Patent Application
20240377493
The invention relates to magnetic resonance imaging, in particular to a gradient system for a magnetic resonance imaging system. The invention also relates to a method of operating a gradient system and a computer program product.
A gradient system 200 consists of multiple components as illustrated in
So far, this concept has been scientifically tested for non-cartesian imaging to derive the actual k-space trajectory present during the MR signal sampling. For this purpose, in parallel to the MR data acquisition also the currents running through the gradient coil 210 are sampled in a synchronized manner. This is a straightforward application because image reconstruction 216 begins after sampling the MR data and at that point the sampled currents that generate the actual gradient waveform 204 and the thus derived k-space trajectory can be made easily available. It should be noted that the current-based correction approach, as discussed here, is limited to the readout gradients, where the k-space trajectory for reconstruction can be adapted to the real fields.
The document WO 2019/179797 A1 discloses a magnetic resonance imaging system with a gradient coil system that comprises a set of gradient coils configured for generating a gradient, a gradient coil amplifier, and a current sensor system configured for measuring current sensor data descriptive of the electrical current supplied to each of the set of gradient coils. Execution of the machine executable instructions causes a processor to: control the magnetic resonance imaging system with the pulse sequence commands to acquire magnetic resonance imaging data; record the current sensor data during the acquisition of the magnetic resonance imaging data; calculate a corrected k-space trajectory using the current sensor data and a gradient coil transfer function; and reconstruct a corrected magnetic resonance image using the magnetic resonance imaging data and the corrected k-space trajectory.
However, when the performance of the gradient amplifiers is too much compromised, also the performance of other gradients involved in the MR scan are compromised. For crusher gradients this is often not much of concern, but it matters in case of MR signal excitation or refocusing processes involving gradients. For optimal performance of spatially selective RF pulses, the behavior of the gradient system must be sufficiently predictable or known in advance.
A real-time feedback of sensed selection gradient currents to the RF transmit object could help to actively steer optimal selection, but it somehow mimics the functionality of the current amplifier control loop with very high bandwidth demands. Such an approach, apart from unknown feasibility and performance, is another undesired cost driver.
U.S. Pat. No. 6,377,043 B1 describes a system that enables the temporal variation of the magnetic field generated by a gradient coil to be derived from the measured temporal variation of the current through gradient coil and from a pulse response which has been determined once and stored for the relevant coil.
A paper titled “Time optimal control-based RFT pulse design under gradient imperfections” by C. Aigner et. al. describes a gradient system imperfections model that fits into a control framework for radio frequency (RF) pulse design.
It is an object of the invention to improve the control of gradient fields for a higher image quality and to minimize the need for continuous monitoring of the output-current of the gradient amplifier.
According to the invention, this object is addressed by the subject matter of the independent claims. Preferred embodiments of the invention are described in the sub claims.
Therefore, according to the invention, a gradient system for a magnetic resonance examination system for generating a gradient magnetic field within an imaging zone of the magnetic resonance examination system is foreseen, the gradient system comprising:
The basic idea behind is, that instead of only monitoring the read-out gradients, also the selection gradients, or potentially other important ones used for MR signal encoding, are monitored via current sensing. Based on the known input gradient pulse shape and the sensed output currents conclusions can be drawn, how the limited amplifier performance is compromising the output current. Due to the non-linear effect occurring in the active amplifier control loop, these compromising effects cannot be simply modelled but need to rely on measured input. The detected currents are related to actual selection gradient fields via the current to field modulation transfer function (CGMTF) of the corresponding gradient coil channel. The transfer function is used to predict, based on the sensed current, the actual selection gradient pulse shape (Gf) in the gradient coil. The invention achieves to avoid the need for continuous monitoring the output-current of the gradient amplifier.
In an embodiment of the invention the selection gradient field is a spatial read-encoding gradient field and/or a crusher gradient field and/or a slice selection gradient field and/or a slice refocusing gradient field and/or a gradient field spatially selective in one or multiple dimensions.
In another aspect the invention provides a magnetic resonance imaging system configured for acquiring magnetic resonance imaging data from an imaging zone, wherein the magnetic resonance imaging system comprises a gradient system as described above.
In another aspect the invention provides a method of operating a gradient system for a magnetic resonance examination system for generating a gradient magnetic field within an imaging zone of the magnetic resonance examination system, wherein the method comprises the following steps:
In an embodiment of the invention the step of obtaining an actual selection gradient pulse shape by applying a current to field modulation transfer function (CGMTF) to the sampled electrical current comprises the step of obtaining an actual selection gradient pulse shape by a convolution operation according to the following formula:
G
f(t)=I(t)*CGMTF(t)
In another embodiment of the invention the current to field modulation transfer function is pre-recorded before applying the function to the actual selection gradient pulse shape.
In an embodiment of the invention the step of predicting gradient distortions of the actual selection gradient pulse shape based on a known input selection gradient pulse shape is performed by means of artificial intelligence.
In another embodiment of the invention the step of predicting gradient distortions of the actual selection gradient pulse shape by means of artificial intelligence comprises the step of using a trained convolutional neural network that returns the actual selection gradient pulse shape from parameter values representing the desired gradient pulse shape.
In an embodiment of the invention the parameters values are selected from the following list: max. gradient strength and/or slew rate and/or strength and/or variation when using variable-rate selective excitation (VERSE).
In another embodiment of the invention the step of sampling the electrical current supplied to each of the set of gradient coils by the gradient coil amplifier by the current sensor system, comprises the step of sampling the electrical current during a start-up phase of the magnetic resonance imaging system with initial sequence dummy shots, or a short prep-scan involving those gradient pulses in question.
In an embodiment of the invention the step of correcting for gradient distortions of the actual selection gradient pulse shape by adapting radio frequency pulse shapes that are emitted concurrently with the actual selection gradient pulses takes place in addition to correcting for gradient distortions of a read-out gradient of the gradient system.
In another aspect the invention provides a computer program product comprising instructions, when the program is executed by a computer, cause the computer to carry out the method as described by the method steps above.
In a further aspect the invention provides a computer program product comprising instructions, when the program is executed by a computer, cause the computer to control a magnetic resonance imaging system to correct for gradient distortions of an actual selection gradient pulse shape by adapting radio frequency pulse shapes that are emitted concurrently with the actual selection gradient pulses.
These and other aspects of the invention will be apparent from and elucidated with reference to the embodiments described hereinafter. Such an embodiment does not necessarily represent the full scope of the invention, however, and reference is made therefore to the claims and herein for interpreting the scope of the invention.
In the drawings:
The magnetic resonance imaging system 100 comprises a gradient system 102, wherein the gradient system 102 is the prime spatial encoding system used in magnetic resonance imaging. The gradient system 102 consists of multiple components. Within the bore 106 of the magnet there is at least one gradient coil 110 which used for acquisition of preliminary magnetic resonance imaging data to spatially encode magnetic spins within the imaging zone 108 of the magnet 104. The gradient coil 110 is connected to a magnetic field gradient coil amplifier 112. The gradient coil 110 can e.g. contain three separate coils for spatially encoding in three orthogonal spatial directions. A magnetic field gradient power supply supplies current to the gradient coil. The current supplied to gradient coil 110 is controlled as a function of time and may be ramped or pulsed. The gradient coil 110 can represent in an embodiment of the invention three separate orthogonal gradient coils for generating a gradient magnetic field within the imaging zone 108. These are typically oriented as the axes 122, 123 and 124 show. The axis 124 is aligned with the axis of the magnet 104. This is typically referred to as the z-axis. 122 and 123 are the x and y-axes, respectively. They are orthogonal to each other and also to the z-axis 124.
The magnetic field gradient coil amplifier 112 is configured for supplying current to the gradient coil. The magnetic field gradient coil amplifier 112 is shown as having a current sensor system 113 for measuring the current supplied to the gradient coil 110. The current sensor system 113 could for example be part of the magnetic field gradient coil amplifier 112 or it could also be integrated into the gradient coil 110. Adjacent to the imaging zone 108 is a radio-frequency coil 114 for manipulating the orientations of magnetic spins within the imaging zone 108 and for receiving radio transmissions from spins also within the imaging zone 108. The radio frequency antenna may contain multiple coil elements. The radio frequency antenna may also be referred to as a channel or antenna. The radio-frequency coil 114 is connected to a radio frequency transceiver 116. The radio-frequency coil 114 and radio frequency transceiver 116 may be replaced by separate transmit and receive coils and a separate transmitter and receiver. It is understood that the radio-frequency coil 114 and the radio frequency transceiver 116 are representative. The radio-frequency coil 114 is intended to also represent a dedicated transmit antenna and a dedicated receive antenna. Likewise the transceiver 116 may also represent a separate transmitter and receivers. The radio-frequency coil 114 may also have multiple receive/transmit elements and the radio frequency transceiver 116 may have multiple receive/transmit channels. For example if a parallel imaging technique such as SENSE is performed, the radio-frequency coil 114 will have multiple coil elements.
The transceiver 116 and the gradient controller 112 are shown as being connected to a hardware interface 128 of a computer system 126. The computer system further comprises a processor 130 that is in communication with the hardware system 128, a memory 134, and a user interface 132. The memory 134 may be any combination of memory which is accessible to the processor 130. This may include such things as main memory, cached memory, and also non-volatile memory such as flash RAM, hard drives, or other storage devices. In some examples the memory 134 may be considered to be a non-transitory computer-readable medium. The memory 134 is shown as containing machine-executable instructions 140. The machine-executable instructions 140 enable the processor 130 to control the operation and function of the magnetic resonance imaging system 100. The machine-executable instructions 140 may also enable the processor 130 to perform various data analysis and calculation functions.
G′
S(t)=I(t)*CGMTF(t).
In one embodiment of the invention, the prediction of GS(t) is learned using artificial intelligence. By comparing the known desired input gradient Gi(t) with the actual gradient G′S(t) realized in the bore and adding some describing parameters to the desired gradient (like max. gradient strength, slew rate, strength, variation when using VERSE, etc. . . . ) a simple convolutional network can be trained that is able to predict based on the nominal input gradient waveform and its basic parameters the actual one. The necessary training data can be generated during initial MRI system installation doing some test scans or during the initial testing phase of the final system. This learning could be performed in each MR system individually or when using identical hardware once in the factory, applying the trained AI model to the whole system fleet. The gradient waveform predicted this way can then be used in the RF design process, e.g. for low tip angles, to realize the target spatial selection profile.
In another embodiment of the invention, the selection gradient currents are monitored during the initial sequence dummy shots that are commonly applied to achieve steady state for the MR signal in a clinical scan. Feeding that back to RF pulse design, changing RF waveforms maybe on the fly during this dummy phase by considering the actually applied gradient fields to achieve better slice definition for upcoming RF pulses fixes the corresponding RF issue accordingly. It is obvious that these adaptations can be done to MR excitation, to refocusing or any spatially selective RF pulse used for instance for appropriate magnetization preparation. It should be further noted that this holds for RF pulses, that are spatially selective in one or multiple dimensions including the application to one or a number of transmit coils.
The proposed AI learning and dummy shot evaluation methods can also be used to better predict actual gradients during signal readout and thus may be a path to removing non-linear effects without continuous reading of output currents or direct gradient fields during imaging.
In an embodiment of the invention the case of sensing the gradient waveform during the start-up phase can be considered. This is basically illustrated in
G′s(t)=Iz(t)*CGMTFz(t)
The CGMTF may be measured once e.g. at the end of the MR system installation. As a result, the actual gradient G′S generated in the gradient coil 110, 310 can be obtained. When the RF excitation pulse belongs to the low tip angle class, for calculation the new and corrected RF pulse, Pauly's approach can be used. For this excitation case also, the refocussing lobes have to be considered although there is no RF present. It facilitates the refocussing of the excited transverse magnetization and should amount in the low tip angle approximation half of the integral of the selection gradient to maintain SNR. Please note, that the same formalism holds when using oblique selection gradients, which consist of all the three elementary gradient components. In this case the convolution given in the equation above has to include more components (at least x,y,z) of the sampled currents and the corresponding transfer functions (CGMTF). Please note, that the CGMTF has even more component than the basic gradients. Apart from the B0 term, higher order terms (including cross terms) could be considered to appropriately predict the actual field from the measured currents.
While the invention has been illustrated and described in detail in the drawings and foregoing description, such illustration and description are to be considered illustrative or exemplary and not restrictive; the invention is not limited to the disclosed embodiments. Other variations to the disclosed embodiments can be understood and effected by those skilled in the art in practicing the claimed invention, from a study of the drawings, the disclosure, and the appended claims. In the claims, the word “comprising” does not exclude other elements or steps, and the indefinite article “a” or “an” does not exclude a plurality. The mere fact that certain measures are recited in mutually different dependent claims does not indicate that a combination of these measures cannot be used to advantage. Any reference signs in the claims should not be construed as limiting the scope. Further, for the sake of clearness, not all elements in the drawings may have been supplied with reference signs.
| Number | Date | Country | Kind |
|---|---|---|---|
| 21190021.2 | Aug 2021 | EP | regional |
| Filing Document | Filing Date | Country | Kind |
|---|---|---|---|
| PCT/EP2022/071191 | 7/28/2022 | WO |
