Validation of the quantitative magnetic resonance imaging protocol

Abstract

A validation technique for quality assurance of quantitative MRI methods compares the measured magnetic properties of a phantom over a range of T1 and T2 values ​​measured by a clinically feasible acceleration protocol with predicted values ​​for that magnetic properties calculated from a set of reference T1 and T2 values ​​measured on the phantom. The prediction is based on the relationship between the values ​​from the acceleration protocol and the values ​​from reference measurements obtained by repeated scanning of one or more phantoms.

Description

Technical Field

[0001] The present invention relates to a method of validating or quality assurance of a protocol for quantitative magnetic resonance imaging. Background Art

[0002] Magnetic resonance (MR) imaging (MRI) has become ubiquitous as a medical imaging technique used to generate images of patient anatomy and physiological processes within both healthy and diseased patients. Essentially, MR images are two-dimensional displays of contrast caused by various magnetic differences in patient tissue. Many different types of sequences (i.e., different combinations of static and oscillating magnetic fields and RF pulses) are well known and are typically preprogrammed into MRI scanners to allow radiographers to select the optimal sequence to detect tissue of interest or tissue abnormalities in a specific patient. These traditional sequences weight multiple MR properties in a nonlinear manner and depend on the distance from the coil used to acquire the signal. In weighted MRI, differences in the image plane are of primary interest, as these differences can be displayed and allow trained observers to visually identify tissue differences. Because the goal is to display images that distinguish one tissue from another (e.g., abnormal tissue from normal tissue), MRI is used as a qualitative technique. Traditionally, the absolute (i.e., quantitative) values ​​of the magnetic properties associated with pixels in MRI images have been of limited interest.

[0003] Further MRI sequences are known to directly measure the specific underlying magnetic properties of a substance, such as spin-lattice relaxation time T1, spin-spin relaxation time T2, proton density (PD), and so on. Consequently, interest has arisen in so-called quantitative MRI, which aims to characterize different tissue types or abnormalities by measuring magnetic properties, such as the tissue's actual T1 or T2 relaxation time or PD. However, a key issue with this approach is that the measured magnetic properties are sensitively dependent on various factors, including the MRI scanner, coil type, setup, signal reconstruction, and calibration, and are therefore not consistent between scans or scanners. Tissue surrogates, known as phantoms, are typically used to study magnetic properties. Even the most complex phantoms cannot fully simulate the complexity of living tissue. Living tissue consists of a multi-layered variability of molecules, organelles, cells, and organs, whose components exhibit a range of magnetic properties. Over any practical measurement time, measurements of MR properties are subject to dynamic exchange (through the exchange of spin properties at the molecular level and through bulk flow and diffusion effects). Ultimately, due to these factors, no single MR property can accurately describe the complex mixture of compartments in any biological system. Therefore, any chosen quantification method describes a weighted estimate of its primary objective and has been shown to be dependent on the choice of the specific sequence used during imaging and its individual settings. Therefore, although quantitative MRI scans will give absolute quantitative values ​​for various properties of tissue, these values ​​cannot be compared between different scans unless it is known that the protocol used in the scans is exactly the same.

[0004] To obtain clear MR images, the patient must remain still during the MRI sequence. The duration of MRI sequences varies widely, and generally speaking, the longer the duration, the more difficult it is for the patient to remain still. Furthermore, in some applications (e.g., cardiac MRI), the patient's heartbeat and respiration significantly limit the clinically acceptable duration of the sequence. Specific accelerated sequences have been developed that are shorter in duration than reference sequences and can therefore be performed in a single breath hold of the patient, for example, for cardiac T1 mapping. These include MOLLI, ShMOLLI, SASHASAPPHIRE, and others (see, for example, Piechnik SK and Jerosch-Herold M., “Myocardial T1 mapping and extracellular volume quantification: an overview of technical and biological confounders,” Int J Cardiovasc. Imaging. 2018 Jan;34(1):3-14. doi:10.1007 / s 10554-017-1235-7. Epub 2017 Aug 28, and Andrew J. Taylor MD, PhD, Michael Salerno MD, PhD, Rohan Dharmakumar, PhD, and Michael Jerosch-Herold MD, PhD, “T1 Mapping: Basic Techniques and Clinical Applications,” JACC: Cardiovascular Imaging. Imaging, Vol. 9, No. 1, January 2016, pp. 67–81).

[0005] These can be referred to as "clinically feasible" or accelerated MRI sequences. However, the magnetic values ​​measured using such clinically feasible sequences often differ from each other and also from the values ​​measured by so-called "reference" sequences of longer duration. This, in turn, presents a significant difficulty when using clinically feasible MRI sequences in quantitative MRI, as these values ​​do not necessarily allow for quantitative comparisons between different scans and scanners to obtain consistent tissue characterizations using these values.

[0006] Recently, a different approach has been proposed, which is called Magnetic Resonance Fingerprinting (MRF). This involves pseudo-randomizing multi-parameter acquisition so that signals from different materials or tissues have unique signal evolutions over time (or "fingerprints" that are simultaneously a function of the multiple material properties being studied). Post-acquisition processing involves pattern recognition algorithms that match the fingerprints with predefined dictionaries that predict the signal evolution. See, for example, MAD., Gulani V., Seiberlich N. et al., "Magnetic Resonance Fingerprinting," Nature, 2013;495:187-192. However, this in itself does not guarantee that quantitative values ​​measured at different scanners or centers are comparable.

[0007] If one can ensure that MR scans follow precisely the same protocol, it will be possible to compare the values ​​measured in different scans and thus characterize the tissue through these values. Summary of the Invention

[0008] Therefore, the object of the present invention is to provide a technique for ensuring the quality of MRI scans by validating the protocol in use. This allows greater confidence that magnetic measurements can be relied upon for tissue characterization for comparison between different centers or over time. The term "sequence" is generally used to refer to a combination of static and oscillating magnetic fields and RF pulses in order to obtain interpretable MRI signals, and the term "protocol" refers to a collection of one or more sequences with associated setup requirements and possible further steps (e.g., pauses, setting of external conditions, e.g., time of artificial triggering or contrast agent injection). An MR experiment is a process performed according to a protocol and comprising one or more sequences.

[0009] According to one aspect of the present invention, a method for verifying a clinically feasible MRI protocol is provided, comprising the following steps: performing at least one first clinically feasible MRI experiment on a phantom according to the clinically feasible MRI protocol to be verified to measure a first magnetism of the phantom; performing a second, different MRI experiment on the phantom to measure a second magnetism of the phantom; performing a third MRI experiment on the phantom, the third MRI experiment being different from the first and second MRI experiments, to measure a third magnetism of the phantom that is different from the first and second magnetisms; calculating a predicted value of the first magnetism from the measured second and third magnetisms based on a predetermined relationship between the first, second and third magnetisms; calculating a difference between the predicted value of the first magnetism and the measured value of the first magnetism; determining that the MRI protocol used for the first MRI experiment is valid if the difference between the predicted value of the first magnetism and the measured value of the first magnetism is lower than a predetermined threshold; and outputting the determined result.

[0010] The predetermined relationship can be a nonlinear, multi-parameter relationship, for example, obtained by performing the first magnetic field measurement, the second magnetic field measurement, and the third magnetic field measurement multiple times on the same phantom or multiple phantoms. The multiple phantoms can have different compositions. The phantoms can include multiple materials with different magnetic properties to cover a range of interest, for example, T1 and T2 values ​​from 50 to 3500 ms.

[0011] The predetermined relationship can be obtained by performing measurements of the first, second, and third magnetic properties at two or more different magnetic field strengths. This is particularly useful if, in practice, an MRI scan will be performed under different field settings. For example, cardiac MRI is typically performed at 1.5 T and 3.0 T. There may be a relationship that is independent of the magnetic field, an individual relationship established for an available choice of magnetic field, or a relationship (formula) in which the magnetic field is a parameter.

[0012] A predetermined relationship (which may be referred to as an "MR model") can be obtained by performing measurements of the first, second, and third magnetic quantities at multiple different parameter values. Parameter values ​​refer to the variables (if any) that affect the relationship between the measured magnetic quantities. For some magnetic quantities, such as ShMOLLI T1 (hereinafter referred to as T1sh), the MR model depends primarily on T1 and T2, with slight differences between 1.5T and 3T. However, for other clinically feasible sequences, or when higher-precision quality measurements are required, the model can be equipped with additional MR parameter dependencies (e.g., magnetization transfer, off-frequency characteristics, T2*, T1ρ, etc.) and experimental correlations (e.g., heart rate or temperature at the time of measurement). The desired MR model can be determined by performing scans on a phantom at various parameter values. For example, it is well known that the relationship between MOLLI T1 and T1 depends on heart rate, which requires modification of the model and the number of measurements to establish and validate the model. Therefore, further parameter values ​​may include at least one of temperature, magnetic transfer ratio, coil sensitivity, heart rate, MR frequency adjustment, etc.

[0013] The first and second magnetic properties can be the same, for example, the spin-lattice relaxation time T1, but measured using two different methods—at least one clinically feasible or accelerated sequence, and the other a set of other, typically longer-duration reference experiments, such as repeated turbo spin echo (TSE) acquisitions with varying inversion times. The third magnetic property can be the spin-spin relaxation time T2, measured using multiple TSE sequences with varying TEs, or a single multi-echo acquisition for increased speed. Typically, the reference sequence will be a simple, standard, long-duration MR routine technique, for example, to allow for transparent and easy transfer between different systems and manufacturers. The second and third experiments can also be performed using a single hybrid sequence, for example, for MR fingerprinting. Furthermore, the first experiment can be performed within such a hybrid sequence.

[0014] The present invention also provides a method for measuring a first magnetic property of human or animal tissue using an MRI protocol validated according to the above method. This allows for quantitative MRI measurements, wherein there is greater confidence that the measured values ​​are comparable to those measured at other locations or scan times, and therefore on a consistent basis. This allows the present invention to be extended to procedures for characterizing tissue (e.g., normal or abnormal) and abnormal properties based on a normal range of measured first magnetic properties established globally, rather than at each individual center.

[0015] According to other aspects of the present invention, there are provided a computer program executable by a computer device and configured to cause the computer device to perform a similar method when executed, a computer-readable storage medium storing such a computer program, and a computer device configured to perform a similar method. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] The present invention will be further described by way of example with reference to the accompanying drawings, in which:

[0017] Figure 1(a) to Figure 1(d) are different views of an example of a phantom used in a method according to an embodiment of the present invention;

[0018] Figure 2 is a flowchart illustrating a quality assurance technique according to a first embodiment of the present invention;

[0019] FIG3( a ) and FIG3 ( b ) show the direct relationship between the spin-lattice T1 relaxation times of the phantom of FIG1 at two different magnetic field strengths measured by a reference MRI experiment and a clinically feasible MRI sequence;

[0020] 4( a ) and 4 ( b ) illustrate how the deviation observed from the linear relationship between the quantities shown in FIG. 3 ( a ) and FIG. 3 ( b ) depends on the spin-spin relaxation time T 2 of the phantom of FIG. 1 ;

[0021] Figure 5 shows a protocol for quality assurance of T1 mapping;

[0022] Figure 6 is a flowchart illustrating a quality assurance technique according to a second embodiment of the present invention;

[0023] Figures 7(a) and 7(b) show the potential heart rate dependence curves that appear between different clinically feasible T1 mapping sequences;

[0024] Figure 8 is a flow chart illustrating an example of a technique for deriving a relationship between magnetic properties that are further measured in magnetic resonance imaging;

[0025] Figure 9is a flow chart illustrating another example for deriving a relationship between magnetic properties measured in magnetic resonance imaging;

[0026] FIG10( a ) shows the residual error and confidence range of the predicted clinically feasible T1 value based on the T2-dependent model established in FIG3( a ), FIG3( b ), FIG4( a ), and FIG4( b );

[0027] FIG10( b ) shows an additional empirical polynomial correction (shown as a dashed line) for the trends identified in the MR model in FIG10( a ); and

[0028] FIG. 11 shows an example of quality assurance of a T1 mapping sequence using the phantom of FIG. 1 . DETAILED DESCRIPTION

[0029] While one could argue that one way to effectively calibrate the magnetic values ​​measured by an MRI scan is to measure the magnetic properties of a material of known magnetic properties (called a "phantom") and then apply some calibration factor, it has been impossible to find a material whose magnetic properties are sufficiently stable and consistent over the desired range of values ​​to achieve this in a cost-effective manner.

[0030] The inventors have found that if clinically feasible (i.e., accelerated) MRI protocols are properly followed, normal values ​​are the same between various scanners and potentially allow direct comparison of pathological deviations. From the perspective of MR theory, there should be a consistent relationship between the magnetic properties measured by clinically feasible sequences and the magnetic properties measured by reference sequences. This relationship should have a universal form, that is, independent of the exact specification of the property being measured. However, although many simulations using the Bloch equations that govern MR physics can be used to guide the understanding of this dependence, there is no guarantee that these simulations will fully describe this relationship because many unknown factors need to be assumed.

[0031] The inventors have found that, alternatively (though nevertheless with any theoretical background, if possible), it is possible to establish this relationship experimentally. This means that MRI scans can be performed on relatively simple phantoms, regardless of the history or properties of the phantom, and it can be determined whether a clinically feasible MRI protocol has been followed by comparing magnetic measurements made using that protocol with reference measurements made on a similar phantom. The phantoms do not need to be identical, just need to span the range of reasonably similar properties required to establish and validate MR models. Therefore, the quality of MRI results from a particular scanning center can be assured by requiring the scanning center to regularly perform clinically feasible MRI measurements and reference measurements on the phantom and compare the results to check whether they have the required relationship.

[0032] Figure 1(a) to Figure 1(d)An example of a phantom suitable for the validation and quality assurance techniques of the present invention, including each of the methods described below, is shown. The phantom has nine compartments 50, each consisting of a clinical sample container filled with an aqueous gel using agar and carrageenan doped with sodium chloride to achieve the desired T1 and T2 combination in the range of 50-3500 milliseconds. Figure 1(a) shows a single container and Figure 1(b) shows nine containers stacked in a 3×3 arrangement in a PCV container 52. As shown in Figure 1(d), this is in turn inserted into a tight-fitting cardboard box 54 to provide some protection against damage and improve internal thermal uniformity. Figure 1(c) schematically shows the containers labeled A to I and the characteristics of the contents of the containers are shown in Table 1 below, where the T1 and T2 combinations are achieved.

[0033] Table 1:

[0034]

[0035]

[0036]

[0037] Figure 2 A first embodiment of a verification or quality assurance technique based on this concept is schematically shown. In step 101 , a clinically feasible MRI experiment is performed on a phantom according to a clinically feasible protocol to measure a first magnetic field.

[0038] An example of such a clinically feasible MRI experiment is ShMOLLI T1 mapping or MOLLI T1 mapping. This experiment is performed on a phantom as shown in FIG1 , which includes compartments containing substances with various T1 and T2 values, which are selected to span the range of interest of the MRI scan. In step 102 , a second MRI experiment (consisting of one or more sequences, with appropriate post-processing) is performed on the phantom to measure a second magnetic field. The second magnetic experiment can be of longer duration, for example, a slice-selective inversion recovery spin echo sequence, to provide a reference T1 value for the phantom. As described above, although both the first and second experiments measure T1 values, the T1 values ​​measured by a clinically feasible accelerated sequence are typically different from the T1 values ​​measured by a longer duration reference sequence. Compared to accelerated clinical sequences, which often have customized proprietary features, reference sequences are considered more standardized and therefore more easily transferable between various devices. FIG3( a) and FIG3( b) illustrate the difference in T1 values ​​(i.e., the difference between T1sh and T1ref in this example) at two different static field strengths, 1.5T and 3T. The expected linear relationship shows deviations associated with additional attributes (most notable indicated by arrows).

[0039] In step 103 , a third MRI experiment is performed on the phantom to measure a third magnetic field, which may be, for example, a sequence for measuring the spin-spin relaxation time T2, such as a multi-echo SE sequence.

[0040] Figures 4(a) and 4(b) show how the deviation between T1 values ​​measured by a clinically feasible sequence and a reference experiment depends on T2 at the two different static field strengths used. Note that these relationships are relatively similar and, in some cases, can be considered identical and pooled between magnetic field settings, as long as this addresses the tolerance range required for sequence identification in clinical applications.

[0041] Then, in step 104, a predicted value of the first magnetic property can be calculated from the measured values ​​of the second and third magnetic properties using a predetermined relationship between the three. If a clinically feasible protocol has been implemented and correctly followed, there should be little difference between the predicted value of the first magnetic property and its measured value. Therefore, in step 105, the difference between the measured value and the predicted value of the first magnetic property is calculated, and in step 106, the difference is compared with a predetermined confidence interval to determine whether the clinically feasible MRI protocol has been correctly followed. That is, if the difference is below a predetermined threshold, it is valid, otherwise it is invalid.

[0042] In step 107, the determined validity is output and can be recorded to demonstrate the current performance of the scanner. Here and in corresponding steps of the following method, the output can be data representing the determined validity. The output can be displayed on a display.

[0043] As now described Figure 5 As shown, further confidence can be gained by taking multiple measurements of any measured magnetic property.

[0044] Figure 5 An example of a scanning experiment list for the first embodiment is shown. Steps 7 to 15 perform repeated ShMOLLI T1 mapping measurements on the phantom, steps 19 to 30 are used to reconstruct reference T1 values, and steps 17, 18, or 31 are used to reconstruct reference T2 values. If the protocol is adjusted to allow for variations in other parameters, steps 7 to 15 or the scanning process will be repeated using different values ​​of these parameters, depending on which magnetic properties are affected, for example, at different ambient temperatures throughout the scanning process or at different heart rate settings in steps 7 to 15.

[0045] Figure 6A second embodiment of the present invention is shown, which is similar to the first embodiment, except that in the first embodiment, it is assumed that the value of the first magnetic field can be predicted based solely on the second and third magnetic fields. However, for some clinically feasible MRI sequences, additional parameters may influence this relationship. For example, in the case of the MOLLI T1 mapping sequence, the difference between the MOLLI T1 value and the reference T1 value has a heart rate dependency, as shown in Figures 7(a) and 7(b). This is shown in Figure 7(a), which shows the lack of heart rate dependency for ShMOLLI, compared to the clear heart rate dependency of MOLLI shown in Figure 7(b). In other clinically feasible sequences, different parameters or other additional parameters may influence this relationship. Therefore, modeling and testing for the presence of such a pattern may require further parameters to fully identify these techniques. The second embodiment accounts for this multi-parameter dependency by repeatedly performing the steps of a clinically feasible MRI sequence on a phantom, each time using different parameter settings. For example, magnetization transfer can be substituted or used in addition to T2 measurement. If heart rate dependence is present, clinically feasible MRI sequences can be performed at different heart rate settings by setting up an artificial electrocardiogram to represent various rhythms or frequencies.

[0046] Some settings and physical properties, while affecting the magnetic values ​​of the phantom, do not affect the relationship. In the first embodiment, this has been confirmed for the temperature and age of the phantom. Unless required for other purposes (e.g., quality control of the phantom state, aging, etc.), there is no need to measure these parameters excluded from the relationship. An example is the ambient temperature, which affects the measurement of T1sh and T1ref in the ShMOLLIT1 validation, but the relationship model between them does not show temperature dependence. The same is true for moderate deviations in the phantom properties, such as age or physical damage, as long as they do not severely limit the range of parameter values ​​used to validate the MR model.

[0047] In steps 202 and 203, as in the first embodiment, the second and third magnetic quantities are measured experimentally by reference MRI, and in step 204, these values ​​are used together with the different parameter values ​​used in step 201 to predict the value of the first magnetic quantity. In step 205, this is compared with the measured value, and in step 206, the validity of the protocol for a clinically feasible MRI sequence is determined based on the difference between the predicted and measured values ​​of the first magnetic quantity, i.e., if the difference is below a predetermined threshold, the protocol is valid, otherwise the protocol is invalid. In step 207, the determined validity is output and can again be recorded as the basis for certification of the scanning protocol.

[0048] Figure 8A method for obtaining the desired predetermined relationship between the first, second, and third magnetic properties is shown. In step 301, a clinically feasible MRI experiment is performed on a phantom such as that shown in FIG1 to obtain a measurement value of the first magnetic property. In steps 302 and 303, a second and third MRI experiment are performed on the same phantom to measure the second and third magnetic properties. These three steps are then repeated at multiple times to exploit natural changes in the phantom properties due to aging and changes in ambient temperature, optionally on multiple different examples of the phantom. The number of repetitions can potentially be reduced by using a more complex phantom. Then, in step 305, the relationship between the three quantities is empirically derived using any pre-existing MR theory, publications, or dedicated simulations and empirical data fitting (least squares and robust methods). This relationship effectively constitutes a fingerprint prediction MR model that can be used to predict the first magnetic quantity given the second and third measurements and any related parameters.

[0049] Figure 9 A corresponding process is shown, wherein additional parameters are taken into account by repeating steps 401 , 402 and 403 with different parameter settings. Figure 9 Steps 401 to 405 additionally correspond to Figure 8 Steps 1 to 3 of the present invention are not discussed in detail. In step 404, other parameters that affect the relationship between the three magnetic properties are changed. For example, in the case where the first magnetic property is MOLLI T1, there is a heart rate dependency. Figure 9 The process is modeled by varying the artificial ECG signal during the measurement of the phantom in step 401. It will be appreciated that the technique is extendible to models containing more than three magnetic fields, in which case additional experiments similar to step 403 are added to measure the additional magnetic fields.

[0050] An example of a relationship or prediction model in the case of ShMOLLI T1 mapping sequences for two different magnetic field strengths (1.5T and 3T, which are the magnetic field strengths typically used in clinical practice) is given below as an example of a clinically feasible experiment. It can be seen that T1sh generally follows a linear relationship with T1ref, with a certain known dependence on T2 as shown in Figures 4(a) and 4(b) that is fitted as an exponential relationship. Correction for this gives the residual error shown in Figure 10(a), which includes a small visible trend that can be further compensated using a third-order polynomial (black dashed line indicated by the arrow in Figure 10(a)) to achieve a unit correlation between T1sh and T1ref. In this case, the final model for predicting the expected T1sh based on the reference T1 and T2 measurements is:

[0051]

[0052] Thus, this model is an example of a model used in step 104 or 204 to calculate a predicted value of the first magnetic property (T1sh in this particular example), which can be compared with the measured value of step 101 or 201. Depending on whether the measured value is within the appropriate confidence interval, it can be considered valid or invalid. In Figure 10(b), a 95% confidence interval and a 99.7% confidence interval and one standard deviation are shown.

[0053] Figure 11 shows the results of an example of a quality assurance program conducted on multiple participating MRI sites. In this case, multiple sites were required to Figure 2 The first embodiment of the method is validated, wherein the clinically feasible MRI experiment is ShMOLLI T1 mapping, and the second and third MRI experiments are measurements of T1 reference and T2 reference as discussed above. These sites are measured on the phantom, such as Figure 5 A protocol (shown in Figure 11) was provided for these sites to perform five replicates of ShMOLLI sequences, IR-SE experiments, and multi-echo SE experiments with the following specifications.

[0054] 1) Repeated ShMOLLI T1

[10] sequence: echo time (TE) = 1.07 ms; repetition time (TR) = 3.57 ms; number of inversions (TI) = 100, 1100, 2100, 3100, 4100, 180, 260 ms; flip angle (FA) 35°; FOV = 270x360 mm; matrix size 384x288; slice thickness 8 mm; body matrix coil; GRAPPAx2.

[0055] 2) Slice-selective IR experiments with turbo spin echo readout with a turbo factor of 7 to provide reference T1 relaxation times: TE = 11 ms; TR = 10,000 ms; TI = 33, 100, 300, 900, 2,700, and 5,000 ms; FOV = 360 x 360 mm; matrix size = 256 x 256; slice thickness 8 mm; body matrix coil.

[0056] 3) Multi-echo SE experiment, providing reference T2 relaxation time: TE = 15-480 ms every 15 ms; TR = 9000 ms; FOV = 360x360 mm; matrix size 256x256; slice thickness 8 mm; body matrix coil.

[0057] The 78 quality assurance scans out of the 28 scans are categorized in Table 2 below.

[0058] Table 2:

[0059] QA results from 78 scans at 28 sites with recommended findings and actions.

[0060]

[0061]

[0062] Each step of the method can be implemented using the following computer equipment.

[0063] Steps 104-107 of Figure 1, Figure 6 Steps 204-207 Figure 8 Step 305 and Figure 9 Step 405 can be performed using a computer device. In this case, the results of the MRI experiments performed in other steps can be input to the computer device as data. Alternatively, the computer device can also control the MRI device to perform other steps.

[0064] In order to achieve this, a computer program that can be executed by a computer device may be provided. The computer program is configured such that when executed, the computer device is caused to perform the relevant steps of the method.

[0065] The computer device, when used, may be any type of computer system, but will generally be of conventional construction. A computer program may be written in any suitable programming language. The computer program may be stored on a computer-readable storage medium, which may be of any type, such as a recording medium that can be inserted into a drive of a computer system and that can store information magnetically, optically, or magneto-optically; a fixed recording medium of a computer system, such as a hard drive; or a computer memory.

Claims

1. A method for validating a clinically feasible MRI protocol, comprising the following steps: performing a clinically feasible first MRI experiment on the phantom according to a clinically feasible MRI protocol to be verified, to measure a first magnetic property of the phantom; performing a different second MRI experiment on the phantom to measure a second magnetic property of the phantom; performing a third MRI experiment on the phantom to measure a third magnetic property of the phantom that is different from the first magnetic property and the second magnetic property, wherein the third MRI experiment is different from the first MRI experiment and the second MRI experiment; calculating a predicted value of the first magnetism from the measured second magnetism and the measured third magnetism based on a predetermined relationship among the first magnetism, the second magnetism, and the third magnetism, calculating a difference between a predicted value of the first magnetic property and a measured value of the first magnetic property; determining that the MRI protocol for the first MRI experiment is valid if the difference between the predicted value of the first magnetic property and the measured value of the first magnetic property is below a predetermined threshold; and A result of the determination is output.

2. The method according to claim 1, wherein The predetermined relationship is a nonlinear multi-parameter relationship.

3. The method according to claim 1, wherein The predetermined relationship is obtained by performing the measurement of the first magnetism, the measurement of the second magnetism, and the measurement of the third magnetism a plurality of times.

4. The method according to claim 3, wherein: The predetermined relationship is obtained by performing one or more measurements of the first magnetic properties, the second magnetic properties, and the third magnetic properties on each of a plurality of phantoms.

5. The method according to claim 4, wherein The plurality of phantoms differ in at least one of composition and age.

6. The method according to any one of claims 3 to 5, wherein The predetermined relationship is obtained by performing the measurement of the first magnetism, the measurement of the second magnetism, and the measurement of the third magnetism at two or more different magnetic field strengths.

7. The method according to any one of claims 3 to 5, wherein The predetermined relationship is obtained by performing measurements of the first magnetism, the second magnetism, and the third magnetism at a plurality of different parameter values, wherein the parameter value is a variable that affects the relationship between the measured magnetic quantities.

8. The method according to claim 7, wherein: The plurality of different parameter values ​​include at least one of the following: T2*, diffusion coefficient, temperature, magnetic transmission ratio, coil sensitivity, and patient heart rate.

9. The method according to any one of claims 1 to 5, wherein The first magnetic property and the second magnetic property are the same.

10. The method according to claim 9, wherein: The first magnetism and the second magnetism are spin-lattice relaxation time T1.

11. The method according to claim 10, wherein: The third magnetic property is the spin-spin relaxation time T2.

12. The method according to any one of claims 1 to 5, wherein The second MRI experiment and the third MRI experiment are performed in a hybrid multi-parameter acquisition to obtain any or all reference magnetic properties.

13. The method according to any one of claims 1 to 5, wherein The first MRI experiment is performed in a hybrid multiparameter acquisition to obtain a set consisting of several first magnetic fields acquired in one go.

14. A method of measuring a first magnetic property of human or animal tissue, comprising using an MRI protocol validated by the method according to any one of claims 1 to 10. The method of claim 14 , further comprising characterizing the tissue based on the measured first magnetic properties.

16. A computer program product, executable by a computer device and configured to, when executed, cause the computer device to perform a method for validating a clinically feasible MRI protocol, the method comprising: receiving a first magnetic property of the phantom measured by performing a first clinically feasible MRI experiment on the phantom according to a clinically feasible MRI protocol to be validated; receiving a second magnetic property of the phantom measured by performing a different second MRI experiment on the phantom; and receiving a third magnetic property of the phantom that is different from the first magnetic property and the second magnetic property, measured by performing a third MRI experiment on the phantom, the third MRI experiment being different from the first MRI experiment and the second MRI experiment; calculating a predicted value of the first magnetism from the measured second magnetism and the measured third magnetism based on a predetermined relationship among the first magnetism, the second magnetism, and the third magnetism, calculating a difference between a predicted value of the first magnetic property and a measured value of the first magnetic property; determining that the MRI protocol for the first MRI experiment is valid if the difference between the predicted value of the first magnetic property and the measured value of the first magnetic property is below a predetermined threshold; and A result of the determination is output.

17. A computer-readable storage medium storing the computer program product according to claim 16.

18. A computer device configured to perform a method for validating a clinically feasible MRI protocol, the method comprising: receiving a first magnetic property of the phantom measured by performing a first clinically feasible MRI experiment on the phantom according to a clinically feasible MRI protocol to be validated; receiving a second magnetic property of the phantom measured by performing a different second MRI experiment on the phantom; and receiving a third magnetic property of the phantom that is different from the first magnetic property and the second magnetic property, measured by performing a third MRI experiment on the phantom, the third MRI experiment being different from the first MRI experiment and the second MRI experiment; calculating a predicted value of the first magnetism from the measured second magnetism and the measured third magnetism based on a predetermined relationship among the first magnetism, the second magnetism, and the third magnetism, calculating a difference between a predicted value of the first magnetic property and a measured value of the first magnetic property; determining that the MRI protocol for the first MRI experiment is valid if the difference between the predicted value of the first magnetic property and the measured value of the first magnetic property is below a predetermined threshold; and A result of the determination is output.

Images