requires at least seven measurements with at least six different coding directions and at least two different diffusion weightings, and thus the b-values. It is not feasible to determine the trajectory weighted image data set independently for a single b-value, since at least two diffusion weightings are taken into account. Finally, the tensor estimate and thus the trajectory weighting is affected by microscopic anisotropy and mesoscopic orientation dispersion. Summary of the invention

[0023] The object underlying the present invention is therefore to specify a method for ascertaining trajectory-weighted image data sets which is efficient, requires few measurements and is less susceptible to microanisotropy and mesoscopic orientation dispersion.

[0024] In order to achieve the object, a method for diffusion imaging of an examination region of a patient by means of a magnetic resonance device comprises the following steps:

[0025] - a number of diffusion gradient pulse sequences for acquiring a diffusion data set by means of a magnetic resonance apparatus is provided, such that the diffusion gradient pulse sequence has a b-matrix describing a planar diffusion encoding, the number of which is at least two and has exactly two non-zero eigenvalues,

[0026] - recording of diffusion data sets with the aid of a defined diffusion gradient sequence,

[0027] - obtaining a trajectory-weighted image dataset by taking the geometric mean of at least two diffusion datasets,

[0028] The diffusion gradient pulse sequence is determined in such a way that the multiplication of the identity matrix with factors characterizing the diffusion weighting, in particular a predefined b value and the number divided by three, results in the sum of all b matrices.

[0029] Here, the b value is usually used as a factor to characterize the diffusion weighting. Therefore, the present invention provides the following feasibility: using planar diffusion coding, from at least two diffusion-weighted images in a diffusion data set with the same b value, a trajectory-weighted image data set is directly calculated. This means that the b matrix can be expressed in particular as the multiplication of the same b value with a matrix filled with, in particular, rational numbers and zeros. Correspondingly, the method according to the present invention proposes to record at least two diffusion data sets with the help of planar diffusion coding and the same diffusion weighting (b value) in order to calculate the trajectory-weighted image data set therefrom. In this way, the advantages of linear diffusion coding and spherical diffusion coding are combined, so that a good compromise between efficiency and contrast performance is achieved. Correspondingly, the trajectory-weighted image data set can be advantageously determined independently for each individual b value of interest. Compared with linear diffusion coding, the influence of mesoscopic orientation dispersion on trajectory weighting is reduced, and compared with spherical diffusion coding, the higher efficiency of diffusion coding is immediately obtained. A small number of measurement processes, i.e., at least two measurement processes are sufficient.

[0030] Planar diffusion coding is characterized in that the b matrix has exactly two non-zero eigenvalues. The trajectory-weighted image can be generated from the number N of diffusion data sets recorded with the aid of planar coding as the geometric mean:

[0031] S Trace (r)= Π n=1…N S n (r) 1 / N , (7)

[0032] If b matrix b n The conditions of equation (1) are satisfied again, so we have:

[0033] S Trace (r) = Π n=1…N S n (r) 1 / N

[0034] =S0(r)exp(-1 / N Tr(D(r)Σ n=1…N b n ))

[0035] = S0(r) exp( -b / 3 Tr(D(r)) ). (8) All b matrices are usually symmetric and real-valued, so they can be represented in diagonal form.

[0036]

[0037] In contrast, for spherical diffusion coding we get:

[0038]

[0039] And for planar diffusion coding:

[0040]

[0041] For planar diffusion coding, x>0 and y>0 apply. The special case x=y is called circular diffusion coding.

[0042] Since it is ultimately desirable to keep the number of measurements, ie the number of recording processes for the diffusion data set, small, the number may suitably be two or three, preferably two.

[0043] Diffusion data sets can be recorded using spin echo sequences and / or EPI readout techniques. EPI (Echo Planar Imaging) provides very efficient spatial encoding, which is beneficial for measuring large or well-resolved examination areas. Nowadays, spin echo techniques are frequently used in diffusion imaging.

[0044] Nevertheless, it should be noted that the method according to the invention can be carried out not only with the aid of a spin echo sequence, but also, for example, with the aid of a gradient echo sequence, which means that no refocusing pulse is required, or also with the aid of a doubly focused spin echo sequence (which means, for example, two refocusing pulses output after the first diffusion gradient pulse and before the last diffusion gradient pulse). In addition, magnetic resonance sequences with stimulated echoes are also possible. In addition to the EPI technique, in particular the single-shot EPI technique, segmented EPI, turbo spin echo (TSE), multi-gradient echo, etc. can also be used as readout techniques.

[0045] Two examples are shown below for a number N=2 and a number N=3. If a trajectory-weighted image data set is to be ascertained from N=2 measurements by means of planar diffusion coding, the following b matrix can be used for the first measurement:

[0046]

[0047] And for the second measurement the following b matrix can be used:

[0048]

[0049] Then condition (1) is satisfied:

[0050]

[0051] The diffusion dataset can then be encoded from two planes

[0052] S 1,2 (r) = S0(r) exp( -Tr(D(r) b 1,2 )) (15) According to:

[0053] S Trace (r) = ( S1(r) S2(r) ) 1 / 2 (16) Calculate the trajectory-weighted image data set. In the case of N=3 measurements with planar diffusion coding, the following b matrix can be used, for example:

[0054]

[0055] As is evident, the conditions according to equation (1) are again satisfied,

[0056]

[0057] Trajectory weighted image dataset from the current three planes encoded diffusion dataset

[0058] S 1,2,3 (r) = S0(r) exp( -Tr(D(r) b 1,2,3 )) (21) as

[0059] S Trace (r) = ( S1(r) S2(r) S3(r) ) 1 / 3 (22) concluded.

[0060] In order to realize such a b-matrix in a concrete technical manner, different approaches are conceivable within the scope of the present invention. It is particularly preferred that the diffusion gradient pulse sequence is determined according to a double diffusion encoding scheme (DDE scheme), while other exemplary conceivable variants provide that the diffusion gradient pulse sequence is determined in two spatial directions by a temporally optimized gradient variation curve or as a sequence of gradient pulses of identical shape and amplitude. It is also conceivable in particular that in at least one of the diffusion gradient pulse sequences, gradient pulses are output simultaneously in multiple spatial directions, in particular on multiple physical gradient axes, preferably on all physical gradient axes. This can achieve a significant improvement in the efficiency of the diffusion encoding.

[0061] The preferred "double diffusion encoding" (DDE) according to the present invention is described, for example, in the article "Conventions and nomenclature for double diffusion encoding NMR and MRI" by Noam Shemesh et al., Magnetic Resonance in Medicine 75 (2016), pages 82 to 87. In this case, for example, two pairs of diffusion gradient pulses can be applied along different encoding directions, in particular two different physical gradient axes (X, Y and Z). It is then conceivable, for example, that in the case of a single-pulse spin echo magnetic resonance sequence embedded in a planar echo readout column with one of the diffusion gradient pulse pairs before the refocusing pulse, another pair of diffusion gradient pulses is output thereafter. The time interval of a pair of diffusion gradient pulses and their width (in the case of at least substantially rectangular diffusion gradient pulses) together with the gradient amplitude determine the corresponding b-value.

[0062] In addition to the use of DDE sequences, it also lies within the scope of the invention to use, for example, optimized time gradient profiles in both encoding directions, as is described, for example, in the article by Filip Szczepankiewicz et al. mentioned at the outset.

[0063] In addition, planar diffusion encoding can also be achieved by a sequence of diffusion gradient pulses with the same shape and amplitude along two spatial directions. For example, the following sequence of idealized rectangular gradient pulses and pauses of the same duration on the x-axis and y-axis generates an almost circular encoding: G x ={+1,+1,+1,-1,-1,-1,-1,-1,-1,+1,+1,+1}

[0064] G y ={-1,0,-1,0,0,-1,+1,0,0,+1,0,+1}.

[0065] As already mentioned, in a technical implementation, the efficiency of planar diffusion encoding can be improved by using diffusion gradient pulses simultaneously on multiple physical gradient axes. If, in the example of a DDE sequence, a first gradient pair with an amplitude G is applied temporally spaced apart on the x-axis, and a second gradient pair with the same amplitude is applied temporally spaced apart on the y-axis, the following b-matrix is ​​obtained:

[0066]

[0067] The b value is calculated using the gyromagnetic ratio γ, the gradient pulse width δ, the gradient pulse interval Δ of a pair of diffusion gradient pulses, the gradient amplitude G, and while neglecting the gradient slope:

[0068] b=2 / 3γ 2 δ 2 G 2 (Δ-δ / 3)+2 / 3γ 2 δ 2 G 2 (Δ-δ / 3)=4 / 3γ 2 δ 2 G 2 (Δ-δ / 3). (24)

[0069] In this context, a more efficient encoding is obtained if, for example, a first pair of diffusion gradient pulses is applied simultaneously on the x-axis (amplitude +G), the y-axis (amplitude +G) and the z-axis (amplitude +G / 2), and a second pair of diffusion gradient pulses is applied on the x-axis (amplitude -G), the y-axis (amplitude +G / 2) and the z-axis (amplitude +G). The two encoding directions are orthogonal to each other and a planar diffusion encoding, here even a circular diffusion encoding, of the form:

[0070]

[0071] After diagonalization

[0072]

[0073] in

[0074] b' = 4 / 3 γ 2 δ 2 G eff 2 (Δ - δ / 3). (27)

[0075] This applies:

[0076] G eff 2 =G 2 (1+1+1 / 4)=9 / 4G 2 , (28)

[0077] This improves the efficiency of diffusion coding by a factor of 2.25.

[0078] If combined with a second and a third measurement, which applies gradient pairs as (+G, -G, +G / 2) and (-G / 2, +G, -G) or as (-G, -G / 2, +G) and (+G / 2, +G, +G), this means that it uses the following b-matrix:

[0079]

[0080] as well as

[0081]

[0082] Then the corresponding trajectory-weighted image dataset can be generated by taking the geometric mean.

[0083] Here, in the context described, the DDE encoding directions do not necessarily have to be perpendicular to each other. If this premise is abandoned, new degrees of freedom are created, through which other efficiency increases are possible. For example, a first pair of diffusion gradient pulses can be applied at (+G, +G, +G) and a second pair of diffusion gradient pulses can be applied at (-G, -G, +G) at the same time. Now, the encoding directions are no longer orthogonal, and a planar diffusion encoding of the following form is obtained:

[0084]

[0085] After diagonalization

[0086]

[0087] Here b′ is derived again from equation (27). However, here we use

[0088] G eff 2 =G 2 (1+1+1)=3G 2 (33)

[0089] The efficiency of diffusion coding is even improved by a factor of 3.

[0090] In combination with a second measurement with applied gradient pairs (+G, -G, -G) and (-G, +G, -G), we can thus obtain the value of

[0091]

[0092] The trajectory-weighted image dataset can be generated again by taking the geometric mean.

[0093] In a particularly advantageous development of the invention, it is proposed that, in the case of imaging gradient pulses which are output primarily on a first physical gradient axis for recording a diffusion data set, the distribution of the diffusion gradient pulses is selected such that, for each diffusion gradient pulse sequence, the diffusion gradient pulses are output primarily on a physical gradient axis different from the first physical gradient axis. In other words, for measurements in which the imaging gradient is first applied on a specific physical gradient axis, namely the first physical gradient axis, the load of the planar diffusion encoding gradient pulses can be advantageously distributed to other physical gradient axes, in particular the second physical gradient axis and the third physical gradient axis. In this way, depending on the gradient device or gradient hardware used, more power can be used for the imaging portion.

[0094] For example, in an EPI readout train, the main load is on the readout gradient axis, such as the x-axis, which can be kept as free of load as possible from planar diffusion encoding gradient pulses.

[0095] Here, in a specific design, it can be provided that, in a diffusion gradient pulse sequence of two, in the first gradient pulse sequence, a diffusion gradient pulse pair with a predetermined amplitude is output on the second physical gradient axis, and a diffusion gradient pulse pair with a predetermined amplitude divided by the square root of two is output on the first gradient axis, and in the second gradient pulse sequence, a diffusion gradient pulse pair with a predetermined amplitude is output on the third physical gradient axis, and a diffusion gradient pulse pair with a predetermined amplitude divided by the square root of two is output on the first gradient axis. If the first physical gradient axis is, for example, the x-axis, two DDE measurements can be implemented as follows, namely, in the first measurement, a first diffusion gradient pulse pair with an amplitude of G / √2 is output on the x-axis, and a second diffusion gradient pulse pair with an amplitude of G is output on the y-axis. In the second measurement, a first diffusion gradient pulse pair with an amplitude of G / √2 is output on the x-axis, and a second diffusion gradient pulse pair with an amplitude of G is output on the z-axis. Then, although the x-axis is loaded in both measurements, it is only with a reduced amplitude.

[0096] Here, it should also be noted that it is also conceivable in principle to determine the trajectory-weighted image data set from a diffusion data set with a combination of linear diffusion coding and planar diffusion coding. For example, a first measurement with the following planar diffusion coding can be used:

[0097]

[0098] And a second measurement with linear diffusion coding as follows is used:

[0099]

[0100] Here b'=b / 2. To realize the second measurement, G'=G / √2 can be selected in the DDE diffusion gradient pulse sequence, but it is also possible to select SDE (Single Diffusion Encoding) encoding for the second measurement. Then the condition of equation (1) is satisfied in any case.

[0101] Similarly, in principle, a combination of planar diffusion coding and elliptical diffusion coding of the following form can also be considered:

[0102]

[0103] where x, y and z are different.

[0104] In addition to the method, the present invention also relates to a magnetic resonance device having a control device configured to carry out the method according to the invention. All embodiments of the method according to the invention can be transferred analogously to the magnetic resonance device according to the invention, so that the advantages mentioned above can also be achieved with the magnetic resonance device.

[0105] Here, the control device may have at least one processor and / or storage device. Specifically, for example, it can be proposed that the control device includes a regulation unit for regulating a diffusion gradient pulse sequence, a sequence unit for recording a diffusion data set, and an acquisition unit for calculating a trajectory-weighted image data set from the diffusion data set.

[0106] For example, the computer program according to the invention can be loaded into a storage device of a control device of a magnetic resonance device and has program means which, when the computer program is executed in the control device of the magnetic resonance device, execute the steps of the method according to the invention. The computer program can be stored on an electronically readable data carrier according to the invention, which thus comprises electronically readable control information, which comprises at least one computer program according to the invention and is designed such that, when the electronically readable data carrier is used in the control device of the magnetic resonance device, the electronically readable control information allows the steps of the method according to the invention to be executed. The electronically readable data carrier is in particular a non-transitory data carrier, for example a CD-ROM. BRIEF DESCRIPTION OF THE DRAWINGS

[0107] Further advantages and details of the invention are apparent from the exemplary embodiments described below and from the accompanying drawings. Here, the following is shown:

[0108] Figure 1 A flow chart showing an embodiment of the method according to the present invention,

[0109] Figure 2 shows a sequence chart for recording a diffusion data set,

[0110] Figure 3 A magnetic resonance apparatus according to the present invention is shown, and

[0111] Figure 4 The functional structure of a control device of a magnetic resonance apparatus is shown. DETAILED DESCRIPTION

[0112] Figure 1 A flow chart of an exemplary embodiment of a method according to the invention is shown in general form. The method is used for diffusion imaging, in particular for determining a trajectory-weighted image data set of an examination region of a patient.

[0113] In this case, in step S1, the diffusion gradient pulse sequences to be used subsequently are specified such that a planar diffusion encoding is provided in all diffusion gradient pulse sequences to be used, preferably in two or three diffusion gradient pulse sequences, which means that the diffusion gradient pulse sequences are specified such that their associated b-matrices have exactly two non-zero eigenvalues. The b-matrix must satisfy the condition described in equation (1), according to which the sum of all b-matrices yields an identity matrix multiplied by the b value and the number divided by three which are predefined here.

[0114] In order to technically implement the diffusion gradient sequence specifically, there are different possibilities, such as determining a temporally optimized gradient variation curve along two spatial directions, determining a sequence of gradient pulses of the same shape and amplitude in two spatial directions, etc., wherein in a preferred embodiment, the diffusion gradient pulse sequence is determined as a DDE sequence (Double Diffusion Encoding). In this case, it can also be considered which magnetic resonance sequence should be used for the imaging component, wherein a spin echo sequence with single-shot EPI readout is preferred in the present case.

[0115] Figure 2 An exemplary sequence diagram is shown, in which a DDE diffusion gradient pulse sequence is integrated into a single-shot spin echo sequence with a planar echo readout train. Here, RF / ADC stands for high frequency activity, G x The gradient pulse G is shown on the first physical gradient axis, here on the x-axis. y and G z Correspondingly for the y-axis as the second physical gradient axis and the z-axis as the third physical gradient axis. Obviously, the single-shot spin echo sequence comprises a high-frequency excitation pulse 1 and a high-frequency refocusing pulse 2. As is known in principle, the high-frequency excitation pulse and the high-frequency refocusing pulse are correspondingly associated with an imaging gradient pulse 3, which is here a slice selection gradient along the z-axis. In the readout module, the main gradient load is in the imaging gradient pulse 4 of the x-axis as the readout direction, while, as is known, in particular for the transformation of the k-space line, further imaging gradient pulses 5 on the y-axis are required.

[0116] In the present case, the planar diffusion encoding uses two pairs of diffusion gradient pulses 6, 7, wherein the first pair of diffusion gradient pulses 6 are output along the x-axis and the second pair of diffusion gradient pulses 7 are output along the y-axis. Figure 2 The diffusion gradient pulse sequence shown in implements a b-matrix of the form described by equation (23), wherein, as already explained above, the b-values ​​are derived according to equation (24). Figure 2 4 and 5. The diffusion gradient pulse width δ and the interval Δ between a pair of diffusion gradient pulses 6 and 7 are also shown in FIG.

[0117] pass Figure 2 The measurement shown is suitable, for example, as a first measurement in the acquisition of three diffusion data sets, wherein the same amplitude G of the diffusion gradient pulses 6 , 7 can be used there.

[0118] In another example, if the number N should be only two, such as when implementing a b matrix of the form:

[0119]

[0120] For example, for the diffusion gradient pulse 6, the amplitude G can be used, while for the other pair of diffusion gradient pulses 7, the amplitude G / √2 can be used. In this case, the value b for the entire diffusion gradient pulse sequence is obtained as:

[0121] b=2 / 3γ 2 δ 2 G 2 (Δ-δ / 3)+1 / 3γ 2 δ 2 G 2 (Δ-δ / 3) = γ 2 δ 2 G 2 (Δ-δ / 3). (39)

[0122] To show it more simply, Figure 2 The figure shows the gradients along different physical gradient axes. Figure 2 In the example of , the diffusion gradient pulse pairs along the x-axis and the y-axis are provided. However, it is preferred to use the measures described with respect to equations (26) to (34) to improve the diffusion efficiency, and / or in Figure 2 After the main load due to the EPI readout is on the x-axis in the sequence diagram of , the gradient load on the x-axis is reduced in a targeted manner, as described.

[0123] As soon as the diffusion gradient pulse sequences are correspondingly defined, the diffusion data sets can be recorded in step S2 using a respective one of the diffusion gradient pulse sequences, see again Figure 1 Then, in step S3, the trajectory weighted image dataset S can be converted into Trace Calculated as the geometric mean of the diffusion data set.

[0124] Figure 3The schematic diagram of the principle of a magnetic resonance device 8 according to the invention is shown. As is known in principle, the magnetic resonance device has a main magnet unit 9, in which a patient accommodating part 10 is formed, into which a patient can be moved for examination by means of an examination couch (not shown in detail here). A gradient coil arrangement with gradient coils for the x-axis, y-axis and z-axis (not shown in detail here for the sake of overview) is usually provided around the examination couch. Both imaging gradient pulses 3, 4, 5 and diffusion gradient pulses 6, 7 can be output via the gradient coil arrangement.

[0125] The operation of the magnetic resonance device 8 is controlled by a control device 11 , which is only indicated and which is also designed to carry out the method according to the invention.

[0126] Figure 4 The functional structure of the control device 11 is shown in more detail. As is known in principle, the control device 11 first has a sequence unit 12, via which the remaining components of the magnetic resonance device 8 for recording magnetic resonance data are controlled according to a magnetic resonance sequence. Therefore, in step S2, the diffusion data set can also be recorded with the aid of the sequence unit 12. For this purpose, the b matrix to be used and, in particular, the diffusion gradient pulse sequence which technically implements the b matrix are previously specified with the aid of a specification unit 13. It should also be noted here that it is possible to specify once before a plurality of measurements, i.e., in particular to carry out once according to a magnetic resonance sequence. Figure 1 This means that a diffusion gradient pulse sequence that is specified once can be used multiple times.

[0127] The control device 11 also has a determination unit 14 for determining the trajectory-weighted image data set according to step S3 .

[0128] The functional units, namely the sequence unit 12, the definition unit 13 and the determination unit 14, can be implemented by at least one processor of the control device 11. In addition, the control device can also have a storage device 15, in which, for example, a defined diffusion gradient pulse sequence and a diffusion data set can be stored at least temporarily, but of course also the final result can be stored. The final result, and thus the trajectory-weighted image data set, can also be output via an interface 16 of the control device 11, wherein the control device 11 can of course also have a display unit, in order to output the trajectory-weighted image data set, for example on a display device of the magnetic resonance device 8.

[0129] Finally, it should be pointed out that within the scope of the invention it is also possible to determine a plurality of trajectory-weighted image data sets for different b values, in particular in order to derive further information and parameters therefrom, for example regarding the diffusion tensor and / or the surface diffusion coefficient (ADC).

[0130] Although the details of the present invention have been explained and described in detail through the preferred embodiments, the present invention is not limited by the disclosed examples and other variations can be derived therefrom by those skilled in the art without departing from the scope of protection of the present invention.

Claims

1. A method for diffusion imaging of an examination region of a patient by means of a magnetic resonance device (8), the method comprising the following steps: - a number of diffusion gradient pulse sequences for recording a diffusion data set by means of the magnetic resonance device (8) is specified, so that the diffusion gradient pulse sequence comprises b matrices describing a planar diffusion encoding, the number being at least two, the b matrices having exactly two non-zero eigenvalues, - recording the diffusion data set with the aid of a determined diffusion gradient sequence, - obtaining a trajectory-weighted image dataset by taking the geometric mean of at least two diffusion datasets, wherein the diffusion gradient pulse sequence is determined in such a way that the sum of all b matrices results in an identity matrix which is multiplied by a factor characterizing the diffusion weighting and the number divided by three.

2. The method according to claim 1, It is characterized in that The factor that characterizes the diffusion weighting is a predetermined b-value.

3. The method according to claim 1 or 2, It is characterized in that The number is two or three, and / or a circular encoding is used.

4. The method according to claim 1 or 2, It is characterized in that The diffusion data set is recorded using a spin echo sequence and / or an EPI readout technique.

5. The method according to claim 1 or 2, It is characterized in that The diffusion gradient pulse sequence is ascertained in two spatial directions according to a double diffusion coding scheme either by means of a temporally optimized gradient profile along the two spatial directions or as a sequence of gradient pulses of identical shape and amplitude.

6. The method according to claim 1 or 2, It is characterized in that In at least one of the diffusion gradient pulse sequences, diffusion gradient pulses (6, 7) are output simultaneously in a plurality of spatial directions.

7. The method according to claim 6, It is characterized in that In at least one of the diffusion gradient pulse sequences, diffusion gradient pulses (6, 7) are output simultaneously on a plurality of physical gradient axes.

8. The method according to claim 6, It is characterized in that In at least one of the diffusion gradient pulse sequences, diffusion gradient pulses (6, 7) are output simultaneously on all physical gradient axes.

9. The method according to claim 1 or 2, It is characterized in that In case imaging gradient pulses (3, 4, 5) outputted predominantly on a first physical gradient axis are used to record the diffusion data set, the distribution of the diffusion gradient pulses (6, 7) is selected such that the diffusion gradient pulses are outputted predominantly on a physical gradient axis different from the first physical gradient axis for each diffusion gradient pulse sequence.

10. The method according to claim 9, It is characterized in that In a diffusion gradient pulse sequence of two, in a first gradient pulse sequence, a diffusion gradient pulse pair having a predetermined amplitude is output on a second physical gradient axis, and a diffusion gradient pulse pair having the predetermined amplitude divided by the square root of two is output on the first gradient axis, and in a second gradient pulse sequence, a diffusion gradient pulse pair having the predetermined amplitude is output on a third physical gradient axis, and a diffusion gradient pulse pair having the predetermined amplitude divided by the square root of two is output on the first gradient axis.

11. A magnetic resonance apparatus (8) having a control device (11) designed to carry out the method according to claim 1.

12. A computer program product comprising a computer program which, when executed on a control device (11) of a magnetic resonance device (8), performs the steps of the method according to any one of claims 1 to 10.

13. An electronically readable data carrier having stored thereon a computer program which, when executed on a control device (11) of a magnetic resonance device (8), executes the steps of the method according to any one of claims 1 to 10.