ASL-MRI USING A SEQUENCE OF MAGNETICALLY LABELED OR NON-LABELED FLUID BOLI

Description

[0001] The invention relates to a system for evaluating diagnostically usable MR data of a flowing fluid, a method for evaluating diagnostically usable MR data of a flowing fluid, and a computer program for evaluating diagnostically usable MR data of a flowing fluid. The fluid is preferably blood.

[0002] US patent US 8,260,396 B2 (Guenther 2012) discloses an MR system that magnetically labels blood in vivo within pre-determined spatial sections of the vessels, e.g., in the carotid artery. This is done using an arterial MR spin labeling process, also known as arterial spin labeling (ASL). These blood sections are called boluses and represent spatially defined areas of either labeled or unlabeled blood within a vessel. The latter serve as control or comparison sections and are often referred to as such. In the MR system described in (Guenther 2012), the blood sections are magnetically labeled or unlabeled at successive initial points in time, creating a flowing sequence of label and control boluses. Different combinations of label and control boluses generate different bolus sequences.Technically speaking, the different sequences of label / control boluses represent different codings of the blood. For coding, a Hadamard matrix is preferably used here, the elements of which determine whether a label or a control bolus is generated, and each row of the matrix determines a sequence of label and / or control boluses. Each row of the matrix is thus used as the basis for an MR measurement of initial MR data. For each of these blood bolus sequences, after a waiting time (often called inflow time (TI)), initial MR images, also called intermediate images, are acquired in a second region. This second region is the region to which the blood bolus sequence flows, in Guenther's case, for example, the brain. These initial MR images are then combined in a known manner in the decoding step, in particular by adding or subtracting them from one another.For decoding, the Hadamard matrix used for coding is preferably first inverted and then used in the actual decoding step to determine the mathematical operations necessary for combination. In particular, the second MR data are added to or subtracted from one another depending on the entry in a row of the matrix. In this way, second MR images are decoded, each indicative of a single blood bolus generated at a specific first time point and flowing to the second region. These each correspond to a first MR image that would have been determined if only a single, selected, marked blood bolus had been generated at the first time point. The various second MR images continue to correspond to different TIs.The MR system described by Guenther (2012) is very efficient because it allows clinically relevant MR data on organ perfusion and / or blood flow to be acquired in a comparatively short time and without deterioration of the signal-to-noise ratio. In an exemplary implementation of (Guenther (2012), N-1 reconstructed MR images can be generated from N measured MR images, and all N measured MR images are required to generate each of these N-1 reconstructed MR images. Using the N-1 reconstructed MR images, an MR signal can be determined at different waiting times TI, and thus the signal course over time can be determined voxel-by-pixel. From these signal-time curves, various parameters relevant for clinical diagnosis can then be determined voxel-by-pixel.

[0003] The article "Time-Encoded pseudoContinuous Arterial Spin Labeling: Basic Properties and Timing Strategies for Human Applications" by W. Teeuwisse et al., Magnetic Resonance in Medicine, Volume 72, pages 1712 to 1722 (2014) discloses a time-encoded pseudocontinuous ASL (pCASL) method in which, for example, the duration of a block is variable to compensate for T1 decay or in which the postlabeling delay time is used to detect arterial transit time (ATT).

[0004] It is an object of the present invention to provide a system, a method and an associated computer program that enable the generation of MR data of a flowing fluid to be improved.

[0005] The object is achieved by a system for evaluating diagnostically usable MR data of a flowing fluid according to claim 1, the system comprising: a data acquisition unit adapted to perform multiple MR measurements and, in each MR measurement, measures first MR data by, in each magnetic resonance measurement, a) at different first points in time in a first region, either magnetically marking or not marking the fluid, such that a flowing sequence of marked and / or unmarked fluid boluses is created, and b) at a second point in time in a second region, at which the fluid flows from the first region, measuring the first MR data, which are indicative of the generated sequence of marked and / or unmarked fluid boluses, wherein the data acquisition unit is adapted to generate a complementary sequence of marked and / or unmarked fluid boluses for each sequence of marked and / or unmarked fluid boluses, wherein the magnetic resonance measurements of the first magnetic resonance data of mutually complementary sequences are performed immediately one after the other,and provide the first magnetic resonance data thus measured, or a data providing unit adapted to provide first magnetic resonance data measured according to a) and b), and a matrix providing unit for providing a two-valued determination matrix, wherein a row of the determination matrix represents a sequence of labeled and / or unlabeled fluid boluses and different rows of the determination matrix are used as the basis for different magnetic resonance measurements of the first magnetic resonance data, wherein the determination matrix is formed by a combination of a Walsh-sorted Hadamard matrix with a complementary Walsh-sorted Hadamard matrix, wherein the formed matrix alternately comprises the rows of the Walsh-sorted Hadamard matrix and the complementary Walsh-sorted Hadamard matrix, an evaluation unit adapted,based on a combination of the first MR data measured in the various MR measurements, taking into account the determination matrix, i) to determine second single fluid bolus magnetic resonance data, each indicative of a single fluid bolus generated at a specific first time point and flowing to the second region, based on a combination of the first magnetic resonance data measured in the various magnetic resonance measurements, and ii) to determine second combination fluid bolus MR data, each indicative of a combination fluid bolus and a time interval between one of the first time points and the second time point, wherein the combination fluid bolus is composed of a plurality of fluid boluses (L, C) marked / unmarked at different first time points and flowing to the second region, wherein the data acquisition unit is adapted to measure the first magnetic resonance data in an order that enablesthat, even before all first magnetic resonance data required to determine the second single fluid bolus magnetic resonance data have been measured, second combination fluid bolus magnetic resonance data can be determined in the evaluation unit, characterized in that the Walsh-sorted Hadamard matrix and the complementary Walsh-sorted Hadamard matrix each have a degree 2 k< with k ≥ 2, and the evaluation unit is adapted to evaluate the first magnetic resonance data in an order that allows second combination fluid bolus magnetic resonance data to be determined even before all first magnetic resonance data required to determine the second single fluid bolus magnetic resonance data have been measured,wherein fewer first magnetic resonance data are used to determine second magnetic resonance data for a combination fluid bolus than to determine second magnetic resonance data for a single fluid bolus.

[0006] The marking / non-marking according to the above-mentioned feature a) can also be regarded as a coding of the fluid - as already done in Guenther 2012. The preceding assembly, which could also be called a combination, consists, as is well known (cf. Guenther 2012), in particular of mathematical operations such as addition and subtraction of the first MR data.

[0007] By using the combination fluid bolus(es), the possibility is created, explained in more detail below, of obtaining information about the flow behavior of the fluid before all of the first MR data have been measured. Therefore, it is not necessary, for example, to wait for an entire measurement process to complete and then determine the second MR data, as described in the publication Guenther 2012 cited above. Particularly in this sense, the system, which can also be referred to as an MR system, can improve the generation of MR data of a flowing fluid.

[0008] As a result, the combination fluid bolus, composed of two or more contiguous labeled / unlabeled individual fluid boluses, effectively creates a comparatively longer bolus, referred to above as a combination fluid bolus. A combination fluid bolus can therefore also be understood as a sequence of two or more individual fluid boluses.

[0009] The level of detail of the second MR data decreases with fewer and longer fluid boluses, i.e., ultimately, with longer sampling intervals. However, this creates the possibility of generating the second MR data from fewer initial MR data sets, thus allowing for earlier acquisition of relevant second MR data, such as clinically relevant information regarding organ perfusion. This information can also be advantageously considered in the further control of the MR measurement.

[0010] The flowing fluid is preferably blood. However, the flowing fluid can also be another liquid or gaseous substance; for example, another body fluid. The first MR data are preferably two-dimensional or three-dimensional MR images consisting of pixels or voxels.

[0011] According to the invention, the evaluation unit is adapted such that fewer first MR data are used to determine those second MR data that are each indicative of a combination fluid bolus and a time interval between one of the first times and one of the second times than for determining those second MR data that are each indicative of an individual fluid bolus. This means that if, for example, some first MR data should be erroneous, for example because a patient has moved, so that second MR data that are each indicative of an individual fluid bolus cannot be determined, second MR data that are each indicative of a combination fluid bolus and a time interval between one of the first times and the second time could still be determined.Furthermore, in principle, even before all first MR data required to determine the second MR data, each indicative of a single fluid bolus, have been measured, second MR data could be determined, each indicative of a combination fluid bolus and a time interval between one of the first time points and the second time point. For example, it is conceivable that the second MR data, each indicative of a combination fluid bolus and a time interval between one of the first time points and the second time point, could be determined in real time during the measurement.

[0012] According to the invention, the MR system comprises a matrix provision unit for providing a two-valued determination matrix, i.e. a matrix whose elements have one of two possible values. A row of the determination matrix represents a sequence of labeled and / or unlabeled fluid boluses, with different rows of the determination matrix being used as a basis for different MR measurements of the first MR data. The determination matrix can also be understood as a matrix that determines the coding of the fluid boluses. The evaluation unit is preferably adapted to a) form a vector, with one vector element being formed by an MR value of the first MR data and different vector elements corresponding to different MR measurements of the first MR data, b) invert the determination matrix, and c) determine the second MR data by applying the inverted determination matrix to the vector, in particular by multiplying it.The entries of a two-valued determination matrix have only two values, for example, +1 and -1, representing labeled and unlabeled fluid boli. The determination matrix therefore comprises several matrix elements, each of which can have one of two possible values. The term "two-valued" therefore refers to the number of possible values of a matrix element. However, it does not refer, for example, to the number of matrix elements.

[0013] A first MR value, i.e., an MR value of the first MR data, which forms a vector element, is preferably an image value, in particular a voxel value or a pixel value, of a first MR image. This means that the second MR data can be determined pixel by pixel or voxel by voxel, with a vector being formed for each pixel or for each voxel of the different first MR images, which in this example represent the first MR data. For example, a vector is formed for the pixel at location (1,1), which vector has the corresponding image values at locations (1,1) of the first MR images acquired at the different first points in time.

[0014] In one embodiment, in the decoding step (such a step is already known in principle from Guenther 2012), for example, first MR images are combined as first MR data, in particular added and / or subtracted, in order to generate second MR images as second MR data. The generation of a second MR image by combining a plurality of first MR images can, in one embodiment, be carried out by adding and / or subtracting the first MR images to form a second MR image according to the entries in each row of the inverted determination matrix. The inverted determination matrix can also be understood here as a decoding matrix. This can be realized computationally, for example, as follows: A vector is formed in which each vector entry is assigned to exactly one first MR image and corresponds to the voxel value at exactly one predetermined voxel location of the respective first MR image.The inverted determination matrix is then multiplied by this vector. The result is a second vector in which each vector entry is in turn assigned to exactly one second MR image and corresponds to the voxel value at the same, previously defined voxel location of the respective second MR image. In the described embodiment, this method is performed for each voxel of the first and second MR images. For example, to determine a voxel value for the voxel location (1,1,1) of a second MR image, a first row of the inverted determination matrix can be multiplied by a vector comprising the voxel values contained in the different first MR images at the voxel location (1,1,1).To determine the voxel value located at position (1,1,2) for this second MR image, the same first row of the inverted determination matrix can be multiplied by a vector containing the voxel values of the first MR images at voxel position (1,1,2). This is performed for all voxels, so that the voxel values of all voxel positions for this second MR image are determined in this way. Different rows of the inverted determination matrix then each correspond exactly to a second MR image.

[0015] According to the invention, the data acquisition unit is further adapted to generate a complementary sequence for each sequence of marked and / or unmarked fluid boluses, wherein a first sequence is complementary to a second sequence if an unmarked fluid bolus is arranged at each location in the first sequence that corresponds to a location in the second sequence at which a marked fluid bolus is arranged, and vice versa. In one embodiment, a first MR image is generated for each flowing sequence of marked and / or unmarked fluid boluses.Since a complementary sequence is preferably present for each sequence of labeled and / or unlabeled fluid boluses, a further first MR image is preferably generated for each first MR image, which has been generated based on a flowing sequence of labeled and / or unlabeled fluid boluses that is complementary to the flowing sequence of labeled and / or unlabeled fluid boluses on which the preceding first MR image is based. Accordingly, in a preferred embodiment, for each first MR image generated based on a first flowing sequence, there is a further first MR image generated based on a flowing sequence complementary to the first.

[0016] The provision of complementary sequences of labeled / unlabeled fluid boluses has the following advantage over the aforementioned publication Guenther 2012: There, only N-1 second MR data can be determined from N first MR data. For example, second MR data indicative of the first fluid bolus cannot be generated. However, using the information from the two complementary fluid bolus sequences, second individual MR data indicative of all N fluid boluses can be generated. Furthermore, second MR combination data indicative of combination fluid boluses containing the first fluid bolus can also be generated from the first MR data.

[0017] According to the invention, the data acquisition unit is adapted to perform MR measurements of the first MR data, which have mutually complementary sequences, directly one after the other. This means that after a first MR measurement with a first fluid bolus sequence, a second such MR measurement is performed with a second sequence that is complementary to the first sequence, after which third and fourth MR measurements of first MR data with third and fourth sequences are performed, wherein the fourth sequence is complementary to the third sequence, and so on. The data acquisition unit can therefore be programmed, for example, to first generate a first MR image based on a first flowing sequence of labeled and / or unlabeled fluid boluses, after which a further first MR image is generated based on a second sequence of labeled and / or unlabeled fluid boluses that is complementary to the first sequence.Thereafter, a further first MR image can be generated by means of the correspondingly programmed data acquisition unit, which is based on a third sequence of labeled and / or unlabeled fluid boluses, after which a further first MR image can be generated which is based on a fourth sequence of labeled and / or unlabeled fluid boluses which is complementary to the third sequence of labeled and / or unlabeled fluid boluses, and so on.

[0018] In one embodiment, the rows of the Walsh-sorted Hadamard matrix represent a first group of sequences of labeled and / or unlabeled fluid boluses, wherein different rows of the Walsh-sorted Hadamard matrix are based on different MR measurements of the first MR data. Furthermore, the rows of the complementary Walsh-sorted Hadamard matrix represent a second group of sequences of labeled and / or unlabeled fluid boluses that are complementary to the sequences of the first group, and wherein different rows of the complementary Walsh-sorted Hadamard matrix correspond to different MR measurements of the first MR data.In addition, the evaluation unit is adapted to a) form a first vector, wherein a vector element of the first vector is formed by an MR value of the first MR data and different vector elements of the first vector correspond to different MR measurements, which in turn correspond to different sequences of the first group, b) form a second vector, wherein a vector element of the second vector is formed by an MR value of the first MR data and different vector elements of the second vector correspond to different MR measurements, which in turn correspond to different sequences of the second group, c) invert the Walsh-sorted Hadamard matrix and the complementary Walsh-sorted Hadamard matrix, and d) determine the second MR data by applying the inverted Walsh-sorted Hadamard matrix to the first vector and the inverted complementary Walsh-sorted Hadamard matrix to the second vector, in particular by multiplication.

[0019] According to the invention, there is therefore a first group of sequences of labeled and / or unlabeled fluid boluses, wherein a first MR image can be generated for each sequence of this first group. Since the rows of the Walsh-sorted Hadamard matrix represent the sequences of labeled and / or unlabeled fluid boluses of the first group, and since a respective sequence of labeled and / or unlabeled fluid boluses characterizes an MR measurement in which a respective first MR image has preferably been generated, an individual row of the Walsh-sorted Hadamard matrix corresponds to an individual MR measurement, which has preferably led to a single first MR image, and different rows of the Walsh-sorted Hadamard matrix correspond to different MR measurements, which have preferably led to different first MR images.

[0020] In one embodiment, the Walsh-sorted Hadamard matrix is inverted, and the inverted Walsh-sorted Hadamard matrix is multiplied by a first vector comprising the different voxel values of the different first MR images at the same voxel location (1,1,1) to calculate voxel values at the same voxel location (1,1,1) for second MR images. A first vector is also generated for each of the other voxel locations and multiplied by the inverted Walsh-sorted Hadamard matrix to generate voxel values at the corresponding voxel locations for the second MR images. In this way, a first group of second MR images is generated for the first group of sequences of labeled and / or unlabeled fluid boluses.In the same way, a complementary Walsh-sorted Hadamard matrix can be generated and inverted for a second group of sequences of labeled and / or unlabeled fluid boluses, wherein the inverted complementary Walsh-sorted Hadamard matrix can be multiplied by corresponding second vectors to generate second MR data also for the second group of sequences of labeled and / or unlabeled fluid boluses.

[0021] According to the invention, the data acquisition unit is adapted to perform MR measurements sequentially ordered in time according to the number of changes between marked and / or unmarked fluid boluses in the corresponding sequence. The MR measurements are ordered sequentially in time according to a Walsh sorting. This means that the determination matrix described above has rows arranged according to the Walsh sorting. In one embodiment, a data acquisition unit, i.e. in particular a corresponding MR system, can therefore be set such that first a first MR image is generated based on a sequence of marked and / or unmarked fluid boluses that has the fewest number of changes between marked and / or unmarked fluid boluses. Next, another first MR image could be generated based on a sequence that has the second fewest changes between marked and / or unmarked fluid boluses, and so on.If multiple sequences have the same number of alternations between labeled and / or unlabeled fluid boluses, the corresponding first MR images are generated sequentially before another first MR image is generated based on a sequence having a larger number of alternations between labeled and / or unlabeled fluid boluses.

[0022] In a preferred embodiment, the evaluation unit is adapted to determine addition and / or subtraction data by adding and / or subtracting first MR data measured in different MR measurements from one another, and to determine the second MR data based on the addition and / or subtraction data. This means, for example, that corresponding pixel values or voxel values from first MR images of different MR measurements are added and / or subtracted to determine the second MR data, which are each indicative of a combination fluid bolus—composed of several fluid boluses generated at different first times and flowing to the second region—and a time interval between one of the first times and the second time.

[0023] The above-mentioned object is also achieved by a method for evaluating diagnostically usable MR data of a flowing fluid according to claim 4.

[0024] The above-mentioned object is further achieved by a computer program for evaluating MR data of a flowing fluid according to claim 6.

[0025] It should be understood that the system according to claim 1, the method according to claim 4 and the above-mentioned computer program have similar and / or identical preferred embodiments as particularly defined in the dependent claims.

[0026] It should also be understood that a preferred embodiment of the present invention can be formed by any combination of the dependent claims and / or the embodiments mentioned in the present application with the respective independent claims, provided that the respective combination falls within the scope of the invention defined by the claims.

[0027] Embodiments of the invention are described with reference to the following figures, wherein Fig. 1 schematically and exemplarily shows an embodiment of an MR system, Fig. 2 schematically and exemplarily shows various sequences of labeled and unlabeled fluid boluses and second MR data, Fig. 3 schematically and exemplarily shows an embodiment of an evaluation system for evaluating first MR data and Fig. 4 shows a flow chart illustrating an MR method.

[0028] Fig. 1 shows schematically and by way of example an embodiment of an MR system. The MR system 1 comprises a data acquisition unit 5 for acquiring first MR data of a patient 3 arranged on a patient couch 2. In this embodiment, the data acquisition unit 5 is designed to perform a plurality of MR measurements in the head region 4 of the patient 3. In each MR measurement, first MR data are measured by a) at various first points in time, the blood flowing in one or more blood vessels, in particular arteries, in a first vessel section in the patient 3 is either magnetically marked or not marked in order to generate a flowing sequence of marked and / or unmarked fluid boluses, here marked and / or unmarked blood boluses ( Fig. 2 ; L,C), and b) at a second time in a second region to which the blood flows from the first location, i.e. the aforementioned vessel section, the first MR data, which are indicative of the generated sequence of labeled and / or unlabeled blood boluses, are measured. The MR system 1 further comprises an evaluation unit 7 for evaluating the first MR data measured in the various MR measurements, wherein the evaluation unit 7 is adapted to determine second MR data, which are each indicative of a combination blood bolus - composed of several blood boluses flowing to the second region at different first times - and a time interval between one of the first times and the second time. The determination of the second MR data is based on a combination of the first MR data measured in the various MR measurements.This combination comprises in particular adding and subtracting first MR data, wherein in particular the elements of a decoding matrix determine whether subtraction or addition takes place.

[0029] This can, for example, analogous to Guenther 2012, also be described by multiplying the decoding matrix by a vector whose individual entries are formed from an image value, e.g., the value of a voxel, of each of the first MR data. The evaluation unit 7 is further adapted to also determine second MR data that are indicative of a single blood bolus generated at a specific first point in time and flowing to the second region, again based on a combination of the first MR data measured in the various MR measurements. In this embodiment, the first and second MR data are MR images.

[0030] The MR system 1 further comprises a matrix provision unit 6 for providing a two-value determination matrix, wherein each row of the determination matrix is formed by a sequence of marking and / or non-marking regions. These marking / non-marking regions determine whether the blood is marked or not and, to this extent, correspond to the marked / non-marked blood boli. The determination matrix thus contains the specification for coding the individual blood boli. The various MR measurements of the first MR data correspond to different rows of the determination matrix, more precisely, to different sequences of blood boli that were coded according to the determination matrix. Furthermore, the evaluation unit 7 is adapted to form a vector, wherein a vector element is formed by a first MR value, which in this embodiment corresponds to an image value, and different vector elements correspond to different MR measurements.The evaluation unit 7 is further adapted to invert the determination matrix - i.e. to make it a decoding matrix - and to determine the second MR data by applying the inverted determination matrix to the vector.

[0031] In this embodiment, 16 MR measurements are performed, with each MR measurement generating a sequence of eight labeled and / or unlabeled blood boluses. The coding of the sequences of blood boluses in the different MR measurements can, for example, be represented by the following matrix M: M = + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 − 1 − 1 − 1 − 1 − 1 − 1 − 1 − 1 + 1 + 1 + 1 + 1 − 1 − 1 − 1 − 1 − 1 − 1 − 1 − 1 + 1 + 1 + 1 + 1 + 1 + 1 − 1 − 1 − 1 − 1 + 1 + 1 − 1 − 1 + 1 + 1 + 1 + 1 − 1 − 1 + 1 + 1 − 1 − 1 + 1 + 1 − 1 − 1 − 1 − 1 + 1 + 1 − 1 − 1 + 1 + 1 + 1 − 1 − 1 + 1 + 1 − 1 − 1 + 1 − 1 + 1 + 1 − 1 − 1 + 1 + 1 − 1 + 1 − 1 − 1 + 1 − 1 + 1 + 1 − 1 − 1 + 1 + 1 − 1 + 1 − 1 − 1 + 1 + 1 − 1 + 1 − 1 − 1 + 1 − 1 + 1 − 1 + 1 − 1 + 1 + 1 − 1 + 1 − 1 + 1 − 1 + 1 − 1 + 1 − 1 + 1 − 1 − 1 + 1 − 1 + 1 − 1 + 1 − 1 + 1

[0032] Here, "-1" denotes a non-labeling region and "+1" a labeling region, thus determining whether the respective blood bolus is unlabeled or labeled. The matrix M has a complementary sequence for each sequence of labeled and / or unlabeled blood boluses, i.e., for each row, with complementary sequences arranged directly one after the other. Furthermore, the individual sequences are sorted according to the number of alternations between labeled and unlabeled blood boluses. This means that the rows of the matrix M are Walsh-sorted. Walsh sorting is described, for example, in Wolfram, S., 2002. A new kind of science 1st ed., Wolfram Media, Inc. and Wang, R., 2013. Sequency Ordered Walsh-Hadamard Matrix, http: / / fourier.eng.hmc.edu / e 16 1 / lectures / wht / node3.html (Wolfram 2002).

[0033] The matrix M is formed by a combination of a first determination matrix, which according to the invention is a Hadamard matrix, with a second determination matrix, which according to the invention is a complementary Hadamard matrix, i.e. one corresponding to the complementary sequences of labeled and / or unlabeled fluid boluses, wherein the rows of the Hadamard matrix and the complementary Hadamard matrix are used alternately to generate the matrix M. The individual sequences of labeled and unlabeled blood boluses, which correspond to the individual rows of the matrix M, are shown schematically and by way of example in the Fig. 2 illustrated. Fig. 2 shows 16 sequences 10 of labeled boluses L and unlabeled boluses C, which could also be considered control boluses. In Fig. 2 The sequences indexed with odd numbers (rows 1, 3, 5,..., 15), which can also be called sequences of a first group, correspond to rows of the Hadamard matrix, and the sequences indexed with even numbers (rows 2, 4, 6,..., 16), which can also be called sequences of a second group, correspond to rows of the complementary Hadamard matrix.

[0034] The second MR data obtained from the first MR data, each of which is indicative of a blood bolus generated at a specific first point in time and flowing to the second region, are preferably determined voxel by voxel. This means that the first MR data are preferably first MR images with different voxels, wherein a first vector and a second vector are further formed. A vector element of the first vector is formed by a voxel value of a first MR image, wherein different vector elements of the first vector correspond to different MR measurements, which in turn correspond to different sequences of the first group (rows 1, 3, 5,..., 15). A vector element of the second vector is formed by a voxel value of a second MR image, wherein different vector elements of the second vector correspond to different MR measurements, which in turn correspond to different sequences of the second group (rows 2, 4, 6,..., 16).For example, a first vector for the voxel (1,1,1) comprises the voxel values for this voxel in the different first MR images generated in the different MR measurements corresponding to the different sequences of blood boluses of the first group (rows 1, 3, 5,...,15). For each voxel, therefore, second MR data, i.e., voxel values for the corresponding voxel, are preferably determined, each indicative of a blood bolus generated at a specific first point in time and flowing to the second region, by multiplying the inverse first determination matrix by the first vector and the inverse second determination matrix by the second vector for each voxel. The two resulting vectors can, for example, be averaged, with the resulting averaged vector representing the second MR data, for example, for the voxel (1,1,1).

[0035] Furthermore, the evaluation unit 7 is adapted such that, for determining the second MR data, which are each indicative of a combination blood bolus composed of several blood boluses generated at different first times and flowing to the second region, and a time interval between one of the first times and the second time, fewer first MR data are combined than for determining the second MR data, which are each indicative of a single blood bolus generated at a specific first time and flowing to the second region. This will be explained below with reference to Fig. 2 explained by example.

[0036] In this exemplary embodiment, the evaluation unit 7 subtracts the first MR image of the first MR measurement, which is based on the first row of the first group (row 1) of the blood bolus sequences 10, from the first MR image of the second MR measurement, which is based on the first row of the second group (row 2) of the blood bolus sequences 10, in order to generate the second MR image 30, i.e., corresponding second MR data. The second MR image 30 is indicative of the combination blood bolus, which is formed from all eight individual blood boluses, i.e., has eight times the length of an individual blood bolus, and of the time interval TI between the second time at which the first MR images were acquired and the first time at which the blood bolus 48 was marked. From the perspective of a combination blood bolus, the individual blood boluses could also be referred to as subboluses or sub-boluses.The second MR images 31, 32, 33 each correspond to a combined blood bolus formed from four individual blood boluses, i.e., four times the length of a single blood bolus. The second MR image 31 was generated by subtracting the first MR image of the fourth MR measurement (4th line) from the first MR image of the first MR measurement (1st line). The second MR image 32 was generated by subtracting the first MR image of the fifth MR measurement (5th line) from the first MR image of the first MR measurement (1st line), and the second MR image 33 was generated by subtracting the first MR image of the third MR measurement (3rd line) from the first MR image of the first MR measurement (1st line). The second MR image 31 corresponds to a time interval TI between the second time point and the time point at which the blood bolus 49 was marked.The second MR image 32 corresponds to a time interval TI between the second time point and the time point at which the blood bolus 50 was marked, and the second MR image 33 corresponds to a time interval TI between the second time point and the time point at which the blood bolus 48 was marked. The length of the individual or combination blood boluses used to generate the blood bolus shown in . Fig. 2 The sequence of the second MR images 30-48 used is illustrated again below the corresponding images in the form of horizontal bars. Subtracting and adding consecutively measured MR images is a fundamental principle of time-coded ASL (cf. Guenther 2012). The inventive Walsh sorting of the determination matrix performed here automatically ensures that the decoded second MR images are based on increasingly shorter combination blood boluses as the number of MR measurements increases.

[0037] A method for determining the decoding rules of the second MR images from the first MR images for matrices of any size, and according to which, for example, images 30-48 were also determined, can be as follows. First, for each second MR image, which is indicative of a single fluid bolus generated at a specific first time and flowing to the second region, it is written down symbolically, i.e., written in symbols, how these second MR images b 1 ,...,b N from the first MR images v 1 ,... v N put together, or the v i to the b i be put together.

[0038] For example, using the first group (rows 1, 3, 5,...,15) of matrix M (see equation 1) (the same procedure can be used for the second group (rows 2, 4, 6,...,16 of matrix M)): b 2 = 1 8 ν 1 + v 2 + ν 3 + ν 4 − ν 5 − ν 6 − ν 7 − ν 8

[0039] It is important to note that this step can be performed symbolically before the first MR data is generated, and no second MR images are generated. The symbolic representation of the individual fluid boluses serves only to derive arithmetic calculation rules for the second MR images, each of which is indicative of a combination fluid bolus generated at a specific first time point and flowing to the second region.

[0040] In a second step, the b i symbolically by adding and / or subtracting all desired second MR data K i The procedure is such that the fluid boluses for the b i Indicative are the combination fluid boluses which result in the K i are indicative. From this symbolic notation, after deleting all summands that cancel out to zero, the arithmetic calculation rule for the desired combination fluid bolus results.

[0041] For example, using the first group (rows 1, 3, 5,..., 15) of matrix M (see equation 1 ) (the same procedure can be used for the second group (rows 2, 4, 6,..., 16 of M)): b 2 + b 3 = 1 8 ν 1 + v 2 + ν 3 + ν 4 − ν 5 − ν 6 − ν 7 − ν 8 + ν 1 + v 2 − ν 3 − ν 4 − ν 5 − ν 6 + ν 7 + ν 8 = 1 4 ν 1 + ν 2 − ν 5 − ν 6 or b 5 + b 6 + b 7 + b 8 = 1 2 ν 1 − ν 2

[0042] If, for example, second MR images are to be determined that are indicative of the first fluid bolus or combination fluid boluses containing the first fluid bolus, the first MR images from the first and second groups are combined. In this example, the first fluid bolus refers to the fluid bolus with the greatest time interval TI between its generation and the second time, and the last fluid bolus refers to the fluid bolus with the shortest time interval TI between its generation and the second time. To determine the second MR images that are indicative of the first fluid bolus, the summands with a negative sign in the calculation rule for the second MR images that are indicative of the last fluid bolus are simply replaced by entries of the second vector, i.e., image values of the first MR images vik< (second group) belonging to the complementary matrix, with the same indices.Similarly, the calculation rules for determining second MR data indicative of combination fluid boluses containing the first fluid bolus can be adapted.

[0043] For example, given the first (rows 1, 3, 5,..., 15) and second (rows 2, 4, 6,..., 16) group of matrix M (see equation 1): b 1 + b 2 + b 3 + b 4 = 1 2 ν 1 − ν 2 k

[0044] In this example, the determination of second MR data indicative of the first fluid bolus or combination fluid boluses containing the first fluid bolus was only made possible by combining the information of the determination matrix and the determination matrix complementary to it.

[0045] The second MR images, i.e., the second MR data corresponding to combination fluid boluses, or more precisely, are indicative of them, are generated using fewer first MR data than the second MR data corresponding to individual fluid boluses. In particular, in the latter case, second MR data are determined for each of the eight individual fluid boluses, i.e., for the associated eight first points in time, by combining at least eight first MR images, in particular, in the case of the aforementioned composite determination matrix (M, cf. Equation 1), all 16 first MR images corresponding to the 16 sequences 10, using the aforementioned first and second determination matrices. In contrast, for determining the second MR data corresponding to the combination fluid boluses, only two first MR images are required, for example, the first MR images based on rows 1 and 2 of the determination matrix M.The second MR data corresponding to combination fluid boluses can therefore be generated even if not all 16 first MR images have been acquired. This means that the second MR images corresponding to the combination fluid boluses can already be generated before all 16 first MR images have been measured. For example, second MR images can already be generated during the measurement of the 16 first MR images, in particular in real time. Furthermore, the second MR images can be generated even if, for example, not all 16 first MR images could be measured due to image artifacts or other interference.

[0046] Using the second MR images, for example, an MR signal can be determined at different time points, thus determining the signal profile over time voxel-by-pixel. From these signal-time curves, various parameters relevant for clinical diagnosis can then be determined voxel-by-pixel. Examples of this are described in Buxton, RB et al., 1998. A general kinetic model for quantitative perfusion imaging with arterial spin labeling. Magnetic Resonance in Medicine, 40(3), pp. 383-96.

[0047] The Fig. 1 The MR system 1 shown further comprises a control unit 8 adapted to control the various units of the MR system 1, and an input unit 9 such as a keyboard, a computer mouse, a touchscreen, etc. Furthermore, the MR system 1 comprises an output unit 25 such as a monitor.

[0048] In the following, an MR method for generating MR data of a flowing fluid will be described with reference to a flowchart shown in Fig. 3 is shown.

[0049] In step 101, first MR data are acquired, wherein multiple MR measurements are performed, and wherein first MR data are measured in each MR measurement. In particular, a first MR image is generated in each MR measurement to measure first MR data. During the generation of the first MR data, the fluid is either magnetically marked or unmarked at various first points in time in a first region in order to generate a flowing sequence of marked and / or unmarked fluid boluses. Furthermore, at a second point in time in a second region to which the fluid flows from the first region, the first MR data, which are indicative of the generated sequence of marked and / or unmarked fluid boluses, are measured. This means that in each MR measurement, a respective sequence of marked and / or unmarked fluid boluses is created, and for each of these sequences, a first MR image is preferably generated as a measurement of the first MR data.

[0050] In addition, in step 101, second MR data are determined, each of which is indicative of a combination fluid bolus—composed of several labeled / unlabeled fluid boluses generated at different first points in time and flowing to the second region—and a time interval TI between one of the first points in time and the second point in time, by combining the first MR data measured in the various MR measurements, as already described. Since not all of the first MR data are necessary to determine this second MR data, second MR data for the combination fluid boluses can already be determined during the acquisition of the first MR data, that is, in this exemplary embodiment, the first MR images.

[0051] After the measurement of the first MR data has been completed, these can also be used in step 102 to determine those second MR data which are indicative of a single fluid bolus generated at a specific first time point and flowed to the second region.

[0052] Fig. 4 shows schematically and by way of example an evaluation system 20 for evaluating MR data of a flowing fluid. The evaluation system 20 comprises a data provision unit 21 for providing first MR data, wherein the data provision unit 21 is adapted to provide first MR data of a plurality of MR measurements, and wherein in each MR measurement, first MR data were measured in that a) at various first points in time in a first region, the fluid was either magnetically marked or not marked in order to generate a flowing sequence of marked and / or unmarked fluid boluses, and b) at a second point in time in a second region, at which the fluid flows from a first region, the first MR data, which are indicative of the generated sequence of marked and / or unmarked fluid boluses, were measured.The evaluation system 20 further comprises an evaluation unit 22, which corresponds to the above-described evaluation unit 7 of the MR system 1, for evaluating the first MR data measured in the various MR measurements. The evaluation unit 22 is adapted to determine second MR data, each of which is indicative of a combination fluid bolus—composed of several labeled / unlabeled fluid boluses generated at different first times and flowing to the second region—and a time interval between one of the first times and the second time, based on the previously described combination of the first MR data measured in the various MR measurements.The evaluation unit 22 is preferably also adapted to determine second MR data, each indicative of a single fluid bolus generated at a specific first time point and flowing to the second region, based on a combination of the first MR data measured in the various MR measurements. Furthermore, the evaluation system 20 comprises an input unit 23, such as a keyboard, a computer mouse, or a touchscreen, and an output unit 24, such as a monitor.

[0053] Although in the embodiments described above the first MR data were measured only at a second time point, in other embodiments the first MR data may also be measured at a plurality of different second time points.

[0054] For example, first MR data in each MR measurement can also be acquired at several second time points t 1 <t 2 <t 3 <... gemessen werden. Für jeden Zeitpunkt t i können dann zweite MR-Daten erzeugt werden. Damit könnte dann zum Beispiel die Signalintensität an mehr Zeitpunkten ermittelt werden, nämlich dem Produkt der Anzahl von zweiten Zeitpunkten und der Anzahl von zweiten MR-Daten, als es mit einer Messung zu nur einem zweiten Zeitpunkt möglich wäre. Damit stünden mehr Datenpunkte für eine Signal-Zeit-Kurve zur Verfügung, was zum Beispiel ein Fitting verbessern kann.

[0055] The following are some exemplary explanations of individual aspects of the revelation.

[0056] For example, the data acquisition unit can be adapted to cause a rotation of the blood magnetization by a desired angle in order to mark the blood at a first point in time and in a selected section, e.g. the carotid artery. In one exemplary embodiment, the magnetization resulting from the spins of the protons in the blood is rotated (inverted) by 180° from the relaxed state (0°). This is also referred to here as marking or labeling. In this way, a clearly defined length section of the flowing blood (bolus, marking area) is created. The blood with inverted magnetization locally changes the magnetic properties wherever it flows and thus the contrast of an MR image generated after labeling (label image). Since these changes are very small and not visible to the naked eye, an additional image without prior labeling, the so-called control image, is generated.For example, the difference between the label and control images makes the effects of perfusion and / or flow visible and, if necessary, quantifiable. In another exemplary embodiment, instead of inversion, a saturation (rotation by 90°) of the magnetization is achieved.

[0057] The rotation (also called tilt or flip) of the magnetization can be achieved, for example, by radiofrequency pulses (RF pulses). An example of such RF pulses is described in Payne, GS & Leach, MO, 1997. Implementation and evaluation of frequency offset corrected inversion (FOCI) pulses on a clinical MR system. Magnetic resonance in medicine, 38(5), pp. 828-33 and Dai, W. et al., 2008. Continuous flow-driven inversion for arterial spin labeling using pulsed radio frequency and gradient fields. Magnetic resonance in medicine, 60(6), pp. 1488-97.

[0058] This blood labeling technique can, for example, be used to generate not just a single labeled or unlabeled fluid bolus prior to measuring an MR image. Rather, sequences of labeled and unlabeled fluid boluses can be created, and multiple MR measurements of initial MR data can be performed sequentially, potentially generating a different sequence of labeled and unlabeled fluid boluses in each measurement. The specific form of these sequences can, for example, be specified in the form of a matrix, with the matrix entries specifying the type of labeling (e.g., labeled or unlabeled), and each row of the matrix corresponding to an MR measurement of initial MR data.

[0059] For an example description of the generation of initial MR data, see Gadian, DG, 1996. NMR & Its Applications to Living Systems 2 Sub., Oxf.UP; Bernstein, M., King, K. & Zhou, X., 2004. Handbook of MRI pulse sequences, Academic Press. For further details of the labeling process and the generation of initial MR data, see Williams, DS et al., 1992. Magnetic resonance imaging of perfusion using spin inversion of arterial water. Proceedings of the National Academy of Sciences of the United States of America, 89(1), pp.212-6, Detre, J., Leigh, J. & Williams, D., 1996. Perfusion imaging. Magnetic Resonance in Medicine, 35(1), pp.70-9, Günther, M., 2007. Habilitationsschrift. Ruprechts-Karl-Universität Heidelberg (Günther 2007) and Günther, M., Oshio, K. & Feinberg, DA, 2005. Single-shot 3D imaging techniques improve arterial spin labeling perfusion measurements. Magnetic resonance in medicine, 54(2), pp. 491-498.

[0060] For example, to obtain second MR data indicative of a combined blood bolus (hereinafter also referred to as second MR combination data), only subsets of all first MR data can be used from the first MR data, and these can still be relevant for clinical diagnosis. In other words, in this example, fewer first MR data are required to reconstruct second MR combination data than to reconstruct second MR data based on individual fluid boluses (hereinafter referred to as second individual MR data). This is based, for example, on the following observation. If the calculation of the second MR combination data from second individual MR data is written down arithmetically in symbols, this observation becomes clear.If, in this calculation, the second individual MR data are written, again in symbols, as a combination of addition and subtraction of the first MR data, it results that some of the first MR data calculate pairwise to zero. From this representation, it follows that the second MR combination data can be described as a combination of addition and subtraction of subsets of first MR data and that fewer first MR data are required to determine second MR combination data than to determine second individual MR data. In the sense of this example, the second MR combination data are generated directly from the first MR data (and not via the detour of generating second individual MR data). This can be advantageous compared to the prior art because, for example, the calculation of second MR combination data performed in the evaluation unit can be accelerated.This is because the detour via calculation of the second individual MR data is no longer necessary, as in (Günther 2007). The second individual MR data therefore only need to be generated when the highest possible detail resolution, in particular temporal resolution, is required. Furthermore, the robustness of the measurement can be increased, for example. If some of the initial MR data are faulty, for example because a patient has moved, or are unavailable, the second individual MR data may not be determined or may only be determined incorrectly. Nevertheless, second MR combination data could still be determined from the intact and available initial MR data.

[0061] For example, if the determination matrix is invertible, meaning its inverse matrix can be determined, it is possible to determine second MR individual and combination data using this matrix and the vector consisting of first MR values. However, other exemplary ways are also conceivable for determining second MR individual and MR combination data from first MR data. One example is the exemplary arithmetic representation of the second MR combination data explained above as a symbolically written combination of addition and / or subtraction of first MR data.

[0062] An example of a conceivable determination matrix is the Hadamard matrix, which is described in (Guenther 2012) or (Wolfram 2002), among others.

[0063] For example, when selecting the determination matrices for an original matrix, the matrix complementary to it (complementary matrix), in which marked and unmarked states are interchanged, can also be used. This has the following advantage over the prior art, for example: Usually, only N-1 second MR data can be determined from N first MR data; for example, second MR data indicative of the first fluid bolus cannot be generated. However, using the information from the original matrix and the complementary matrix, second individual MR data indicative of all N fluid boluses can be generated. Furthermore, second MR combination data indicative of combination fluid boluses that would contain the first fluid bolus can also be generated from the first MR data.This can, for example, also allow the generation of second MR combination data from two MR measurements, that of the first row of the original matrix and that of the first row of the complementary matrix, which are indicative of the combination fluid bolus consisting of all N individual fluid boluses. Another conceivable implementation of the determination matrix could be such that only certain rows of an original matrix alternate with the rows of its complementary matrix. The remaining rows of the original matrix would then be non-alternating in the determination matrix and in the order in which they appear in the original matrix. The remaining rows of the complementary matrix would also be located elsewhere in the determination matrix, non-alternating and in the order in which they appear in the complementary matrix.

[0064] An example of ordering based on the number of changes between labeled and unlabeled fluid boluses is so-called sequence sorting. Here, the rows are ordered according to their sequence, and the sequence is defined as the number of sign changes (i.e., zero crossings) of the row's entries. In the case of sequence-sorted Hadamard matrices, this can also be referred to as a Walsh sort (see, for example, (Wolfram 2002)). Using a sequence-sorted determination matrix, for example, the sequence of MR measurements can be rearranged so that the first MR data required to generate second MR combination data are measured as early as possible.One advantage of this compared to the state of the art is, for example, that second MR combination data could be determined before all of the first MR data required to determine the second individual MR data have been measured, and this second combination data is already relevant for clinical diagnostics. For example, it is conceivable that the second MR combination data could be determined in real time during the measurement. If the first MR data are generated in a suitable sequence, second MR combination data could be generated from them during the measurement. This allows, for example, the medical personnel accompanying the measurement to receive initial information at a very early stage of the examination. If necessary, the ongoing measurement can then be modified and / or further diagnostic steps can be planned. Here, too, the robustness of the measurement can be increased.Patient movements, which can render images unusable, typically increase with acquisition time. Therefore, if the subsets of the first MR data that already allow the generation of second MR combination data are generated as early as possible, the probability that all second MR individual and / or MR combination data cannot be generated or can only be generated incorrectly due to patient movement is reduced.

[0065] In other words, it can be said that, for example, it is possible to select the sequence of MR measurements and / or evaluation steps such that the second MR data are determined first, which can be generated from the smallest possible number of first MR data. This number can even be minimal.

[0066] A general example of a Walsh sort is given below.

[0067] Walsh sorting can be performed in three steps: a given sequence S initially written as a binary number (also called dual number): S = S n − 1 S n − 2 … S 1 S 0 2 = ∑ i = 0 n − 1 S i 2 i

[0068] This number in binary code is then converted to Gray code: g i = S i ⊕ S i + 1 , i ∈ 0 , n − 1 , S n ≡ 0

[0069] Where ⊕ stands for the exclusive or.

[0070] By a bit reversal k i = g n − 1 − i are then made from the g i 's the binary values k i The line number corresponding to the sequence k The Hadamard matrix can then be derived from the binary representation k = k n − 1 k n − 2 … k 1 k 0 2 = ∑ i = 0 n − 1 k i 2 i be determined. If the rows k of the Hadamard matrix are arranged in such a way that the sequences S belonging to the k are in ascending or descending order, the matrix is called Walsh-sorted.

[0071] In particular, for N=8 rows the steps result: S 0 1 2 3 4 5 6 7 Binär 000 001 010 011 100 101 110 111 Gray code 000 001 011 010 110 111 101 100 Bit-reverse 000 100 110 010 011 111 101 001 k 0 4 6 2 3 7 5 1

[0072] For example, if quantitative perfusion data are also to be determined (see for example (Buxton et al. 1998)), the evaluation unit can also enable further arithmetic operations for further image processing.

[0073] For the sake of completeness, reference is made to the following literature.

[0074] A process very similar to Guenther 2012 is described in US patent US 8,610,433 B2.

[0075] Guenther (2007) proposes combining individual boluses (labeling areas), called subboluses, into longer, so-called virtual (sub)boluses. This is because the sum of several (sub)boluses can be used to generate any desired longer virtual (sub)bolus, and the interval to image readout, the TI, also called post-labeling delay (PLD), can also be varied. In practice, this involves first reconstructing all reconstructed MR images belonging to the individual (sub)boluses from the measured MR images. These images are then added in a further reconstruction step to form doubly reconstructed MR images, so that the (sub)boluses belonging to the reconstructed MR images form virtual (sub)boluses belonging to the doubly reconstructed MR images.

[0076] This and other aspects are explicitly discussed in (Günther 2007), Dai, W., Shankaranarayanan, A. & Alsop, DC, 2013. Volumetric measurement of perfusion and arterial transit delay using hadamard encoded continuous arterial spin labeling. Magnetic resonance in medicine, 69(4), pp.1014-22 and Teeuwisse, WM et al., 2014. Time-encoded pseudocontinuous arterial spin labeling: Basic properties and timing strategies for human applications. Magnetic resonance in medicine described.

[0077] In the claims, the words "comprising" and "including" do not exclude other elements or steps, and the indefinite article "a" does not exclude a plurality.

[0078] A single unit or device may perform the functions of multiple elements recited in the claims. The fact that individual functions and elements are recited in different dependent claims does not mean that a combination of those functions or elements could not also be advantageously used.

[0079] The control of the MR system and / or the evaluation system according to the MR method or the evaluation method can be implemented as program code of a computer program and / or as corresponding hardware.

[0080] A computer program may be stored and / or distributed on a suitable medium, such as an optical storage medium or a solid-state storage medium, distributed with or as part of other hardware. The computer program may also be distributed in other forms, such as over the Internet or other telecommunications systems.

[0081] The reference signs in the claims are not to be understood as limiting the subject matter and scope of protection of the claims by these reference signs.

Claims

1. System (1, 20) for evaluating diagnostically usable magnetic resonance data (Ki, bi) of a flowing fluid, wherein the system (1, 20) comprises - a data collection unit (5) configured to perform a plurality of magnetic resonance measurements, wherein in each magnetic resonance measurement the data collection unit (5) measures first magnetic resonance data (vi), in that in each magnetic resonance measurement a) at different first points in time in a first region, the fluid is either magnetically labeled or not labeled such that a flowing sequence (10) of labeled (L) and / or unlabeled (C) fluid boli is generated, and b) at a second point in time in a second region, to which the fluid, starting at the first region, flows, the first magnetic resonance data (vi), which are indicative of the generated sequence (10) of labeled and / or unlabeled fluid boli (L, C), are measured, wherein the data collection unit (5) is configured to generate, for each sequence (10) of labeled and / or unlabeled fluid boli (L, C), a complementary sequence of labeled and / or unlabeled fluid boli (L, C), wherein the magnetic resonance measurements of the first magnetic resonance data (vi) of sequences (10) complementary to each other are carried out one immediately after the other, and to provide the first magnetic resonance data (vi) thus measured, or - a data provision unit (21) configured to provide first magnetic resonance data (vi) measured according to a) and b), and - a matrix provision unit (6) for providing a two-valued determination matrix, wherein a row of the determination matrix represents a sequence (10) of labeled and / or unlabeled fluid boli (L, C), and different rows of the determination matrix form the basis for different magnetic resonance measurements of the first magnetic resonance data (vi), wherein the determination matrix is formed by a combination of a Walsh-sorted Hadamard matrix with a complementary Walsh-sorted Hadamard matrix, wherein the resulting matrix alternately comprises the rows of the Walsh-sorted Hadamard matrix and the complementary Walsh-sorted Hadamard matrix, - an evaluation unit (7, 22) configured to, based on a combination of the first magnetic resonance data (vi) measured in the different magnetic resonance measurements, taking into account the determination matrix, i) determine second individual fluid bolus magnetic resonance data (bi; 41-48), each of which are indicative of an individual fluid bolus (L, C) generated at a certain first point in time and flowed to the second region, based on a combination of the first magnetic resonance data (vi) measured in the different magnetic resonance measurements, and ii) determine second combination fluid bolus magnetic resonance data (Ki; 30-40), each of which are indicative of a combination fluid bolus and a time interval (TI) between one of the first points in time and the second point in time, wherein the combination fluid bolus is combined from a plurality of fluid boli (L, C) labeled / unlabeled at different first points in time and flowed to the second region, wherein - the data collection unit (5) is configured to measure the first magnetic resonance data (vi) in an order that allows second combination fluid bolus magnetic resonance data (Ki; 30-40) to be determined in the evaluation unit (7, 22) even before all first magnetic resonance data (vi) required to determine the second individual fluid bolus magnetic resonance data (bi; 41-48) have been measured, characterized in that the Walsh-sorted Hadamard matrix and the complementary Walsh-sorted Hadamard matrix each are of order 2k with k ≥ 2, and - the evaluation unit (7, 22) is configured to evaluate the first magnetic resonance data (vi) in an order that allows second combination fluid bolus magnetic resonance data (Ki; 30-40) to be determined even before all first magnetic resonance data (vi) required to determine the second individual fluid bolus magnetic resonance data (bi; 41-48) have been measured, wherein fewer first magnetic resonance data are used to determine second magnetic resonance data for a combination fluid bolus than are used to determine second magnetic resonance data for an individual fluid bolus.

2. System (1, 20) according to claim 1, characterized in that the evaluation unit (7, 22) is configured to evaluate the first magnetic resonance data (vi) in an order in which the number of changes between labeled and unlabeled fluid boli (L, C) in the fluid boli sequences (10) belonging to the respective first magnetic resonance data entering into the evaluation increases.

3. System (1, 20) according to claim 1 or 2, characterized in that the rows of the Walsh-sorted Hadamard matrix represent a first group of sequences (10) of labeled / unlabeled fluid boli (L, C), wherein different rows of the Walsh-sorted Hadamard matrix form the basis for different magnetic resonance measurements of the first magnetic resonance data (vi), and the rows of the complementary Walsh-sorted Hadamard matrix represent a second group of sequences (10) of labeled / unlabeled fluid boli (C, L), which are complementary to the sequences of fluid boli of the first group, wherein different rows of the complementary Walsh-sorted Hadamard matrix form the basis for different magnetic resonance measurements of the first magnetic resonance data (vi), wherein the evaluation unit (7, 22) is further configured - to form a first vector, wherein the vector elements of the first vector are each formed by a magnetic resonance value of the first magnetic resonance data (vi), and different vector elements of the first vector correspond to different magnetic resonance measurements of the first magnetic resonance data (vi), which in turn correspond to different fluid boli sequences of the first group, - to form a second vector, wherein a vector element of the second vector is formed by a magnetic resonance value of the first magnetic resonance data (vi), and different vector elements of the second vector correspond to different magnetic resonance measurements of the first magnetic resonance data (vi), which in turn correspond to different fluid boli sequences of the second group, - to invert the Walsh-sorted Hadamard matrix and the complementary Walsh-sorted Hadamard matrix, and - to determine the second magnetic resonance data by applying the inverted Walsh-sorted Hadamard matrix to the first vector and the inverted complementary Walsh-sorted Hadamard matrix to the second vector.

4. Method for evaluating diagnostically usable magnetic resonance data of a flowing fluid, wherein the method comprises: - collecting and providing first magnetic resonance data (vi), wherein a plurality of magnetic resonance measurements are carried out, wherein in each magnetic resonance measurement first magnetic resonance data (vi) are measured, in that in each magnetic resonance measurement a) at different first points in time in a first region of the fluid, the fluid is either magnetically labeled or not labeled such that a flowing sequence (10) of labeled (L) and / or unlabeled (C) fluid boli is generated, and b) at a second point in time in a second region, to which the fluid, starting at the first region, flows, the first magnetic resonance data (vi), which are indicative of the generated sequence (10) of labeled and / or unlabeled fluid boli (L, C), are measured, wherein for each sequence (10) of labeled and / or unlabeled fluid boli (L, C), a complementary sequence of labeled and / or unlabeled fluid boli (L, C) is generated, wherein the magnetic resonance measurements of the first magnetic resonance data (vi) of sequences (10) complementary to each other are carried out one immediately after the other, or - providing first magnetic resonance data (vi) measured according to a) and b), and - providing a two-valued determination matrix, wherein a row of the determination matrix represents a sequence (10) of labeled and / or unlabeled fluid boli (L, C), and different rows of the determination matrix form the basis for different magnetic resonance measurements of the first magnetic resonance data (vi), wherein the determination matrix is formed by a combination of a Walsh-sorted Hadamard matrix with a complementary Walsh-sorted Hadamard matrix, wherein the resulting matrix alternately comprises the rows of the Walsh-sorted Hadamard matrix and the complementary Walsh-sorted Hadamard matrix, - evaluating the provided first magnetic resonance data (vi), taking into account the determination matrix, wherein the evaluating comprises, based on a combination of the first magnetic resonance data (vi) measured in the different magnetic resonance measurements i) to determine second individual fluid bolus magnetic resonance data (bi; 41-48), each of which are indicative of an individual fluid bolus (L, C) generated at a certain first point in time and flowed to the second region, and ii) to determine second combination fluid bolus magnetic resonance data (Ki; 30-40), each of which are indicative of a combination fluid bolus and a time interval (TI) between one of the first points in time and the second point in time, wherein the combination fluid bolus is combined from a plurality of fluid boli (L, C) labeled / unlabeled at different first points in time and flowed to the second region, wherein - the first magnetic resonance data (vi) are collected in an order in which, even before all first magnetic resonance data (vi) required to determine the second individual fluid bolus magnetic resonance data (bi; 41-48) have been measured, second combination fluid bolus magnetic resonance data (Ki; 30-40) can be determined, characterized in that the Walsh-sorted Hadamard matrix and the complementary Walsh-sorted Hadamard matrix each are of order 2k with k ≥ 2, and - the first magnetic resonance data (vi) are evaluated in an order that allows second combination fluid bolus magnetic resonance data (Ki; 30-40) to be determined even before all first magnetic resonance data (vi) required to determine the second individual fluid bolus magnetic resonance data (bi; 41-48) have been measured, wherein fewer first magnetic resonance data are used to determine second magnetic resonance data for a combination fluid bolus than are used to determine second magnetic resonance data for an individual fluid bolus.

5. Method according to claim 4, wherein the first magnetic resonance data (vi) are evaluated in an order in which the number of changes between labeled and unlabeled fluid boli (L, C) in the fluid boli sequences (10) belonging to the respective first magnetic resonance data entering into the evaluation increases.

6. Computer program for evaluating diagnostically usable magnetic resonance data (Ki, bi) of a flowing fluid, characterized in that the computer program comprises program code means which cause the system according to claim 1 to carry out the method according to claim 4 when the computer program is executed on a computer which controls the system.