A nuclear magnetic resonance imaging device and method for rapid on-site detection

Abstract

The application provides a nuclear magnetic resonance device and method which can be used for on-site rapid detection. The application relates to a simple, portable and easy-to-operate nuclear magnetic resonance device. Through the design of a core probe component, the device can provide rapid analysis and nondestructive detection equipment for different on-site application scenarios. A new rapid measurement relaxation analysis method is provided. Through corresponding data processing on the collected data of the technology, important information such as the internal structure, micropore characteristics and fluid saturation occurrence of the detected sample can be quickly analyzed, so that the technology has important application value in the fields of online biomedicine, chemical engineering and geological on-site exploration.

Description

Technical Field

[0001] This invention relates to the field of nuclear magnetic resonance, and more particularly to a nuclear magnetic resonance imaging device and method that can be used for rapid on-site detection. Background Technology

[0002] Nuclear magnetic resonance imaging (NMR) technology, as an advanced non-destructive testing method, has extremely wide applications in many fields such as medicine, biology, energy, materials, agriculture and forestry, food, safety monitoring, and chemical engineering. Currently, there is an urgent need for NMR imaging equipment and related testing methods that enable in-situ testing, independent of indoor testing and analysis. In-situ NMR imaging measurements first require equipment with low magnetic leakage and portability, among other safety features. Meanwhile, indoor NMR equipment contains large, expensive gradient systems and water-cooling systems, which are not conducive to on-site testing. Therefore, to maximize the flexibility of on-site testing, on-site NMR equipment needs to be as compatible as possible with simple and lightweight gradient systems.

[0003] Measurement methods suitable for on-site MRI detection are equally important. Taking the longitudinal relaxation time T1 parameter as an example, due to the differences in longitudinal relaxation time among different types of components in the detected object, MRI T1 imaging technology can provide the most direct and effective evidence for in-situ detection of lesion mechanisms in biological tissues, and is therefore a conventional MRI weighted imaging method. This invention also simultaneously proposes a relaxation measurement method based on rapid on-site detection of MRI devices. Summary of the Invention

[0004] The purpose of this invention is to solve the above-mentioned problems existing in the prior art, and to propose a magnetic resonance imaging device and method that can be used for rapid on-site detection.

[0005] To achieve the above objectives, the present invention adopts the following technical solution:

[0006] A magnetic resonance imaging (MRI) device for rapid on-site detection includes a magnet. The magnet part consists of a main magnet unit and an auxiliary magnet unit. The main magnet system uses rare-earth permanent magnet blocks with a specific cross-sectional shape (e.g., hexagonal), arranged in multiple magnet arrays along the axial direction. Each magnet array consists of a specific number of permanent magnets (e.g., 16 blocks) arranged circumferentially along the cross-section, forming a hollow cylindrical permanent magnet device. The static magnetic field in the middle part of this magnet device is a uniform field. To achieve the imaging encoding function, the device of the present invention adds an auxiliary magnet unit on the basis of the main magnet unit. The auxiliary magnet unit consists of two pairs of magnets with opposite polarization directions arranged along the axial direction, thereby forming an auxiliary magnetic field along the axial direction in the middle magnet cavity. The gradient of this auxiliary magnetic field is G. By applying a corresponding pulse sequence, MRI relaxation imaging measurement is achieved.

[0007] Simultaneously, this invention patent proposes a novel rapid T1-weighted magnetic resonance imaging method, the method specifically comprising:

[0008] Step 1: In the GRD channel, due to the presence of the additional magnetic field unit, an additional magnetic field with a specific magnetic field gradient G will be formed along the z direction.

[0009] Step 2: Apply a 90° radio frequency pulse to the hydrogen-containing proton spin system of the sample under test in the TRS channel to rotate the macroscopic magnetization vector M0 to the transverse plane;

[0010] Step 3: After waiting for a very short time τ, apply a 180° radio frequency pulse to the hydrogen-containing proton spin system of the sample under test on the TRS channel to obtain the transverse planar magnetization vector after re-aggregation of the dephase.

[0011] Step 4: After waiting for a very short time τ again, apply a 90° radio frequency pulse to the hydrogen proton spin system of the sample under test on the TRS channel to rotate the re-aggregated transverse planar magnetization vector by 90° to the longitudinal axis.

[0012] Preferably, the longitudinal axis is aligned with the direction of the static magnetic field;

[0013] Step 5: Next, apply a small-angle α radio frequency pulse to the TRS channel, followed by a frequency-coded gradient magnetic field on the GRD channel.

[0014] Preferably, the height of the frequency-coded gradient pulse is fixed at G;

[0015] Step 6: Repeatedly apply 180° radio frequency pulses to the TRS channel. After a certain period of time, acquire a complete discrete spin echo signal in the ACQ channel. When starting to acquire the spin echo signal, first apply a frequency decoding gradient pulse to the GRD channel. The amplitude of the frequency decoding gradient pulse is the same as that of the frequency encoding gradient pulse in Step 5.

[0016] Step 7: Apply another 180° RF pulse to the TRS channel to flip the magnetization vector remaining in the longitudinal direction;

[0017] Step 8: From the application of a small-angle α radio frequency pulse on the TRS channel in Step 5 to the application of the last constant amplitude gradient on the GRD channel, the duration of the entire timing sequence is Δ; the timing sequence is continuously cycled N times, and N spin echo signals will be finally acquired in the ACQ channel; the echo train signal M(k, NΔ) is obtained, where k is a defined wave function; the acquired echo train signal M(k, NΔ) is processed by nuclear magnetic resonance data to obtain the fast nuclear magnetic resonance one-dimensional T1 imaging results.

[0018] Furthermore, this invention patent proposes a novel rapid nuclear magnetic resonance T1-T2 correlation imaging method, the method specifically comprising:

[0019] Step 1: In the GRD channel, due to the presence of the additional magnetic field unit, an additional magnetic field with a specific magnetic field gradient G will be formed along the z direction.

[0020] Step 2: Apply a 90° radio frequency pulse to the hydrogen-containing proton spin system of the sample under test in the TRS channel to rotate the macroscopic magnetization vector M0 to the transverse plane;

[0021] Step 3: After waiting for a very short time τ, apply a 180° radio frequency pulse to the hydrogen-containing proton spin system of the sample under test on the TRS channel to obtain the transverse planar magnetization vector after re-aggregation of the dephase.

[0022] Step 4: After waiting for a very short time τ again, apply a 90° radio frequency pulse to the hydrogen proton spin system of the sample under test on the TRS channel to rotate the re-aggregated transverse planar magnetization vector by 90° to the longitudinal axis.

[0023] Preferably, the longitudinal axis is aligned with the direction of the static magnetic field;

[0024] Step 5: Next, apply a small-angle α radio frequency pulse to the TRS channel, followed by a frequency-coded gradient magnetic field on the GRD channel.

[0025] Preferably, the height of the frequency-coded gradient pulse is fixed at G;

[0026] Step 6: Repeatedly apply 180° radio frequency pulses to the TRS channel. After a certain period of time, acquire a complete discrete spin echo signal in the ACQ channel. When starting to acquire the spin echo signal, first apply a frequency decoding gradient pulse to the GRD channel. The amplitude of the frequency decoding gradient pulse is the same as that of the frequency encoding gradient pulse in Step 5.

[0027] Step 7: Apply another 180° RF pulse to the TRS channel to flip the magnetization vector remaining in the longitudinal direction;

[0028] Step 8: From the application of a small-angle α radio frequency pulse on the TRS channel in Step 5 to the application of the last constant amplitude gradient on the GRD channel, the duration of the entire timing sequence is Δ; the timing sequence is continuously cycled N times, and N spin echo signals will eventually be acquired in the ACQ channel.

[0029] Step 9: Change the length of the waiting time τ in steps 3 and 4, and then edit the transverse relaxation time T2; continuously cycle the time sequence P times, and finally acquire P*N spin echo signals in the ACQ channel; obtain the echo train signal M(k, NΔ, Pτ), where k is a defined wave function; perform nuclear magnetic resonance data processing on the acquired echo train signal M(k, NΔ, Pτ) to obtain the fast nuclear magnetic resonance T1-T2 correlated imaging results.

[0030] The specific steps for processing the nuclear magnetic resonance data are as follows:

[0031] First, the data is subjected to Fourier transform to decompile it in the imaging dimension; then, the decomputed data is subjected to Inverse Laplace inversion to obtain T1 imaging or T1-T2 correlated imaging results.

[0032] Beneficial effects:

[0033] 1. The entire magnet system can be assembled from small, lightweight, and inexpensive permanent magnets.

[0034] 2. The permanent magnet block cross-section of the magnet system has the magnet polarization direction perpendicular to the surface of the magnet block, which facilitates processing, fixing and installation.

[0035] 3. The leakage magnetic field of the magnet structure is weak, which has little impact on the operation of nearby motors and other ferromagnetic components, thus meeting the safety requirements of on-site nuclear magnetic resonance analysis and measurement.

[0036] 4. The main magnet system can generate a sufficiently strong magnetic field for nuclear magnetic resonance analysis experiments and theoretical research.

[0037] 5. The additional magnet system can directly provide a gradient environment for imaging encoding for the whole device. This design ensures that the invention imaging device no longer needs a large and expensive gradient amplifier and water cooling system. Moreover, the magnitude of the gradient can be changed according to the actual imaging needs by adjusting the physical properties of the additional magnet system, which greatly facilitates the practicality and adaptability of the whole device in on-site testing.

[0038] 6. The unique structural design of this nuclear magnetic resonance analyzer removes the limitation on the length of the sample being measured. In addition, it can perform motion scanning measurements on the sample, which speeds up the measurement process, saves measurement time, and ultimately enables the evaluation and description of the sample at both the microscopic and macroscopic levels.

[0039] The relaxation imaging method designed in this invention starts from the perspective of quantum mechanics. By explaining the basic theory, it rationally arranges and optimizes the corresponding pulses in different time periods of the pulse sequence, which can greatly shorten the operation time of nuclear magnetic resonance imaging technology and provide a fast solution for nuclear magnetic resonance imaging technology in a variety of potential application fields. Attached Figure Description

[0040] Figure 1 This is a schematic diagram of the coordinate axis definition of the magnet system and the arrangement of a single layer of magnets in the main magnet unit according to an embodiment of the present invention;

[0041] Figure 2 This is a schematic diagram of the axial cross-section of the main magnet unit and the magnet polarization and magnetic field distribution according to an embodiment of the present invention;

[0042] Figure 3 This is a schematic diagram of the axial cross-section of the magnet polarization and magnetic field distribution of the all-magnet unit according to an embodiment of the present invention;

[0043] Figure 4 This is a novel nuclear magnetic resonance T1 imaging pulse sequence according to an embodiment of the present invention;

[0044] Figure 5 This is a novel nuclear magnetic resonance T1-T2 imaging pulse sequence according to an embodiment of the present invention;

[0045] Figure 6 Reconstruct the flowchart based on actual data and results.

[0046] Figure 7 This is an example of T1 imaging results.

[0047] Figure 8 Example of T1-T2 imaging results.

[0048] In the diagram: TRS is the pulse emission channel of the nuclear magnetic resonance system, GRD is the gradient pulse emission channel, and ACQ is the signal reception channel of the nuclear magnetic resonance system. Detailed Implementation

[0049] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments.

[0050] First, the definitions and physical meanings of the relevant technical terms involved in this invention are introduced as follows:

[0051] The static magnetic field B0 is provided by a magnet and determines the signal-to-noise ratio of the NMR signal. When the sample is placed in the static magnetic field, energy level splitting occurs within the spin system, generating a macroscopic magnetization vector M0 along the direction of the static magnetic field. M0 is determined by parameters such as the static magnetic field strength B0 and temperature. The magnet materials are usually either permanent magnets or superconductors. Permanent magnets are mainly used for low-field NMR measurements; superconductors are usually used in medical imaging and chemical spectroscopy analysis of high-field instruments in the laboratory, requiring the use of liquid helium and liquid nitrogen to maintain a constant magnet temperature.

[0052] Radio frequency (RF) magnetic field B1 and the pulse: The RF pulse is an electromagnetic signal, usually generated by a coil. The magnetic field generated by the RF pulse is the RF magnetic field. The direction of the RF magnetic field is perpendicular to the direction of the static magnetic field, realizing the manipulation of the magnetization vector formed in the static magnetic field. The manipulation angle is: θ = γB1t p Where γ is the gyromagnetic ratio of the proton, B1 is the radio frequency magnetic field strength, and t p The duration of the radio frequency pulse is the key factor. Therefore, the spin angle can be changed by controlling the amplitude or duration of the radio frequency pulse. The nuclear magnetic resonance pulse sequence is composed of radio frequency pulses of different numbers and frequency attributes in a set sequence. By adjusting the time interval between pulses, the pulse angle, and the frequency selectivity of the pulse, the relaxation and diffusion of the spin system can be measured.

[0053] Magnetic Field Gradient and Imaging: The pulsed magnetic field gradient is generated by a gradient coil. During application, the eddy current effect between the pulsed gradient coil and the radio frequency coil is usually considered, and shielding effectiveness is crucial. By understanding the relationship between the spatial magnetic field strength and the gradient value, the sample under test can be encoded with corresponding spatial phase, frequency, and layer selection, enabling spatial imaging in different dimensions. For a specific direction in space, taking z as an example, after applying a gradient pulse of amplitude G in that direction, the proton Larmor frequencies at different spatial locations are:

[0054] ω(z)=γB0+γGz

[0055] Where γ is the gyrometry ratio of the proton, therefore, the relationship between the acquired echo signal M(k) and the spatially calculated imaging proton density ε(z) is as follows:

[0056] M(k)=∫ε(z)e i2πkz dz

[0057] ε(z)=∫M(k)e -i2πkz dk

[0058] Where k is a defined wave function, which is related to the parameters of the gradient pulse. When the system uses frequency coding mode for imaging experiments, k = γg max δ / 2π, as can be seen from the above formula, M(k) and ε(z) are Fourier transform pairs. Therefore, the imaging result can be obtained by performing a Fourier transform on the acquired echo signal.

[0059] Spin echo: Spin echo is one of the most common signals in NMR measurements. First, a 90° pulse is applied to the sample, rotating the magnetization vector M0 to a transverse plane perpendicular to the static magnetic field. Due to molecular diffusion and the spatial inhomogeneity of the static magnetic field, the magnetization vector M0 undergoes dephasing. If the signal acquisition channel is opened during this period, a free decay signal is obtained. After a certain time τ, a 180° pulse is applied, and the dephased magnetization vector will reconverge after the same time τ, forming an echo signal. This echo signal is called the spin echo signal. Spin echoes are observed in NMR measurements. The main applications are as follows: (1) By applying a series of 180° pulses, spin echoes are repeatedly formed and the echo train signal is recorded. This pulse sequence is the CPMG pulse sequence. This signal is extremely important for studying the transverse relaxation characteristics of porous media. Under certain conditions, information related to pore size can be obtained; (2) By changing the gradient amplitude or gradient duration under a gradient magnetic field, the change in the amplitude of the spin echo can be recorded, and the self-diffusion coefficient of fluid molecules can be obtained; (3) By applying paired frequency-coded or phase-coded gradients, the spatial spin density information of the sample under test can be analyzed to realize nuclear magnetic resonance imaging.

[0060] Relaxation: The process by which a spin system recovers from a resonant state to a thermal equilibrium state. This process is characterized by the longitudinal relaxation time T1 or the transverse relaxation time T2 in different directions. T1 is also called the spin-lattice relaxation time, reflecting the energy exchange between the spin system and the external environment, while T2 is called the spin-spin relaxation time, reflecting the energy loss within the spin system. The relaxation process of the spin system can be described by the Bloch equation. The longitudinal relaxation time T1 can be measured using a saturation recovery pulse sequence, by changing the time interval T between two pulses. W Record the signal amplitude to reflect the evolution of the longitudinal magnetization vector at different editing times:

[0061]

[0062] The above methods require a relatively long time. Each step, T... W Each of these processes requires the proton spin system to wait for a considerable period of time and reach thermal equilibrium before the next experiment can proceed, making the data collection process extremely slow.

[0063] The T1 imaging technology of this invention is based on a fast T1 measurement method, which uses a small-angle α radio frequency pulse train to acquire the final measurement result. By applying this pulse train containing N small-angle radio frequency pulses, the components of the magnetization vector of the spin system of the sample under test can be expressed as:

[0064]

[0065]

[0066] By employing a relevant adaptive RF pulse phase cycle and analyzing the amplitude of the acquired signal, the magnetization vector under each small-angle α RF pulse can be obtained as follows:

[0067]

[0068] Where N is the number of small-angle radio frequency pulses, and Δ is the time interval between two adjacent small-angle pulses. Compared to the conventional T1 measurement method, since the longitudinal magnetization vector is only operated on once, the above method can complete the T1 measurement in a shorter time.

[0069] Relaxation imaging. In practical measurement applications, it has been found that simply obtaining the proton density information of the sample, i.e., imaging, is far from sufficient for analyzing the microscopic information of the sample. Therefore, combining the acquisition of relaxation information with imaging technology can perfectly realize the observation of both the macroscopic and microscopic aspects of the sample.

[0070] The longitudinal relaxation time (T1) varies significantly among different biological tissues or samples, and is therefore often chosen as a weighted information method combined with imaging techniques. However, if conventional T1 measurement methods are used in conjunction with imaging, the actual measurement time is relatively long, which is not conducive to rapid dynamic observation of the sample's own information. Therefore, this invention integrates a rapid longitudinal relaxation time (T1) measurement method with relevant imaging techniques, and, through optimizing and adjusting relevant parameters, presents a feasible scheme for rapid longitudinal relaxation time (T1) imaging technology.

[0071] Based on the relationship between the longitudinal relaxation time T1 and the spatial gradient encoding of the measured sample, a design is made as follows: Figure 4 The fast MRI one-dimensional T1 imaging pulse sequence shown can be used to obtain the following response formula by acquiring the signal:

[0072] M(k Z , NΔ)=∫∫F(z, T1)·K1·K2dzdT1

[0073] The specific forms of the two kernel functions K1 and K2 are as follows:

[0074] K1=exp(i2πk Z z

[0075]

[0076] The frequency coding mode shown in this invention, k z =γGΔ / π. The obtained data is then processed using subsequent data inversion steps to obtain the one-dimensional T1 imaging results of the tested sample.

[0077] Based on the relationship between the longitudinal relaxation time T1, the transverse relaxation time T2, and the spatial gradient encoding of the measured sample, a design is made as follows: Figure 5The fast magnetic resonance imaging T1-T2 correlated pulse sequence shown can be used to obtain the following response formula by acquiring the signal:

[0078] M(k z , NΔ, Pτ)=∫∫F(z, T1, T2)·K1·k2·K3dzdT1dT2

[0079] The specific forms of the three kernel functions K1, K2, and K3 are as follows:

[0080] K1=exp(i2πk z z)

[0081]

[0082]

[0083] The frequency coding mode shown in this invention, k z =γGΔ / π. The obtained data is processed using subsequent data inversion steps to obtain the T1-T2 correlation imaging results of the tested sample.

[0084] Because the data acquired by the imaging method described in this application differs from that of conventional methods, special attention is required during actual data processing. This invention provides a corresponding data processing method for rapid T1-weighted magnetic resonance imaging (MRI).

[0085] A method for processing data acquired by a nuclear magnetic resonance (NMR) data acquisition system is described below.

[0086] Step 1: Determine the dimension of the acquired nuclear magnetic resonance data M.

[0087] Step 2: If M is two-dimensional data, the final result can be determined to be a one-dimensional imaging result. First, perform a Fourier transform on the data to decompile it in the imaging dimension. Perform a one-dimensional Inverse Laplace inversion on the decomputed data to obtain the T1 distribution at different spatial locations, which is considered as T1 imaging. The Fourier transform is a linear transform and is a non-ill-conditioned problem, so it will not be elaborated here. Since the Inverse Laplace inversion is an ill-conditioned problem, it will be briefly described here. A regularization term will be introduced here to invert this data matrix. To obtain a stable and accurate solution F, the Tikhonov regularization method is usually used, introducing a smoothing term to solve the problem:

[0088]

[0089] Here, s is the regularization factor, related to the signal-to-noise ratio of the collected data, and the ||·|| term represents the Frobenius norm of the matrix. The introduced regularization term determines the stability and accuracy of the solution. If the regularization factor is too large, although the obtained distribution is more stable, the accuracy of the solution is worse, i.e., oversmoothing; if the regularization factor is too small, the solution is more accurate, but the stability of the solution decreases, and more spurious signals appear, i.e., undersmoothing. Therefore, considering both the authenticity and stability of the solution, using a reasonable regularization factor is the key to this method. The non-negative constraint solution F under a specific regularization factor s can be obtained through the non-negative constraint step.

[0090] Step 3: If M is three-dimensional data, the final result may be a two-dimensional imaging result. First, perform a Fourier transform on the data to decompile it in the imaging dimension. Perform a two-dimensional Inverse Laplace inversion on the decomputed data to obtain the T1-T2 distribution at different spatial locations, which is considered the T1-T2 associated imaging result. The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. A method for rapid nuclear magnetic resonance T1-T2 correlation imaging, characterized in that, Includes the following steps: Step 1: In the GRD channel, due to the presence of the additional magnetic field unit, an additional magnetic field with a specific magnetic field gradient G will be formed along the z direction. Step 2: Apply a 90° radio frequency pulse to the hydrogen-containing proton spin system of the sample under test in the TRS channel to obtain the macroscopic magnetization vector. M 0. Rotate to the horizontal plane; Step 3, wait for a very short time τ Then, a 180° radio frequency pulse is applied to the hydrogen-containing proton spin system of the sample under test on the TRS channel to reassemble the transverse planar magnetization vector after the phase separation. Step 4: Wait a very short time again. τ Then, a 90° radio frequency pulse is applied to the hydrogen proton spin system of the sample under test on the TRS channel, which rotates the re-aggregated transverse planar magnetization vector by 90° to the longitudinal axis, which is consistent with the direction of the static magnetic field. Step 5: Next, apply a small angle to the TRS channel. The radio frequency pulse then undergoes a frequency-coded gradient magnetic field on the GRD channel, wherein the height of the frequency-coded gradient pulse is fixed at G; Step 6: Repeatedly apply 180° radio frequency pulses to the TRS channel. After a certain period of time, acquire a complete discrete spin echo signal in the ACQ channel. When starting to acquire the spin echo signal, first apply a frequency decoding gradient pulse to the GRD channel. The amplitude of the frequency decoding gradient pulse is the same as that of the frequency encoding gradient pulse in Step 5. Step 7: Apply another 180° RF pulse to the TRS channel to flip the magnetization vector remaining in the longitudinal direction; Step 8: Apply a small angle to the TRS channel as described in step 5. The entire timing duration, from the start of the RF pulse to the application of the last constant-amplitude gradient on the GRD channel, is [duration missing]. The timing sequence is continuously repeated N times, and N spin echo signals will eventually be acquired in the ACQ channel. Step 9: Change the waiting time in steps 3 and 4. τ The length of the time interval is adjusted, and the transverse relaxation time T2 is edited accordingly. The timing sequence is continuously looped P times, and P*N spin echo signals will eventually be acquired in the ACQ channel. Obtain echo train signal , The wave function is defined; for the acquired echo train signal By processing the MRI data, rapid MRI T1-T2 correlation imaging results can be obtained.

2. The method for rapid T1-T2 correlation imaging with nuclear magnetic resonance imaging according to claim 1, characterized in that, The nuclear magnetic resonance data processing method in step 9 is as follows: Step 9.1, for Perform a Fourier transform to decompile the data along the imaging dimension; Step 9.2: Perform two-dimensional Inverse Laplace inversion on the obtained decoded data to obtain the final T1-T2 correlation distribution at different spatial locations, which is regarded as the T1-T2 correlation imaging result.

Images