Magnetic resonance imaging and methods for reducing image distortions caused by low-frequency magnetic fields
Patent Information
- Application Number
- DE502022004487
- Authority / Receiving Office
- DE · DE
- Patent Type
- Patents
- Current Assignee / Owner
- Filing Date
- 2022-11-11
- Publication Date
- 2025-07-31
- Estimated Expiration
- 2042-11-11
AI Technical Summary
Magnetic resonance imaging (MRI) devices are sensitive to variations in static magnetic fields due to limited bandwidth of receiving systems, making it difficult to distinguish between protons in water and fat, and are susceptible to interference from external low-frequency magnetic fields.
Incorporation of magnetic field sensors outside the MRI scanner to detect and compensate for external interfering magnetic fields by adjusting the controller to generate counteracting magnetic fields or shift the center frequency of the MRI signal, using methods such as variable-frequency oscillators and gradient coils.
Effectively compensates for external magnetic interference, improving image quality by reducing distortions caused by low-frequency fields without additional energy expenditure and maintaining image reconstruction integrity.
Description
[0001] Regardless of the grammatical gender of a particular term, persons with male, female or other gender identities are included.
[0002] The invention relates to a magnetic resonance imaging device with a magnetic field sensor and a method for reducing interference caused by external magnetic fields.
[0003] Magnetic resonance imaging scanners are imaging devices that, to create images of a subject, align the nuclear spins of the subject with a strong external magnetic field and then excite them to precess around this alignment using an alternating magnetic field. The precession, or return, of the spins from this excited state to a lower-energy state, in turn generates a response alternating magnetic field, which is received via antennas.
[0004] Using magnetic gradient fields, a spatial coding is applied to the signals, which subsequently allows the received signal to be assigned to a volume element. The received signal is then evaluated, and a three-dimensional image of the object under examination is provided. Local receiving antennas, so-called local coils, are preferably used to receive the signal. These antennas are positioned directly on the object under examination to achieve a better signal-to-noise ratio.
[0005] Despite the high magnetic fields, imaging with magnetic resonance imaging is sensitive to variations in the static or quasi-static B0 magnetic field due to the limited bandwidth of the receiving systems. Furthermore, it is often necessary to distinguish protons in water from protons in fat in order to image the different tissue types separately. Due to the minimal influence of chemical bonds on the atomic nuclei, the differences in the magnetic moments and the resulting frequency differences of the magnetic resonance signals are very small, which leads to the high sensitivity to deviations in the static magnetic field.
[0006] A magnetic resonance imaging device with permanent magnets is known from document JP H06 181903 A. The magnetic resonance imaging device is equipped with a first magnetic field sensor and a second magnetic field sensor, which are arranged remotely from the device, and a first correction magnetic field coil and a second correction magnetic field coil, which correct the magnetic field in the measurement space based on the detection of the magnetic field fluctuation by the first and second magnetic field sensors.
[0007] It is therefore an object of the present invention to improve the imaging properties of the magnetic resonance imaging device.
[0008] The object is achieved by a magnetic resonance imaging device according to the invention according to claim 1 and a method according to the invention according to claim 5.
[0009] The magnetic resonance imaging scanner according to the invention has a controller. A controller is understood to be a computer or controller that controls the execution of a sequence in the magnetic resonance imaging scanner during image acquisition, in particular the temporal coordination and signals for the gradients, but also the radio-frequency transmission and reception. The magnetic resonance imaging scanner according to the invention further has a plurality of magnetic field sensors. A magnetic field sensor is understood to be a sensor that can detect the strength of at least one magnetic field component or spatial component of a magnetic field or the magnetic field vector. The magnetic field to be detected has a magnetic field strength that is smaller than that of the static magnetic field B0, for example, less than 0.1 T, 10 mT, 1 mT, or 100 microtesla. In particular, the resolution of the magnetic field sensor is better than 100 microtesla or 1 microtesla.The magnetic field sensor has a temporal resolution of less than 100 Hz, 10 Hz, 1 Hz, or 0.1 Hz. The disruptive low-frequency magnetic fields can be caused, for example, by passing trams, trains, or other electric vehicles. Possible sources can also be electric motors in elevators or machinery.
[0010] The magnetic field sensors are arranged in an environment of the magnetic resonance imaging scanner. The "environment" refers to an area outside the magnetic resonance imaging scanner or at least on the inside of a housing of the magnetic resonance imaging scanner. The magnetic field sensors are preferably arranged at such a distance from the magnetic resonance imaging scanner that the magnetic field sensors are not driven to saturation by the static magnetic field of the field magnet and can still detect external interfering magnetic fields, for example, at a distance greater than 1 m or 2 m. The magnetic field sensors are preferably arranged within the space in which the magnetic resonance imaging scanner or the field magnet is located.
[0011] The magnetic field sensors are connected to the controller via a signal connection and can transmit information about the detected magnetic field, either proactively or upon request / polling. The signal connection can be analog or digital, via electrically conductive cables, optical cables, or even wireless, for example via Wi-Fi, Bluetooth, or another radio technology.
[0012] Advantageously, the magnetic resonance imaging device can detect interfering magnetic fields by means of the magnetic field sensors and, by means of the control, as stated below in relation to the method and in the subclaims, can compensate for the interfering magnetic fields on the basis of this information.
[0013] Furthermore, the magnetic resonance imaging device according to the invention is designed to carry out the method described below.
[0014] The method according to the invention comprises the step of detecting a disturbing magnetic field or interference field in the environment of the magnetic resonance imaging scanner with the magnetic field sensors by controlling the magnetic field sensors. In this case, an amount or strength of at least one magnetic field component is detected at a plurality of points in the space around the magnetic resonance imaging scanner. The magnetic field sensors preferably surround the magnetic resonance imaging scanner such that the magnetic field sensors are located at the corner points of a polyhedron surrounding the magnetic resonance imaging scanner. It is also conceivable for the polyhedron to have openings in directions in which no variable source is expected.If the magnetic field on a surface surrounding the magnetic resonance imaging scanner is known, the magnetic field inside the scanner (taking into account known sources inside) and thus also in the imaging area can be advantageously determined using the field equations for magnetic fields. This is also known as the Huygens box.
[0015] Acquisition can also occur continuously or at predetermined intervals. Acquisition may also include preprocessing. For example, filtering or predistortion is conceivable to account for the time or frequency dependence of the sensors or the transfer function of the magnetic field between the sensor and the image acquisition area. For example, higher-frequency alternating magnetic fields are more strongly attenuated by eddy currents in metallic parts of the magnetic resonance imaging scanner.
[0016] Once the magnetic resonance imaging system or its control system is aware of the disturbing magnetic field in the image acquisition area, its effect on the image to be generated can be compensated for in different ways according to the method according to the invention.
[0017] Firstly, according to one embodiment, the effects of this interfering magnetic field on the magnetic resonance data can be compensated or taken into account during reconstruction. A disturbing magnetic field that is, to a first approximation, homogeneous in the image acquisition range leads to a shift in the Larmor frequency or its center frequency. In one step of the method, this center frequency or its shift is determined as a function of the disturbing magnetic field.
[0018] The magnetic resonance data for image acquisition is then acquired in a further step based on this center frequency. A variable-frequency oscillator, e.g., an NCO (numerically controlled oscillator), is provided in the receive path to downconvert the magnetic resonance signal to an intermediate frequency. The controller adjusts the variable-frequency oscillator based on the determined center frequency or the interfering magnetic field in the image acquisition area so that the downconverted magnetic resonance signal has a constant center frequency, thus compensating for the effect of the interfering magnetic field.
[0019] Advantageously, the method according to the invention does not require any structural changes to generate a compensation field and does not require any additional energy to generate it.
[0020] In another embodiment of the method according to the invention, however, it is provided that the effects of the interfering magnetic field are at least partially compensated before the magnetic resonance data are acquired by canceling it out with a compensation magnetic field.
[0021] To this end, a quasi-static compensation magnetic field and / or gradient field offset is determined in a single step depending on the detected magnetic field, which fully or partially compensates for the detected interfering magnetic field for image acquisition. Ideally, the compensation magnetic field is a field with the same magnitude as the interfering magnetic field in the image acquisition area, but with reversed polarity.
[0022] For sources in the same plane as the magnetic resonance imaging scanner, determination is possible using known values on a border of the magnetic resonance imaging scanner in the same plane. This also applies to determining the center frequency. The controller then outputs the determined static magnetic field and / or the gradient field offset. Separate compensation coils are conceivable, for example. Helmholtz coil pairs could be conceivable for one or all spatial directions. The gradient coils can also be controlled with a current that generates parallel and rectified magnetic fields in both coils of a gradient coil pair, also known as gradient field offset.
[0023] In a further step, magnetic resonance data for imaging are acquired in the presence of the compensation fields, using the known sequences, also using methods such as fat saturation.
[0024] Time-limited compensation fields are also conceivable, the strength of which is changed in such a way that the integral effect on a partial sequence is equivalent.
[0025] In this embodiment, the compensation is advantageously transparent to the image reconstruction, so that no changes are required here.
[0026] Preferably, an image is reconstructed from the acquired magnetic resonance data. As already explained, in one embodiment, the reconstruction can also be performed depending on the interfering magnetic field detected by the magnetic sensors in order to compensate for its effect on the image. It is conceivable that values for the interfering magnetic field are stored with the magnetic resonance data. The reconstruction can be performed on the control system of the magnetic resonance imaging scanner, but also by a dedicated reconstruction computer as part of the magnetic resonance imaging scanner or on a remote resource such as a server or as a cloud resource. Finally, the image is displayed to a user on a screen, for example, on the magnetic resonance imaging scanner itself or at a diagnostic workstation.
[0027] In particular, it is also conceivable to advantageously combine the two embodiments. For example, the variable-frequency oscillator can compensate for constant center frequency shifts, preferably using the acquired magnetic resonance data. A gradient of the interfering magnetic field, on the other hand, would require a location-dependent change in the center frequency. The gradient of the interfering magnetic field can, in turn, be compensated by varying the field gradient generated by the gradient coils. Advantageously, the usually stronger zero-order components are thus compensated by the center frequency without additional energy expenditure, while the weaker first- and higher-order components are eliminated with reasonable energy expenditure by generated compensating magnetic fields.
[0028] In an advantageous manner, the method according to the invention with the magnetic resonance imaging device according to the invention enables a complete or at least partial compensation of low-frequency magnetic interference fields.
[0029] Further advantageous embodiments are specified in the subclaims.
[0030] In one possible embodiment of the magnetic resonance imaging system according to the invention, the controller is configured to modify a quasi-static magnetic field and / or a gradient field during image acquisition depending on the information about the magnetic fields, thereby reducing the influence of the detected magnetic field on the image acquisition. In other words, the controller is advantageously configured to implement the previously described method according to the invention by determining and outputting a compensation field based on the magnetic fields detected by the magnetic field sensors and to reduce the effects of an external interference field.A quasi-static magnetic field is defined as a magnetic field that, in contrast to a magnetic field B0 of the superconducting magnet that is constant for a measurement, changes on a time scale with the magnetic field detected by the magnetic field sensors, but also at a frequency orders of magnitude lower than the Larmor frequency of the B1 field of the nuclear spins.
[0031] In one conceivable embodiment of the magnetic resonance imaging system according to the invention, the magnetic field sensors comprise a semiconductor sensor. Hall sensors, for example, are conceivable.
[0032] Advantageously, semiconductor sensors enable the provision of small, cost-effective and robust magnetic field sensors.
[0033] In one possible embodiment of the magnetic resonance imaging device according to the invention, the magnetic field sensors are designed to detect a magnetic field direction of a disturbing magnetic field. For example, the combination of three magnetic field sensors such as z.B. Hall sensors, each of which detects the strength of a component perpendicular to the planar orientation of the semiconductor. The three sensors are preferably aligned orthogonally to each other so that all three components of the magnetic field are detected, thus completely defining the field. However, it is also conceivable to combine individual sensors at different locations with different sensitivity directions to detect the magnetic field. This is particularly feasible when the source of the interfering magnetic field is far away, thus minimizing spatial variation at the location of the magnetic resonance imaging scanner.
[0034] Advantageously, a complete spatial detection of the disturbing magnetic field enables better compensation.
[0035] In one conceivable embodiment of the method according to the invention, the method further comprises the step of outputting a predetermined magnetic calibration field by means of a calibration source. The calibration source can be, for example, a well-defined direct current source or a source of low-frequency alternating current with a frequency, for example, less than 100 Hz, 10 Hz, 1 Hz or 0.1 Hz, which conducts a current through a calibration coil and thus generates a magnetic field. Preferably, the magnetic field of the coil is well known for a known current flowing through the calibration coil. For example, an annular calibration coil would be conceivable, for which the magnetic field can be calculated analytically using the Biot-Savart law.With a known position and orientation of the calibration coil, the magnetic field generated by the calibration coil is then known, at least at the location of the calibration coil and in a free environment thereof, in particular at the location of the magnetic field sensors in an environment of the calibration coil.
[0036] In a further step, the magnetic calibration field is recorded using the magnetic field sensors in the vicinity of the magnetic resonance imaging scanner. This allows the magnetic field sensor to be calibrated using the known magnetic calibration field, i.e., a function can be determined as the relationship between the sensor signal and the magnetic field at the sensor location.
[0037] In another step, the magnetic calibration field in an image acquisition area is recorded using a calibration sequence. Sequences such as those used to record "GIRF" (Gradient Impulse Response Functions, described, for example, in "Signe Johanna Vannesjo et al. GIRF Gradient system characterization by impulse response measurements with a dynamic field camera, February 2013, Magnetic Resonance in Medicine 69(2)") are also conceivable here. This allows the temporal progression or frequency and / or time dependence of the transfer function to be recorded. From this, a functional relationship, optionally also time- or frequency-dependent, between the calibration field and the deviations caused by it in the image acquisition area can be determined as the first transfer function. This measurement is preferably repeated with different calibration fields. The magnetic fields can differ in direction and / or strength.In particular, if the number of different measurements is greater than or equal to the number of magnetic field sensors used, a clear and reversible functional relationship can be determined.
[0038] This first transfer function is used in the step of determining a static magnetic field and / or gradient field offset, for example, by determining the interfering magnetic field in the image acquisition area or FoV using this first transfer function from the measured values of the magnetic field sensors.
[0039] The first transfer function can also be used in one embodiment to determine the center frequency or its shift due to the disturbing magnetic field.
[0040] This advantageously allows the interfering magnetic field in the FoV to be determined more accurately and easily from the measured values of the magnetic field sensors and thus the required compensation field or the shift of the center frequency.
[0041] In one possible embodiment of the method according to the invention, the method further comprises the step of outputting a predetermined magnetic field and / or a gradient field offset via the controller. For example, one or more of the gradient coils or one or more compensation coils can be subjected to a predetermined calibration current. The calibration current can be a direct current or a low-frequency alternating current. Preferably, the temporal profile of the alternating current depends on the detected temporal profile of the magnetic interference field; for example, both have similar frequencies or a similar frequency spectrum.
[0042] In a further step, as already described above, the magnetic field generated by the calibration current with the compensation coil(s) or gradient coil(s) is recorded in an image acquisition area using a calibration sequence. This measurement is preferably repeated with different calibration currents. In particular, if the number of different measurements is greater than or equal to the number of compensation coils or gradient coils used, a clear and reversible functional relationship can be determined. In this way, a functional relationship or a second transfer function is determined between the currents output or set by the controller through the gradient or compensation coil and the magnetic field generated thereby in the image acquisition area / FoV.
[0043] In the step of determining a static or low-frequency magnetic field and / or gradient field offset, the determination is then carried out depending on the second transfer function. For example, the first transfer function can be used to determine the interfering magnetic field in the image acquisition area from the measured values of the magnetic field sensors. A magnetic field to compensate for the interfering magnetic field must then have the same strength and the opposite direction. Using an inverse function of the second transfer function, the required currents to be applied to the compensation coil(s) or gradient coil(s) can be determined. The inverse function can also be determined using numerical methods.
[0044] Since image acquisition is primarily sensitive to the strength of the magnetic field, but less to the direction, it is also conceivable that the magnetic field sensors and the resulting transfer functions do not completely map the disturbing magnetic field in all components, but have a smaller dimensionality than would be necessary.
[0045] Advantageously, the second transfer function or its inverse function can be used to determine the required correction currents more quickly and thus improve the upper limit frequency of the compensation.
[0046] In principle, it would also be conceivable for the first transfer function and the inverse of the second transfer function to be provided by a trained neural network or, more generally, an AI. The input values of the neural network are the measured values of the magnetic field sensors. The output values of the neural network are the control channels for the compensation coils. Artificially interfering magnetic fields can be generated as training data using the calibration coils. The magnetic field in the image acquisition area is measured on a phantom, as already described, for example, using a sequence such as that used to capture the "GIRF." The use of a field camera is also conceivable. The target value is the undisturbed homogeneous magnetic field B0 of the field magnet, or, in the case of combined B0 shimming, an ideal homogeneous magnetic field.The deviation of the detected magnetic field from this target value is then used to correct the parameters of the neural network by back-propagation.
[0047] The above-described properties, features and advantages of this invention, as well as the manner in which they are achieved, will become clearer and more clearly understood in connection with the following description of the embodiments, which are explained in more detail in connection with the drawings.
[0048] They show: Fig. 1 shows a schematic representation of a magnetic resonance tomograph according to the invention; Fig. 2 shows a schematic representation of an exemplary magnetic field sensor; Fig. 3 shows a magnetic resonance tomograph according to the invention with an exemplary arrangement of magnetic field sensors; Fig. 4 shows an exemplary schematic arrangement of coils for generating a compensation field in a magnetic resonance tomograph according to the invention; Fig. 5 shows a schematic flow chart of an exemplary method according to the invention.
[0049] Fig. 1 shows a schematic representation of an embodiment of a magnetic resonance imaging device 1.
[0050] The magnet unit 10 has a field magnet 11 that generates a static magnetic field B0 for aligning nuclear spins of samples or the patient 100 in a recording area. The recording area is characterized by an extremely homogeneous static magnetic field B0, wherein the homogeneity particularly relates to the magnetic field strength or magnitude. The recording area is almost spherical and arranged in a patient tunnel 16 that extends in a longitudinal direction 2 through the magnet unit 10. A patient bed 30 is movable in the patient tunnel 16 by the traversing unit 36. The field magnet 11 is typically a superconducting magnet that can generate magnetic fields with a magnetic flux density of up to 3T, and even higher in the latest devices. However, permanent magnets or electromagnets with normally conducting coils can also be used for lower magnetic field strengths.
[0051] Furthermore, the magnet unit 10 has gradient coils 12, which are designed to superimpose temporally and spatially variable magnetic fields in three spatial directions on the magnetic field B0 for spatial differentiation of the acquired imaging regions in the examination volume. The gradient coils 12 are typically coils made of normally conducting wires that can generate mutually orthogonal fields in the examination volume.
[0052] The magnet unit 10 also has a body coil 14 which is designed to radiate a high-frequency signal supplied via a signal line into the examination volume and to receive resonance signals emitted by the patient 100 and to emit them via a signal line.
[0053] A control unit 20 supplies the magnet unit 10 with the various signals for the gradient coils 12 and the body coil 14 and evaluates the received signals.
[0054] The control unit 20 has a gradient control 21 which is designed to supply the gradient coils 12 with variable currents via supply lines, which provide the desired gradient fields in the examination volume in a time-coordinated manner.
[0055] Furthermore, the control unit 20 has a radio-frequency unit 22, which is designed to generate a radio-frequency pulse with a predetermined temporal profile, amplitude, and spectral power distribution for exciting a magnetic resonance of the nuclear spins in the patient 100. Pulse powers in the kilowatt range can be achieved. The excitation signals can be transmitted into the patient 100 via the body coil 14 or via a local transmitting antenna.
[0056] A controller 23 communicates via a signal bus 25 with the gradient controller 21 and the high-frequency unit 22.
[0057] To receive the magnetic resonance signal, a local coil 50 according to the invention is arranged on the patient 100 in the patient tunnel 16 in order to acquire magnetic resonance signals from an examination area in the immediate vicinity with the greatest possible signal-to-noise ratio. The local coil 50 is in signal communication with a receiver in the radio-frequency unit 22 via a connecting line 33.
[0058] A plurality of magnetic field sensors 60 are arranged around the magnetic resonance imaging scanner 1. The magnetic field sensors 60 exchange information with the controller 23 of the magnetic resonance imaging scanner 1 via a signal connection, indicated here as a wireless radio connection, e.g., via WLAN or Bluetooth. For example, the magnetic field sensors 60 transmit information about the magnetic fields they detect. This information can, for example, comprise an amount of the detected magnetic field or the amount of one, several, or all three spatial magnetic field components. It is conceivable that the magnetic field sensors 60 transmit this information upon request or polling of the controller 23. However, it would also be possible for the magnetic field sensors 60 to transmit this information only at certain intervals or when a change in the magnetic field greater than a threshold value has occurred.
[0059] The signal connection can also be wired, e.g., via electrical or optical cables. The information can be transmitted analogically via voltage, current, or a modulated carrier, or digitally.
[0060] The magnetic field sensors 60 are designed to detect and also resolve magnetic fields on a scale of the magnetic interference fields, i.e., to reproduce them sufficiently for compensation in the signal output by the magnetic field sensor 60. The resolution is preferably less than 5%, 1%, 0.1%, or 0.01% of a maximum amplitude of an interference field. The maximum amplitude of the magnetic interference field can be considered to be at least the Earth's magnetic field, typically 50 microtesla. The frequency-dependent limit values for low-frequency magnetic fields, which can reach up to 40 mT, can also be considered the maximum amplitude.
[0061] Preferably, the magnetic field sensor 60 is still in a substantially linear operating range at the expected magnetic field strengths. This can be achieved, on the one hand, by ensuring that the distance between the magnetic field sensors 60 and the field magnet 10 and the gradient coils 12 is large enough that their magnetic field at the location of the magnetic field sensor 60 is below the maximum amplitude of the expected magnetic interference field. On the other hand, the magnetic field sensors 60 are designed not to saturate at the maximum amplitude of the expected magnetic interference field. Typical values for the strength of the interference fields are between 100 nT and 20,000 nT. The magnetic field sensors 60 are preferably arranged at a distance of between 1 and 5 m from the field magnet.
[0062] Fig. 2 shows a schematic representation of an exemplary magnetic field sensor 60. By way of example, Hall sensors 61 are used here to detect the magnetic field. These sensors have a flat semiconductor material to which a voltage is applied via first contacts on two opposite sides of the surface, so that a current flows between the contacts. A magnetic field component that flows through the surface parallel to the normal vector of the surface causes a voltage at second contacts, which is tapped on two opposite sides of the surface that are perpendicular to the current direction. The voltage is essentially proportional to the strength of the magnetic field component parallel to the normal vector. Using three Hall sensors 61, whose normal vectors span a three-dimensional space, preferably a Cartesian one, the direction and strength of the magnetic field can be detected.
[0063] The voltages tapped from the Hall sensors 61 are amplified by amplifiers 62 and fed to an A / D converter 64 via a multiplexer 63 in time-division multiplexing. In a signal processor 65, the signals are further processed, for example, filtered, and prepared for transmission. For example, data packets can be provided with an assignment of the individual measured values to the respective Hall sensors and thus to the spatial direction and a time reference. It is also conceivable that an initial evaluation of magnitude and direction is already carried out, or that only changes in the magnetic field are detected and transmitted. The data packets are then transmitted wirelessly via a radio-frequency unit 66. The transmission can, for example, be carried out using a standard such as Bluetooth or WLAN.
[0064] Various solutions regarding the magnetic field sensor are conceivable within the scope of the invention. For example, instead of the Hall sensors 61, other magnetic field sensors can be used, which are based, for example, on micromechanics / MEMS or other effects such as magneto-optical or quantum effects. It is conceivable that a magnetic field sensor 60 detects only one component of the magnetic field, but that a plurality of magnetic field sensors 60 are distributed with different orientations relative to the room or the magnetic resonance imaging scanner. The signals can also be transmitted via a cable, electrically or optically. Analog signal transmission is also conceivable.
[0065] Fig. 3 shows an exemplary arrangement of magnetic field sensors 60 of a magnetic resonance imaging scanner 1 according to the invention. The magnetic field sensors 60 are preferably arranged at corners of a polyhedron 70, inside which the magnetic resonance imaging scanner 1 is located. A cuboid or cube is shown here as an example. A tetrahedron as the simplest polyhedron or a polyhedron with multiple corners is also conceivable. The polyhedron 70 is not necessarily regular but depends on the boundary conditions. For example, the magnetic field sensors 60 are preferably arranged in the same room as the magnetic resonance imaging scanner 1, so that the polyhedron 70 must fit into the room. Typically, the ceiling height is less than the length and width dimensions, so that the polyhedron 70 is, so to speak, flattened to fit into the room.As previously discussed, the distances of the magnetic field sensors 60 from the field magnet 10 and the gradient coils 12 are selected such that the magnetic field sensors 60 do not become saturated due to the stray field of the field magnet 10 and / or the gradient coils 12. It is also conceivable that predetermined corner points are not occupied by magnetic field sensors 60 if boundary conditions for the magnetic interference fields, such as predetermined directions or exclusion directions of the sources in relation to the magnetic resonance imaging scanner 1, exist. The number of magnetic field sensors 60 influences the quality of the compensation. The more accurately a magnetic interference field is detected by the magnetic field sensors 60, the more accurately a resulting interference field in the image acquisition space can be determined using the first transfer function. The order of the correction also increases.With two magnetic field sensors 60, a gradient of an interference field can be determined and correspondingly higher orders with more magnetic field sensors 60. This corresponds to the correction of magnetic fields such as dipole fields, quadrupole fields, etc. However, the possibilities of compensation are also limited by the coils used to output compensation fields.
[0066] In Fig. 4 An exemplary arrangement of coils for generating a compensation field is shown. The image acquisition area is arranged within the polyhedron 70, in this case a cuboid. The compensation coils 15 are arranged in pairs on opposite surfaces of the polyhedron 70, thus forming Helmholtz coil pairs, with which homogeneous correction fields can be set in each axis direction x, y, and z, as well as gradients by applying different current intensities to opposing coils.
[0067] Preferably, the existing gradient coils 12 are used as compensation coils 15, which are usually already designed to generate gradients in the x, y, and z directions. By adding compensation currents to the currents required to generate the gradients, or in sections of the sequence using the compensation currents alone, homogeneous magnetic fields and gradient fields can be generated to compensate for external interference fields. Additional coils, particularly for generating higher-order magnetic fields such as quadrupole fields, are also conceivable.
[0068] Fig. 5 shows a schematic flow chart of an exemplary method according to the invention.
[0069] In a step S30, the controller detects a magnetic field in the environment of the magnetic resonance imaging scanner 1 using the magnetic field sensors 60. For example, it is conceivable that the controller 23 queries the measured values of the previously described magnetic field sensors 60 at regular intervals. The magnetic field sensors 60 can also automatically deliver their measured values to the controller 23 at predetermined intervals. It is also conceivable that the measured value sensors 60 only report changes in the detected magnetic field. The detection can, in particular, also include the subtraction of a constant component or offset, such as a stray field of the field magnet 10; however, it is also conceivable that this only occurs in step S40 when a compensation field is determined.It is also possible that acquisition only takes place when the magnetic resonance imaging device 1 itself does not generate any time-varying interference fields, for example in phases of sequences in which no gradient fields are generated by the controller 23.
[0070] In a step S40, the controller 23 determines a magnetic field and / or a gradient field offset for image acquisition to compensate for the detected magnetic field depending on the detected magnetic field. For example, using the first transfer function described below, an interfering magnetic field in the image acquisition space can be determined from the magnetic field detected by the magnetic sensors 60. For compensation, a magnetic field is then required that has an opposite direction in the image acquisition area with the same magnetic field strength, so that the magnetic field vectors of the interfering magnetic field and the compensation field in the image acquisition area cancel each other out. It is necessary for the stray fields caused by the field magnet 10 and / or by the gradient coils 12 to have already been subtracted from the detected magnetic field in step S30, or this is done here in step S40.
[0071] According to the invention, a shift in the center frequency or a changed center frequency caused by the interference field is determined in step S40. To a first approximation, the Larmor frequency and thus also the center frequency of the magnetic resonance signal are proportional to the magnitude of the magnetic field, the sum of the magnetic field B0 and the interference field, at the location of the nuclear spin and the magnetic moment of the nucleus.
[0072] In a step S50, the controller 23 outputs the determined magnetic field and / or the gradient field offset for compensation. In doing so, the controller 23 controls the compensation coils and / or gradient coils 12 each with a compensation current that essentially generates the determined compensation field. It is fundamentally impossible to generate arbitrary magnetic fields with a limited number of compensation coils 15. The output of the magnetic field or currents to the compensation coils 15 and / or gradient coils 12 therefore always remains an approximation and can be solved, for example, by solving an optimization problem in which the interfering magnetic field in the image acquisition area is minimized. A second transfer function, described below, can be used to specify the relationship between the output compensation currents and the magnetic fields generated in the image acquisition area, or their inverse relationship.For example, it is conceivable to vary the compensation currents as parameters in an LSR process until a remaining disturbing magnetic field in the image acquisition area has minimal energy or the field strength in the entire image acquisition area is below a predetermined threshold value.
[0073] It would also be conceivable to train a neural network with different interfering magnetic fields, providing the compensation currents or the changed center frequency as output values, either using the interfering magnetic field in the image acquisition area as input parameters or directly using the magnetic field values detected by the magnetic field sensors 60. It would also be possible to use artificial intelligence based on a different algorithm or hardware.
[0074] In a further step S60, the controller 23 of the magnetic resonance imaging system 1 acquires magnetic resonance data for imaging, with the output magnetic field being output at least during a portion of the sequence required to acquire the magnetic resonance data. For example, it might be advantageous not to apply an offset to the gradient coils 12 during the gradient in order to achieve the maximum gradient strength. Preferably, however, the entire acquisition of the magnetic resonance data is performed with active compensation.
[0075] According to the invention, in step S60, the changed center frequency of the magnetic resonance signal is compensated, sometimes in addition to the compensation magnetic field. Using a variable oscillator for mixing the magnetic resonance signal to an intermediate frequency, the oscillator frequency can, for example, be changed by the controller 23 as a function of the determined interference field such that the effect of the interference field in the acquired magnetic resonance data is at least partially compensated. Likewise, in one embodiment, it is conceivable that the determined interference field or data derived therefrom, which reflect an effect of the interference field on the magnetic resonance data, such as the center frequency, are acquired and stored together with the magnetic resonance data in step S60. Thus, it is conceivable to take these into account in the subsequent signal processing and / or image reconstruction in order to at least partially compensate for the effect of the interference field in the image.
[0076] Preferably, in step S70, an image is reconstructed from the acquired magnetic resonance data. This can be performed by the controller 23 or by an external dedicated reconstruction computer or a cloud resource.
[0077] In one embodiment, it is conceivable that the image reconstruction in step S70 is carried out as a function of the data on the interference field acquired with the magnetic resonance data, so that the effect of the interference field on the reconstructed image is reduced.
[0078] In a step S80, the reconstructed image is output to a user.
[0079] In one possible embodiment of the method according to the invention, a calibration measurement is used to determine a relationship between the measured values of the magnetic field sensors 60 of a calibration field and the magnetic interference field generated thereby in the image capture area in the form of a first transfer function.
[0080] For this purpose, in a step S10, a predetermined magnetic calibration field is generated using a calibration source. The calibration source can, for example, be a current source and a coil through which a current from the current source flows. For ring-shaped or circular coils, for example, the generated magnetic field is well known as a function of the current. The coil is arranged at predetermined locations with a predetermined orientation and is subjected to predetermined currents. It is also conceivable that a low-frequency alternating current is used, for example, to distinguish the field of the calibration coil from other fields such as the Earth's magnetic field or a stray field of the field magnet 11.
[0081] Using time-dependent calibration fields, a frequency-dependent attenuation of the interfering magnetic field on its way into the image acquisition area, for example due to eddy currents in the metal cladding, can also be detected.
[0082] In a step S11, the magnetic calibration field generated by the calibration source is then detected with the magnetic field sensors 60 in an environment of the magnetic resonance imaging scanner 1. By modulating it with a low-frequency alternating current, the calibration field can be distinguished from background magnetic fields using a filter.
[0083] Furthermore, in step S12, the magnetic calibration field in an image acquisition area is acquired using a calibration sequence, for example, with a homogeneous phantom in the image acquisition area. For the acquisition of time-dependent magnetic fields, a previously mentioned GIRF sequence can be used, for example.
[0084] In a step S13, a first transfer function is determined, which determines a relationship between the external magnetic interference fields detected by the magnetic field sensors 60 and the resulting interfering magnetic fields in the image acquisition area. The functional relationship can, for example, be an nxm matrix that represents a linear relationship between the signals of the n magnetic field sensors and m detection points in the image acquisition area. The coefficients of the matrix can be determined, for example, by m repeated measurements with different calibration fields. Analytical methods are also conceivable in which, using a method comparable to a Huygens box, the fields within a space enclosed by the magnetic field sensors 60 can be determined based on the fields on the surface of the shell of the magnetic field sensors 60.It would also be conceivable to use a neural network with n inputs and m outputs, which is trained using the calibration measurements to determine the magnetic field in the image capture area from the measured values of the magnetic field sensors.
[0085] The determined first transfer function is then used during compensation, as already described, to determine or estimate the interfering magnetic field in the image capture area from the measured values of the magnetic field sensors 60 and subsequently to determine the currents for a compensation field.
[0086] The determined transfer function can also be time or frequency dependent.
[0087] In one conceivable embodiment of the method according to the invention, a second transfer function is determined, which establishes a functional relationship between currents through the compensation coil and the offset currents through the gradient coils. Here, too, with p compensation coils or gradient coils involved in the compensation and q measurement points in the image acquisition area, which are acquired using the calibration sequence already described, at least q repeated measurements with linearly independent current configurations are required to determine the pxq transfer matrix. Here, too, it is conceivable, instead of determining a matrix, to train a corresponding neural network or AI that estimates the magnetic fields generated in the image acquisition area from the compensation currents.
[0088] For this purpose, in a step S20, a predetermined magnetic field and / or gradient field offset is output by the controller 23 by applying corresponding predetermined currents to the compensation coils 15 and / or gradient coils 12. Subsequently, in a step S21, the magnetic field generated by the compensation coil 15 and / or gradient field offsets in the gradient coils 12 is detected in an image acquisition area using a calibration sequence. From this, in a step S22, as already described, the second transfer function can then be determined, for example, as a matrix or as a trained neural network.
[0089] The second transfer function can be used later when determining the compensation currents to determine the magnetic fields caused by the compensation currents or, via an inverse function, the required compensation currents for a given compensation field.
[0090] Since an exact match between the interfering magnetic field in the image acquisition area and the compensation field generated there by the compensation coils is only possible in rare exceptional cases, optimization methods are required to generate the appropriate compensation currents. For example, minimizing the sum of the deviation amounts or the maximum deviation would be conceivable. This function could also be performed by an appropriately trained neural network or other artificial intelligence. A neural network that uses the measured values of the magnetic field sensors 60 as input and directly provides the specifications for the compensation currents as output would also be conceivable. A linear time invariant approach, as used in GIRF, would also be possible.
[0091] Although the invention has been illustrated and described in detail by the preferred embodiment, the invention is not limited to the disclosed examples and other variations may be derived therefrom by those skilled in the art without departing from the scope of the invention.
Claims
1. Magnetic resonance tomography unit, wherein the magnetic resonance tomography unit (1) comprises a controller (23) and a plurality of magnetic field sensors (60), wherein the magnetic field sensors (60) are arranged in an environment of the magnetic resonance tomography unit (1) and are embodied to capture low-frequency magnetic fields in the environment of the magnetic resonance tomography unit (1) and to transmit information about the magnetic fields to the controller (23) via a signal connection, wherein the controller (23) is embodied to change signal processing during image capture in dependence on the information about the magnetic fields so that an influence of the captured magnetic field on image capture is reduced, wherein the controller (23) is embodied to capture a magnetic field in an environment of the magnetic resonance tomography unit (1) by way of the controller (23) by means of the magnetic field sensors (60) and to ascertain a centre frequency of the magnetic resonance signal for image capture in dependence on the captured magnetic field by way of the controller (23) characterised in that a variable-frequency oscillator is provided in the receive path to down-mix the magnetic resonance signal to an intermediate frequency, wherein the controller is embodied to set the variable-frequency oscillator in dependence on the ascertained centre frequency in the image capturing region such that the down-mixed magnetic resonance signal has a constant centre frequency and thus the effect of the interfering magnetic field is compensated.
2. Magnetic resonance tomography unit according to claim 1, wherein the controller (23) is embodied to change a quasistatic magnetic field and / or a gradient field during image capture in dependence on the information about the magnetic fields so that the influence of the captured magnetic field on image capture is reduced.
3. Magnetic resonance tomography unit according to one of the preceding claims, wherein the magnetic field sensors (60) comprise a semiconductor sensor.
4. Magnetic resonance tomography unit according to claim 3, wherein the magnetic field sensors (60) are embodied to capture a magnetic field direction.
5. Method for operating a magnetic resonance tomography unit (1), wherein the magnetic resonance tomography unit (1) comprises a controller (23) and a plurality of magnetic field sensors (60), wherein the magnetic field sensors (60) are arranged in an environment of the magnetic resonance tomography unit (1) and embodied to capture low-frequency magnetic fields in the environment of the magnetic resonance tomography unit (1) and to transmit information about the magnetic fields to the controller (23) via a signal connection, wherein the method comprises the steps: (S30) capturing a magnetic field in an environment of the magnetic resonance tomography unit (1) by way of the controller (23) by means of the magnetic field sensors (60); (S40) ascertaining a centre frequency of the magnetic resonance signal for image capture in dependence on the captured magnetic field by way of the controller (23) (S60) ascertaining magnetic resonance data for imaging in dependence on the centre frequency, characterised in that a variable-frequency oscillator is provided in a receive path to down-mix the magnetic resonance signal to an intermediate frequency, wherein the controller sets the variable-frequency oscillator in dependence on the ascertained centre frequency in the image capturing region such that the down-mixed magnetic resonance signal has a constant centre frequency and thus the effect of the interfering magnetic field is compensated.
6. Method according to claim 5, wherein the method furthermore comprises the steps: (S70) reconstructing mapping from the captured magnetic resonance data; (S80) outputting the mapping to a user.
7. Method according to one of claims 5 or 6, wherein the method furthermore comprises the steps: (S10) outputting a predetermined magnetic calibration field by means of a calibration source; (S11) capturing the magnetic calibration field with the magnetic field sensors (60) in an environment of the magnetic resonance tomography unit (1); (S12) capturing the magnetic calibration field in an image capturing region by means of a calibration sequence by way of the controller (23); (S13) ascertaining a first transfer function for a magnetic field between the magnetic field sensors and the image capturing region, wherein, in the step (S40) of ascertaining a centre frequency and / or a quasistatic magnetic field and / or gradient field offset, the ascertaining takes place in dependence on the first transfer function.
8. Method according to one of claims 5 to 7, wherein the method furthermore comprises the steps: (S20) outputting a predetermined magnetic field and / or gradient field offset by way of the controller (23); (S21) capturing the predetermined magnetic field and / or gradient field offset in an image capturing region by means of a calibration sequence; (S22) ascertaining a second transfer function for the magnetic field and / or the gradient field offset and the image capturing region; wherein, in the step (S40) of ascertaining a quasistatic magnetic field and / or gradient field offset, the ascertaining takes place in dependence on the second transfer function.
9. Method according to claim 7 or 8, wherein the steps of ascertaining a first transfer function and / or ascertaining a second transfer function are executed by training a neural network, wherein input values of the neural network are measured values of the magnetic field sensors and output values of the neural network are the actuation channels for the compensation coils, wherein artificial interfering magnetic fields can be generated as training data by means of the calibration coils.
10. Computer program product, which can be loaded into a memory unit of a controller (23) of a magnetic resonance tomography unit (1) according to claim 1, comprising program code means for executing a method according to one of claims 5 to 9 when the computer program product is executed in the controller (23) of the magnetic resonance tomography unit (1) according to claim 1.
11. Computer-readable medium on which program code means are stored which can be read by a controller (23) of a magnetic resonance tomography unit (1) according to claim 1 and can be executed by the controller (23) in order to execute a method according to one of claims 5 to 9.