METHOD FOR OPERATING AN NMR MEASURING SYSTEM AND AN NMR MEASURING SYSTEM WITH A FAST-ADJUSTING NMR-LOCK CONTROLLER

Description

[0001] The invention relates to a method for operating an NMR measuring system with a main field magnet for generating a static main magnetic field in a z-direction, a measuring RF coil for exciting nuclear spins of a measuring substance, an NMR lock controller for detecting field disturbances and for generating a correction field by means of a compensation coil, wherein an NMR lock control is carried out in which spins of atomic nuclei of a reference substance are excited.

[0002] NMR measurement systems require an extremely stable static magnetic field to ensure the required measurement accuracy. To compensate for magnetic field fluctuations, field stabilization techniques are known. These techniques utilize an NMR lock controller to compensate for dynamic changes in the main magnetic field that would otherwise interfere with the NMR measurement. The nuclear spins of specific isotopes (e.g., deuterium) of a reference substance (e.g., water) in the substance to be measured are excited at regular intervals, and the interfering field deviation is calculated from the quotient of the two measured spin-magnetization vector components (M y / M x ). Based on the field deviation thus determined, a corresponding compensation field is then generated in a compensation coil to compensate for the deviation as effectively as possible. Such systems are known from [1], [2], [3].Furthermore, such systems are known from the publications Galuppini et al, Applied Magnetic Resonance, 50(8):125-1047, 2019, Galuppini et al, IFAC-PapersOnLine 50(1):13020-13025, 2017 and Henry et al, Magn Reson Med 42(4):636-642, 1999.

[0003] The systems mentioned in [1], [2], [3] describe methods for field stabilization in a closed control loop, in which the spin magnetization of the reference isotope is used as a measured variable and a corresponding current is fed into an additional coil (H0 coil / compensation coil) as a controlled variable to compensate for the disturbances. These methods assume that the system is already in an equilibrium state or that the amplitude of the y-components My of the spin magnetization is small compared to the amplitude of the x-components Mx of the spin magnetization.

[0004] However, this is not the case when starting the NMR lock control. At the beginning of the NMR lock control, the excitation power is increased to accelerate the growth of the x-component of the spin magnetization. However, this overshoots the target equilibrium value and only reaches it after a corresponding delay. There are unpublished methods known in which different control parameters are selected for starting the NMR lock controller to compensate for the suboptimal conditions during the start-up phase.

[0005] Furthermore, the known methods do not provide for measures to be taken after disturbances caused by the NMR pulse program (gradients, decoupling, etc.) that could destabilize the spin magnetization. Gradient pulses, for example, can influence the spin magnetization in the XY plane. If the gradient pulse is sufficiently strong and long enough, the spin magnetization is "cancelled" in an uncontrolled manner. The settling of the spin magnetization vectors after such a gradient pulse therefore takes a correspondingly long time. During these settling phases, the NMR lock control does not function optimally. In the case of disturbances repeated at short intervals, e.g., due to gradient pulses, the NMR lock control system can be used to compensate for the spin magnetization.Applying gradient pulses or if the radio frequency channel, which is normally intended for the lock control (lock channel), is used in between for the actual NMR measurement, for example for decoupling, results in a loss of measurement accuracy. Object of the invention

[0006] It is an object of the invention to provide a method for operating an NMR measuring system and a corresponding NMR measuring system which allows the NMR measurement to be started more quickly and / or which improves the measurement accuracy in NMR measurements with gradients and / or decoupling. Description of the invention

[0007] This object is achieved according to the invention by a method according to claim 1 and an NMR lock system according to claim 12.

[0008] In the method according to the invention, an RF pulse is applied whose strength is selected such that the x-component of the spin magnetization M of the reference substance immediately after the end of the RF pulse corresponds to the x-component Mxeq of the spin magnetization in the lock equilibrium state at a desired lock excitation power, whereby the x-components Mx and Mxeq of the spin magnetization are aligned in an x-direction that is perpendicular to the z-direction. According to the invention, the equilibrium state in lock mode (lock equilibrium state) of the spin magnetization (i.e., the state of spin magnetization in which the longitudinal magnetization Mz and the x-component Mx of the magnetization in the lock equilibrium state are kept constant) is established by irradiating a targeted RF pulse, in order to then immediately begin NMR lock control.By applying the RF pulse according to the invention, the spin magnetization is "forced" prematurely into its lock equilibrium state. This accelerates the attainment of the equilibrium state for the active NMR lock control from the resting state or after strong disturbances, such as gradient pulses. The method according to the invention complements the already known lock methods and enables the lock equilibrium state to be reached quickly after disturbance events. Existing lock methods can be improved using the method according to the invention.

[0009] The strength of the RF pulse is selected such that the RF pulse brings the spin magnetization into a state where its x-component corresponds to the x-component of the magnetization in the lock equilibrium state with a desired lock excitation strength. The strength of the RF pulse is preferably determined in advance by a known calibration measurement for pulse calibration.

[0010] Depending on the probe head design, the reference substance (typically a deuterium compound) can be excited with the same coil as the substance to be measured, i.e., with the measurement RF coil, or with another coil designed to excite nuclear spins of the corresponding reference substance. The reference substance is also called the lock substance.

[0011] An NMR lock control includes: Excitation of a spin system of atomic nuclei of the reference substance, preferably at regular intervals, detection of field disturbances (interference signals) by determining a deviation between the phase of the signals emitted by the reference substance and their target phase using a field detector of the NMR lock controller, transmission of the detected field disturbances to the control electronics of the compensation coil, generation of a correction field using the compensation coil.

[0012] In a preferred variant, the x-component Mx and the y-component My of the spin magnetization M of the reference substance are 0 before applying the RF pulse, with the y-component My of the spin magnetization M being determined along a y-direction perpendicular to the z-direction and the x-direction. In this method variant, the spin magnetization is thus oriented in the z-direction before applying the RF pulse. This can be an equilibrium state without excitation (spin magnetization = resting magnetization before the start of NMR lock control) or a non-equilibrium state achieved from the state with lock excitation (during NMR lock control) by applying an RF pulse.

[0013] At a first variantIn the method according to the invention, the RF pulse is applied before starting the NMR lock controller. The initial state of the spin magnetization is thus the rest magnetization (equilibrium state without excitation). By applying the RF pulse, the direction of the spin magnetization is tilted so that its x-component is exactly as large as the x-component of the spin magnetization, which develops over the long term with a lock RF excitation with a field strength B1, which corresponds to the rotation frequency ω1 of the spin magnetization of the reference substance.

[0014] The spin magnetization of the reference substance can thus be brought to an "optimal" angle immediately before the initial start of the NMR lock control, i.e., in a time that is very small compared to the time constants of the spin system, which range from a few hundred milliseconds to a few seconds. In reality, the spin magnetization of the reference substance can be brought to an "optimal" angle within a few microseconds, so that the Mx component of the spin magnetization of the reference substance already has the equilibrium value Mxeq at the start of the NMR lock control.

[0015] In this variant, the orientation of the spin magnetization M of the reference substance is preferably adjusted by an angle α which depends on the two time constants T1, T2 of the reference substance and on the strength of the NMR lock excitation k (see below). αis the angle between the z-direction and the rest magnetization vector "tilted" by the RF pulse (i.e., the spin magnetization vector with the magnitude of the rest magnetization before the RF pulse was applied, and with the x-component of the magnetization in the lock equilibrium state). The z-component of the spin magnetization then reduces to the equilibrium value Mzeq. Thus, after the tilting of the spin magnetization direction by the RF pulse, there is a shortening and rotation (further tilting) of the spin magnetization vector with a roughly constant (except for slight fluctuations) Mxeq component. After the z-component of the spin magnetization has reached the equilibrium value, the direction of the spin magnetization exhibits an angle β (= equilibrium angle) with respect to the z-axis.

[0016] Preferably, the angle α elected as follows: sin α = k 1 + k 2 T 2 T 1

[0017] The angle αis determined in advance based on the time constants T1 and T2 of the reference substance used for the NMR lock and the strength of the NMR lock excitation k. The strength of the NMR lock excitation k ranges from 0 to 1, with k = 1 maximizing the value Mxeq of the x-component of the equilibrium spin magnetization of the reference substance. The value k can be determined for the lock excitation power selected by the user, for example, using the following formula: k = ω 1 T 1 T 2 , with ω1: rotation frequency of the spin magnetization of the reference substance due to the lock RF excitation with field strength B1. The relationship between excitation amplitude and angular velocity should be determined by pulse angle determination, and the influence of pulse shaping for the lock excitation should be taken into account.

[0018] Preferably, the NMR signal occurring after application of the RF pulse is detected and a field deviation is determined from it, whereby the determined field deviation is used to generate the correction field before the lock control is switched on.

[0019] At a second variant In the method according to the invention, the RF pulse is applied after the termination of an expected disturbance, in particular after application of a gradient pulse of a gradient system of the NMR measuring system, wherein the orientation of the spin magnetization M of the reference substance is changed by an angle β which depends on the two time constants T1, T2 of the reference substance and on the strength of the NMR lock excitation k.

[0020] After the expected / planned field disturbance (e.g. gradient irradiation or spin decoupling), the direction of the spin magnetization is "tilted" from the z-axis to the equilibrium magnetization Mxeq by the RF pulse according to the invention.

[0021] The initial state of the spin magnetization is preferably a spin magnetization that is aligned in the z-direction (Mx=0, My=0), but whose amplitude has the value of the magnetization in the lock equilibrium state. The angle β is then the angle between the direction of magnetization in the lock equilibrium state and the z-direction. The angle β can be determined in advance using the time constants T1 and T2 of the reference substance used for the NMR lock and the strength of the NMR lock excitation k.

[0022] Preferably, before the expected perturbation, the spin magnetization M of the reference substance is brought from the equilibrium state in lock mode (lock equilibrium state) to a temporary state in which the x-component Mx and the y-component My of the spin magnetization M have the value 0 by applying a preceding RF pulse. This temporary state serves, on the one hand, to eliminate the transverse magnetization before the expected perturbation occurs while simultaneously preserving the magnetization amplitude. The temporary state can therefore also be referred to as a "magnetization amplitude conservation state." By applying this preceding RF pulse before the expected field perturbation (gradient) or decoupling, the direction of the equilibrium magnetization Mxeq is flipped to the z-axis with an RF pulse.Since the application of gradients typically only affects the x-component of the spin magnetization, but not the z-component, the spin magnetization is removed from the influence of the perturbation by "folding back" the spin magnetization in the z-direction. The preceding RF pulse thus causes a controlled "switching away" of the x-components of the spin magnetization, thus preventing the field perturbation from uncontrollably influencing the x-components of the spin magnetization. By controlled "switching away" of the x-components of the spin magnetization (flipping the spin magnetization from the lock equilibrium state to the z-axis), the spin magnetization is brought into a defined state (Mx=0, My=0), from which it can be quickly returned to the lock equilibrium state after the perturbation has ended, since the angle β required for this is already known.

[0023] The preceding RF pulse is inverse to the RF pulse after the expected disturbance has ended (i.e., it causes a tilting of the x-component Mx of the spin magnetization by the angle -β) and can be radiated through the measuring RF coil or through another RF coil.

[0024] The value of the angle β can be determined using the time constants T1 and T2 of the reference substance used for the NMR lock and the strength of the NMR lock excitation k.

[0025] Preferably, the angle β elected as follows: tan β = k T 2 T 1

[0026] The invention also relates to an NMR lock system according to claim 12, comprising a compensation coil, an NMR lock controller for detecting field disturbances and for generating a correction field using the compensation coil, and an RF coil configured to excite nuclear spins of a reference substance. According to the invention, the compensation coil and the RF coil configured to excite nuclear spins of the reference substance are controlled by a common electronic system.

[0027] The NMR lock controller includes a lock channel, i.e., an additional high-frequency channel for the actual NMR measurement, which is not present in the NMR lock controller. This channel continuously measures the NMR signals of the reference substance in parallel with the actual NMR experiment and corrects the main magnetic field with a correction current in the compensation coil whenever the phase of these signals changes. The lock channel also has transmit and receive functionality. The compensation coil is a coil for generating a static or low-frequency field that is added to the static main magnetic field.

[0028] The inventive joint control of the compensation coil and the RF coil designed to excite nuclear spins of the reference substance has the advantage that the RF coil can be used to determine the field deviation (due to the phase deviation between the transmit and receive signals), and this deviation can be immediately "corrected" with the compensation coil without sacrificing a fast data connection between two separate electronic components. This allows the system to be designed in a simpler and more cost-effective manner.

[0029] According to the invention, the NMR lock system according to the invention is designed to carry out the method described above.

[0030] Further advantages of the invention will become apparent from the description and the drawings. Likewise, the above-mentioned and further-described features can be used individually or in combination, provided the resulting feature combination falls within the scope of the invention as defined by the claims. The embodiments shown and described are not to be understood as an exhaustive list, but rather as examples for describing the invention. Detailed description of the invention and drawing

[0031] Fig. 1 shows the transition of the spin magnetization from the thermal equilibrium state with Mx=0 and My =0 to a state with Mx=Mxeq, as well as to the lock equilibrium state. Fig. 2a shows a pulse sequence with an initial RF pulse according to the method according to the invention, as well as the simulated NMR signals obtained from the reference substance and the corresponding course of the spin magnetization. Fig. 2b shows a pulse sequence according to the prior art, as well as the simulated NMR signals obtained from the reference substance and the corresponding course of the spin magnetization. Fig. 3 shows the transition of the spin magnetization between a lock equilibrium state and a non-equilibrium state with Mx=0 and My =0. Fig. 4 shows a pulse sequence with an RF pulse according to the method according to the invention before and after a gradient pulse, as well as the NMR signals obtained from the reference substance. Fig. 5 shows schematically the structure of an NMR lock system according to the invention.

[0032] In the method according to the invention, by irradiating a targeted RF pulse 1, 1' the lock equilibrium state of the spin magnetization is enforced. In contrast to RF pulses 2, The RF pulse 1, 1' according to the invention is characterized by the fact that it causes the magnetization to tilt by a predetermined angle, so that the x-component reaches its equilibrium value. Based on the NMR signal occurring after the RF pulse 1, 1', the current field deviation can be determined (by means of the rotation frequency around the z-axis or - with the onset of equilibrium - by means of a phase deviation) and the field can be corrected before the lock controller is switched on.

[0033] Since the y-component of the spin magnetization is controlled away by the lock controller, i.e., it is only present when the field is disturbed and is very small or quickly disappears again due to the effect of the controller, it is assumed in the following considerations that no y-component of the magnetization is present, which essentially corresponds to the real state.

[0034] Fig. 1 shows the states of the spin magnetization of the reference substance 6 (see Fig. 5 ) at the first variant of the method according to the invention, in which the RF pulse 1 according to the invention is radiated before the NMR lock control with its lock RF pulses 2 is started. The nuclear spins of the reference substance 6 are in a non-excited state with rest magnetization M∞(M = M∞ = magnetization in thermal equilibrium, i.e., without excitation). The rest magnetization vector M∞ is oriented in the z-direction (direction of the static main magnetic field B0).

[0035] By irradiating the RF pulse 1 according to the invention, the spin magnetization M∞ is rotated relative to the z-direction by the angle α tilted in such a way that it has an x-component Mx which corresponds to the x-component Mxeq of the magnetization in the lock equilibrium state at a desired lock excitation power. The magnitude of the spin magnetization after irradiation of the RF pulse 1 according to the invention initially corresponds to the rest magnetization M∞ ( Fig. 1 left). With a delay (depending on a material-specific time constant), the z-component of the spin magnetization is reduced by ΔMz to the equilibrium value Mzeq ( Fig. 1right), so that finally the equilibrium magnetization Meq with Mx=Mxeq and Mz=Mzeq results in (lock equilibrium state).

[0036] Fig. 2a The upper time axis shows a schematic progression of the spin magnetization when applying the method according to the invention. The middle time axis shows the excitation signals RF TX a lock pulse sequence with lock RF pulses 2 (shown as a dome) and an initial RF pulse 1 according to the invention (shown as a rectangle). The lower time axis shows the corresponding MR signals RF RX.

[0037] Fig. 2bThe upper time axis shows a schematic progression of the spin magnetization using a state-of-the-art method. The middle time axis shows the excitation signals RF TX of a corresponding lock pulse sequence. The state-of-the-art lock sequence includes initial lock RF pulses 2' with increased amplitude and "regular" lock RF pulses 2. The lower time axis shows the corresponding received MR signals RF RX.

[0038] The comparison of the Fig. 2a and Fig. 2b The curves of the spin magnetization shown make it clear that when the method according to the invention is used, the lock equilibrium state is reached much faster than with conventional methods in which excitation is initially carried out with an increased power (initial lock RF pulse 2').

[0039] Fig. 3 shows the states of the spin magnetization of the reference substance at a second variantof the method according to the invention, in which before a disturbance (here: before a gradient pulse 3 ) a preceding RF pulse 4 and after the disturbance a subsequent RF pulse 1' The subsequent RF pulse 1' as well as the preceding RF pulse 4 are radiated during lock operation in order, on the one hand, to move the spin magnetization of the reference substance 6 out of the sphere of influence of the gradient pulse 3, which would otherwise disrupt the transverse magnetization, and, on the other hand, to restore magnetization in the lock equilibrium state as quickly as possible after the disturbance has ended in order to ensure reliable lock operation.

[0040] The nuclear spins of reference substance 6 are initially in the lock equilibrium state (Mx=Mxeq, Mz=Mzeq). The magnetization vector Meq in the lock equilibrium state is oriented relative to the z-axis by the angle β By applying the preceding RF pulse 4, the spin magnetization is "flipped" to the z-axis, i.e. by the angle - β tilted and thus brought into a magnetization amplitude conservation state (Mx=0), in which the subsequent gradient pulse 3 has no influence on the spin magnetization. The preceding RF pulse 4 thus causes the transverse magnetization to be switched off. After the gradient pulse 3 has ended, the subsequent RF pulse 1' is radiated, which tilts the spin magnetization M relative to the z-direction by the angle βThe spin magnetization is thus brought back to the lock equilibrium state (Mx=Mxeq, Mz=Mzeq). The time interval between the preceding RF pulse 4 and the following RF pulse 1' is so small compared to the time constants T1 and T2 of the spin system that in the temporary state in which the x-component of the spin magnetization is zero, the amplitude of the spin magnetization does not change, or does not change significantly. The strength of the subsequent RF pulse 1' can therefore be chosen to be equal to the strength of the preceding RF pulse 4.

[0041] Fig. 4 shows the excitation signals on the upper time axis RF_TX A lock pulse sequence with lock RF pulse 2 (shown as a dome), the preceding RF pulse 4, and the following RF pulse 1' (each shown as a rectangle). The middle time axis shows the corresponding MR signals. Gradient pulse 3 is shown on the lower time axis.

[0042] Fig. 5shows schematically the structure of an inventive NMR lock system 5 for carrying out the method according to the invention. The reference substance used for the NMR lock control 6 is in the main magnetic field B0, which is generated by a main magnet (not shown) of an NMR measurement system. The NMR lock system 5 comprises an NMR lock controller 7 , a compensation coil 8 and an RF coil 9 , which is designed to excite nuclear spins of a reference substance and receive the spin signals. The reference substance 6 is located in the effective range of the RF coil. 9, with which the nuclear spins of the reference substance 6 can be excited and measured in order to detect disturbing field deviations. The RF coil 9 is preferably designed as a saddle coil and Fig. 5only schematically indicated. Furthermore, the reference substance 6 is located in the effective range of the compensation coil 8 for generating a compensation field with which the B0 field can be corrected. The compensation coil 8 and the RF coil 9 are controlled by a common electronics 10 of the NMR lock controller 7. List of reference symbols

[0043] 1Initial RF pulse 1'RF pulse after disturbance 2Lock RF pulse 2'Lock RF pulse with increased amplitude 3Gradient pulse 4Preceding RF pulse 5NMR lock system 6Reference substance 7NMR lock controller 8Compensation coil 9RF coil 10Electronics of the NMR lock controller B0Main magnetic field H0Compensation channel RF_TXExcitation signal RF-RXReceive signal Reference list

[0044] [1]EP 0 522 191 B1 [2]EP 1 887 375 B1 [3]EP 1 621 894 B1

Claims

1. Method for operating an NMR measurement system having a main field magnet for generating a static main magnetic field B0 in a z direction, a measurement RF coil for exciting nuclear spins of a measurement substance, an NMR lock controller (5) for detecting field disturbances and for generating a correction field by means of a compensation coil (8), lock control being carried out, in which spins of atomic nuclei of a reference substance (6) are excited, characterized in that an RF pulse (1, 1') is applied, the strength of which is selected so that the x component of the spin magnetization M of the reference substance (6) immediately after the end of the RF pulse corresponds to the x component Mxeq of the spin magnetization in the lock equilibrium state at a desired lock excitation power, the x components Mx and Mxeq of the spin magnetization being oriented in an x direction which is perpendicular to the z direction.

2. Method according to claim 1, characterized in that the state in which the x component Mx of the spin magnetization has the value Mxeq is achieved by applying a single RF pulse (1, 1').

3. Method according to either of the preceding claims, characterized in that the x component Mx and the y component My of the spin magnetization of the reference substance (6) are zero before the RF pulse (1, 1') is applied, the y component My of the spin magnetization being established along a y direction which is perpendicular to the z direction and the x direction.

4. Method according to any of the preceding claims, characterized in that the RF pulse (1) is applied before the NMR lock controller (5) is started.

5. Method according to claim 4, characterized in that the orientation of the spin magnetization of the reference substance (6) is changed by an angle α which depends on the two time constants T1, T2 of the reference substance (6) and on the strength k of the NMR lock excitation, where k = ω 1 T 1 T 2 , where ω1 is rotation frequency of the spin magnetization of the reference substance (6) due to the lock RF excitation with the field strength B1.

6. Method according to claim 5, characterized in that the angle α is selected as follows: sin α = k 1 + k 2 T 2 T 1 7. Method according to any of the preceding claims, characterized in that after the RF pulse (1, 1') is applied, the occurring NMR signal is detected and a field deviation is determined therefrom, the determined field deviation being used to generate the correction field before the lock controller (5) is switched on.

8. Method according to any of the preceding claims, characterized in that the RF pulse (1') is applied after an expected disturbance is terminated, in particular after a gradient pulse (3) of a gradient system of the NMR measurement system is applied, the orientation of the spin magnetization M of the reference substance (6) being changed by an angle β which depends on the two time constants T1, T2 of the reference substance (6) and on the strength k of the NMR lock excitation k.

9. Method according to claim 8, characterized in that before the expected disturbance (3), the spin magnetization M of the reference substance (6) is brought from the equilibrium state in lock mode into a temporary state in which the x component Mx and the y component My of the spin magnetization M have the value 0 by applying a preceding RF pulse (4).

10. Method according to claim 8 or 9, characterized in that the value of the angle β is determined on the basis of the time constants T1 and T2 of the reference substance used for the NMR lock (6) and on the basis of the strength of the NMR lock excitation k.

11. Method according to claim 10, characterized in that the angle β is selected as follows: tan β = k T 2 T 1 12. NMR lock system (5) comprising a compensation coil (8), an NMR lock controller (7) for detecting field disturbances and for generating a correction field by means of the compensation coil (8), and an RF coil (9) which is designed to excite nuclear spins of a reference substance (6), characterized in that the compensation coil (8) and the RF coil (9) designed to excite nuclear spins of the reference substance are actuated by a common electronics unit (10), and in that the NMR lock system (5) is designed to carry out the method according to any of claims 1 to 10.