Magnetic resonance imaging with zero echo time and slice selection

Abstract

This invention provides a novel radio frequency sequence suitable for performing magnetic resonance imaging (MRI) of two-dimensional (2D) slices (rather than three-dimensional volumes) of samples exhibiting short magnetization coherence times (i.e., hard tissue). The SS-ZTE pulse sequence includes the following steps: a) magnetizing all spins in the sample from the longitudinal direction to the transverse plane; b) exciting the 2D slice of interest, including selective locking of the magnetization of the 2D sample slice while perturbing the magnetization in the remainder of the sample volume; c) ensuring that the magnetization of the selected 2D slice is unaffected by the reconfiguration of the magnetic field gradient from slice selection to encoding and readout; d) reading out the free-induction decay signal of the sample; e) repeating steps (a to d) in different readout directions to collect a corresponding number of radial spokes of the plane defined by the 2D sample slice.

Description

Technical Field

[0001] This invention relates to the field of magnetic resonance imaging (MRI), which is based on nuclear excitation of the nuclear spins of a subject using radio frequency (RF) signals, and the reconstruction of an image of the subject as a result of the MRI signal generated by this excitation. More specifically, this invention relates to a zero-echo time (ZTE) imaging method that provides two-dimensional (2D) slice selection for rapid MRI of samples with short magnetization coherence times. Background Technology

[0002] In MRI, imaging protocols rely on the excitation and detection of the spin degrees of freedom of nucleons in a sample or object of study. When subjected to an external magnetic field in the longitudinal direction, these nucleons possess magnetic energy (proportional to the field strength) and dipole moments (which tend to align with the lines of the external magnetic field). If a radio frequency (RF) pulse with a 90° orientation relative to the magnetic field is applied to the sample, the net magnetization tilts downward, causing the longitudinal magnetization to disappear and the transverse magnetization to appear. Following this pulse, the transverse component of the magnetization precesses at a Larmor frequency, and shortly thereafter, an alternating current is induced that can be detected in a coil. The induced signal (e.g., detected by an RF receiver) (more preferably called the free induction decay (FID) signal) decays with a transverse relaxation time constant (known as T²*). Sample temperature causes magnetic dipole fluctuations, which constitute a noisy environment for the surrounding spins. Due to this and other interactions, magnetization coherence (also known as spin coherence) decays exponentially according to T²*, thus setting an upper limit on the time before sample information can be obtained before it must be re-excited with RF radiation. To recap, in the classical sense, magnetization coherence in MRI is the physical state when sample spins align and rotate at the same velocity around the direction of the magnetic field. In quantum mechanics, coherence is the state in which spins oscillate between different energy states (eigenstates) with the same velocity or phase. This coherence is disrupted by intrinsic processes of material properties over time T2 (such as dipole-dipole interactions and molecular tumbling), and by inhomogeneities of the main magnetic field B0, improper configuration of B0 or ​​its gradient, or the susceptibility of boundaries between tissues. The latter effect also contributes to dephasing, resulting in a total (extrinsic and intrinsic) T2* time (which is the sum of both effects). There is also magnetization relaxation in the longitudinal direction, described by another time constant (called T1).

[0003] In many cases (including the vast majority of clinical applications of MRI), these spin-spin interactions are primarily dipole-dipole type, and their strength depends largely on the angle θ between the line connecting the dipoles and the direction they point. In non-solid samples, θ changes continuously due to molecular motion, thus averaging isotropically, suppressing coupling between adjacent nucleons and resulting in the strong signal used by MRI scanners for image reconstruction. However, this averaging effect does not occur in solids because both nucleons and magnetic fields are static in the laboratory coordinate system.

[0004] Therefore, MRI imaging of hard biological tissues remains technically challenging. In particular, the penetration of MRI in dental clinical applications is extremely limited, primarily due to the extremely short signal response of dentin and enamel (the main components of human teeth) to the MRI protocol. MRI imaging of samples with short-axis T2 requires spatial coding and data acquisition to begin and be completed rapidly after signal generation.

[0005] When describing MRI RF sequences, two main parameters are used: echo time (TE, which is correlated with the time between the application of the RF excitation pulse and the signal peaks included in the RF receiver), and repetition time (TR, measured from the application of the excitation pulse until the application of the next excitation pulse). TR determines the degree of longitudinal magnetization recovered between each pulse, while TE determines the transverse (T2*) relaxation.

[0006] To read fleeting signals from the hardest tissues, two families of rf pulse sequences were employed: SWIFT imaging with Fourier Transform (see, for example, M. Weiger et al., “MRI with zero echo time: Hard Versus Sweep Pulse Excitation,” Magnetic Resonance in Medicine, 66 (2011), 379–89) and zero echo time (ZTE) with hard rf pulse excitation (i.e., pulses with a rectangular form). The ZTE sequence (see M. Weiger et al., “Sweep MRI with Algebraic Reconstruction,” Magnetic Resonance in Medicine, 64 (2010), 1685–95) is a robust, fast, and quiet method for three-dimensional (3D) imaging of short T2 samples. Both of the aforementioned pulse families rely on opening the encoded magnetic inhomogeneities (gradients) before the rf excitation pulse. Furthermore, they present engineering challenges: the former requires simultaneous RF transmission and reception (thus, the SWIFT sequence introduces an initial acquisition dead time), while the latter requires a high-power RF transmission line to achieve sufficient excitation bandwidth. Additionally, SWIFT and ZTE share a common property: they are essentially volumetric sequences. In fact, they encode spatial information of a 3D field of view (FoV), and mathematical algorithms can then be used to reconstruct 3D objects.

[0007] Given the rapid attenuation of MRI signals in hard tissue imaging, rapid signal encoding and acquisition are essential, where there is no dead time between excitation and MRI signal recording. One possible option for addressing these requirements is to utilize 3D radial center-outward k-space encoding and apply an rf ZTE sequence, where the k-space is an array representing the spatial frequencies in the acquired MRI image, obtained, for example, by applying a Fourier transform to the MRI image. The center frequency corresponds to the image contrast, making the recovery of k-values ​​around this point crucial. The goal of MRI is to acquire the maximum amount of k-space (frequency space) content to reliably reconstruct the 3D sample. It is worth noting that the ZTE technique minimizes the dead time between excitation and acquisition, thereby minimizing the gap at the center of the k-space. In particular, the Pointwise Encoding Time Reduction with Radial Acquisition PETRA technique (see, for example, M. Grodzki et al., “Ultrashort Echo Time Imaging Using Pointwise Encoding Time Reduction with Radial Acquisition (PETRA)”, Magnetic Resonance in Medicine, 67 (2012), 510-18) can overcome this limitation and fill the remaining dead time gaps in the k-space pointwise.

[0008] Other researchers have conducted preliminary studies aiming to bring spatial or slice-selective capabilities to MRI methods. For example, various types of spin-locking (SL) sequences have been published in R.A. Wind et al., “Spatial Selection in NMR by Spin-Locking,” Journal of Physics C, 11 (1978), L223-26. Specifically, SL begins at conventional resonance (π / 2). x’ A pulse is applied to magnetize the sample along its initial longitudinal direction (along the z-axis) in the direction x', which precesses at a Larmor frequency in the laboratory coordinate system; and is therefore static in the rotating coordinates of the resonant RF excitation. Once magnetization is directed along the -y' direction (which is always orthogonal to both x' and z), the SL involves the application of a longer locking pulse around ±y'. If linear magnetic inhomogeneity is applied during this pulse, the locking effect occurs only for sample slices perpendicular to the magnetic gradient, to which the RF excitation is resonant. During the SL pulse, the remaining spins of the sample's 3D FoV are dephased (losing their coherence); and slice selection can be performed in this manner.

[0009] The references contain numerous MRI sequences defined by different combinations of rf excitation and gradient pulses, which are selected to be optimally suited for specific applications. In standard MRI sequences, the Delay Alternating with Nutation for Tailored Excitation (DANTE) pulse sequence provides a combination of narrow-bandwidth coherent rotation and high-bandwidth pulses. To make this approach compatible with short T2 samples, DANTE pulses can be combined with dipole-decoupled pulses based on MREV sequences (90° pulses, which eliminate the effects of dipole interactions at the sequence ends), for example, as in DGCory et al., “DANTE Slice Selection for Solid-State NMR Imaging,” Journal of Magnetic Resonance (1969), 90 (1990), 544-50.

[0010] Another example is the application of magnetization storage (MS pulses) to transfer transverse magnetization (which undergoes decoherence and T2* decay) to the longitudinal dimension z, where it is protected from decoherence and undergoes only slow relaxation characterized by another time constant T1 much larger than T2*. In the MS pulse (which can be simple (π / 2)... -x’ After that, if the magnetization is along the -y' direction, there is no spin precession and therefore no signal is sensed on the scanner detector. However, during this time, the magnetic field gradient can be turned on and off without affecting the spin phase within the slice (which is meaningless when it is magnetized longitudinally) and can be used to de-phase the coherence outside the slice.

[0011] In the field of MRI, a special pulse sequence has recently been disclosed, designed and used to simultaneously remove dipole-dipole interactions (thus prolonging the coherence time of transient signals) and inhomogeneous phase loss. This sequence is also known as Combined Hahn and Solid Echo (CHASE), see AMWaeber et al., “PulseControl Protocols for Preserving Coherence in Dipolar-Coupled Nuclear SpinBaths,” Nature Communications, 10 (2019), 1–9. Its simplest form, CHASE-5, combines four solid echo pulses based on the WAHUHA and MREV sequences with a single 180° pulse to dynamically rephase contributions, for example, from undesirable magnetic field inhomogeneities. These five pulses in the CHASE-5 block can be considered as a single sub-block of a larger CHASE pulse block. For example, the first CHASE-5 sub-block {(π / 2)-x,(π / 2)y,πx,(π / 2)y,(π / 2)x} is followed by the second CHASE-5 sub-block {(π / 2)x,(π / 2)-y,π-x,(π / 2)-y,(π / 2)-x} to form a CHASE-10 pulse block, which corrects for finite pulse durations (CHASE-5 requires pulses much shorter than T2). Even longer pulse blocks can be designed to incorporate other effects.

[0012] While volumetric (3D) encoding is convenient in some cases, excitation and acquisition of images of 2D slices of a sample (slice selection) is significantly faster. Currently, there are no pulse sequences specifically designed for 2D slice selection of short T2 materials or tissues. Due to the insufficient rf bandwidth used to excite the full 3D FoV in ZTE, the aforementioned sequences SWIFT and ZTE are considered necessarily 3D, and the observed slice selection process is merely unintentional and harmful. The objective of this invention is to solve the aforementioned problems of 2D slice selection (SS) in MRI, and more particularly, to image samples with extremely short T2 (less than 1 ms, i.e., hard tissue). Summary of the Invention

[0013] As outlined in previous chapters, MRI image reconstruction of short T2 materials (such as hard tissue) is only possible for 3D FoV. This invention overcomes the aforementioned limitations of the prior art by providing a method for 2D FoV (slices) based on a specifically designed ZTE pulse sequence (hereinafter referred to as SliceSelective Zero Time Echo (SS-ZTE)).

[0014] In a first aspect, the present invention provides an MRI method for imaging 2D slices of a sample placed within a magnetic field (B0) oriented along a longitudinal direction, such that the magnetization of the sample is initially parallel to said direction of the magnetic field. This method is characterized by performing an imaging scan of the sample based on a ZTE sequence, wherein the imaging scan comprises four steps for a single radial spoke (considering the radially encoded direction in k-space, as in standard ZTE):

[0015] a) The excitation step includes activating the slice selection gradient pulse (G) SS (It is responsible for selecting the 2D slice of the sample to be imaged), followed by at least one rf excitation pulse (B1) (which provides a coherent rotation of the magnetization of all spins in the sample toward a plane transverse to B0), followed by an initiation slice selection gradient pulse (G). SS (It is responsible for selecting the 2D slices of the sample to be imaged) after that.

[0016] b) The slice selection step involves selectively locking the sample magnetization such that only selected 2D slices remain excited within the FoV. Thus, in this step, a sequence of at least one rf pulse is applied to the selective locking (SL) of the sample magnetization.

[0017] c) The holding step includes several rf and magnetic pulses designed to maintain the magnetization coherence of the 2D slice, while selecting gradient G. SS Turn off and image encoding gradient G ro Open, the magnetization coherence of 2D slices is not affected by subsequent G SS and G ro The impact of reconfiguration.

[0018] d) The final step, including the acquisition of FID signals, is directed to G... ro Each defined radial spoke provides a code.

[0019] Steps a) through d) are performed sequentially along several reading directions to collect the corresponding number of radial spokes in the k-space plane, where only 2D slices of the sample contribute to the acquired signal. Finally, all the data collected from the magnetic FID signal and arranged in k-space is used together with mathematical tools (e.g., inverse Fourier transform) to reconstruct an image of the 2D slices of the sample.

[0020] Specific implementations of slice selection (SS) step b) may include spin-locked pulses or DANTE sequences, which combine excitation and slice selection into a combined sequence. Specific implementations of hold step c) may be based on magnetized storage pulses or CHASE sequences. Some of these steps are briefly described in the following paragraphs.

[0021] In a particular embodiment of the invention, slice selection step b) may be implemented as a standard spin-locked pulse (SL).

[0022] Another specific embodiment of the invention includes a slice selection step (b) with at least one rotating echo SL pulse. These sequences are divided into two segments with equal durations but opposite phases. The advantage of using them is that they compensate for the inhomogeneity of B1, which can induce artifacts in MRI reconstruction.

[0023] In another embodiment of the invention, the excitation step a) and the slice selection step b) can be combined into a unique homonuclear dipolar-decoupled version of the DANTE selection-excitation sequence. To recap, the DANTE sequence consists of a periodic alternation of free spin evolution and short, hard RF pulses (i.e., rectangular pulses), making its implementation in an RF emitter relatively simple. These sequences enable slice selection of 2D slices of the sample while maintaining the coherence of its magnetization.

[0024] In other embodiments of the invention, the holding pulse sequence in step c) begins with a hard rf pulse to rotate the transverse component of the magnetization to the longitudinal axis. Once magnetized longitudinally, G... SS The gradient decreases to zero and G ro Reduced in the vertical (image encoding) direction. G ro The presence of [something] allows for the reading process in the acquisition step, and also disrupts the remaining coherence beyond 2D slice d). Finally, due to G ro At its target value, an excitation RF pulse is applied to the sample, thereby net magnetizing it at a specific angle (better called the flip angle (θ)). x The rotation is oriented laterally. This angle is typically used to define the excitation angle of a field echo pulse sequence. In practice, θ... x The angle by which the net sample is magnetized relative to the direction of the main magnetic field is rotated or tilted by applying an rf excitation pulse at the Larmor frequency.

[0025] An alternative embodiment of the invention includes holding step c) performed together with the MS pulse as described in the preceding paragraphs, but includes an additional turbulence gradient pulse; this additional turbulence gradient pulse in G SS Termination and G ro Alternating between starts, an increased disturbance is responsible for the residual magnetization outside the 2D slice, which has not yet been completely lost during the large SL pulse in b).

[0026] Additional embodiments of the objective of this invention include SS-ZTE sequences, wherein the holding step c) comprises a CHASE sequence. More specifically, the CHASE sequence may be a single CHASE-5 pulse train, wherein the slice selection gradient G SS It linearly decreases to zero before the first 90° pulse and increases after the last pulse, and the gradient (G) is read. ro First rise and then fall (with G) SS On the contrary, as in Figure 4 In this process, the holding step attempts to maintain the magnetization coherence of the 2D slice and keep it in the lateral plane while simultaneously turning the encoding (slice selection) gradient on (off). Because the magnetization is already lateral, encoding can begin immediately without dead time. More complex CHASE sequences (e.g., CHASE-10) can be used for further improvement of magnetization coherence. Additionally, G... SS and G ro The complexity is reduced, but it must be emphasized that the corresponding time interval is compensated along the CHASE sequence in each of the quantum spin directions, so that the 2D selected slice does not experience gradient dephase after the CHASE sequence.

[0027] Another particularly convenient embodiment of the objective of this invention includes maintaining step c) performed together with the MS pulse, and acquisition step d) following the ZTE scheme. This is because the MS step is followed by an excitation pulse that rotates the 2D slice to a flip angle θ. x Therefore, a dead time exists. Thus, PETRA sequences can also be used to provide better image reconstruction. This combination addresses the dead time problem, which is associated with the time required for RF electronics to switch between transmit and receive modes. This dead time hinders the acquisition of a specific range of data around the k-space center; however, PETRA is able to overcome this limitation.

[0028] Another embodiment of the objective of this invention comprises a holding step (c) involving the CHASE sequence, such that the dead time can be incorporated into the waiting time between pulses (τ), thereby allowing for a further reduction in acquisition time, which is particularly critical for hard tissue imaging. Therefore, this embodiment can scan k-space without gaps at the center.

[0029] Another object of the present invention relates to an MRI device, comprising:

[0030] —A magnet, operable to provide a magnetic field (B0);

[0031] — An RF transmitter configured to emit an RF field (B1) onto a sample placed in a magnetic field;

[0032] —RF receiver, configured to receive magnetic resonance signals;

[0033] —Data acquisition unit, recording magnetic resonance signals; and

[0034] — An instrument used to process information provided by the data acquisition unit;

[0035] Advantageously, the rf emitter is configured to generate a ZTE sequence according to any of the described embodiments, such that only 2D slices of the sample are magnetized using the applied rf field.

[0036] It is also possible to have an RF transmitter and receiver integrated into the same physical element (RF coil).

[0037] Another objective of the present invention is to apply MRI pulse sequences to MRI imaging of the maxillofacial region for clinical dental applications, or to imaging of bones or tendons for physiological therapeutic applications, according to any of the preceding embodiments (particularly useful for which is the MRI technique, which enables imaging of samples exhibiting extremely short T2 times).

[0038] The ultimate objective of this invention relates to imaging solid samples using MRI pulse sequences according to any of the preceding embodiments, which is particularly useful in any of the following fields: archaeology, mineralogy / gemology, soil exploration, analysis of precious minerals, and analysis of the chemical composition of solids. Attached Figure Description

[0039] To complete the description and to provide a better understanding of the invention, a set of accompanying drawings is provided. These drawings form part of the overall description and illustrate embodiments of the invention, which should not be construed as limiting the scope of the invention, but rather as examples of how the invention can be practiced. The drawings specifically include the following figures:

[0040] Figure 1 The diagram illustrates an SS-ZTE sequence according to an embodiment of the present invention and includes a sample radio frequency (RF) pulse, a slice selection (G)... SS Gradient pulses and readouts (G) ro Gradient pulse.

[0041] Figure 2 It shows Figure 1 Two possible embodiments of the slice selection (SS) step are shown: naked selection lock (SL) pulse (A) or rotating echo SL pulse (B), the latter being insensitive to inhomogeneities of the rf excitation pulse (B1) and suppressing phase loss within the slice.

[0042] Figure 3 It shows Figure 1 The possible embodiment of the SS-ZTE sequence shown includes a hold (P) step comprising a magnetization storage pulse, a gradient perturbation pulse, and an excitation pulse.

[0043] Figure 4 It shows Figure 1 A possible embodiment of the SS-ZTE sequence is shown, wherein the hold (P) step includes a CHASE pulse train of length 5, thereby allowing zero dead time before data acquisition.

[0044] Figure 5 The slice selection (SS) step is shown to be performed in order to apply a bare spin-locked pulse, wherein the sample consists of a test tube filled with copper sulfate solution.

[0045] Figure 6 It shows having with Figure 5 The slice selection (SS) step is performed with the same sample and conditions, but the bare spin-lock pulse is replaced with a rotating echo spin-lock pulse. Detailed Implementation

[0046] Figure 1 To illustrate the block diagram of the SS-ZTE pulse sequence according to an embodiment of the present invention, it includes four main steps: an excitation (E) step, in which the full 3D FoV magnetization is rotated to the transverse plane; a slice selection (SS) step, in which the sample magnetization is selectively locked / spoiled to obtain selected slices for excitation individually; a hold (P) step, designed to ensure that the magnetization and coherence of the selected slices are unaffected by the reconfiguration of the magnetic field gradient; and finally, an acquisition (A) step, in which the FID signal is detected, recorded, and discretized. Each execution of the above process (each iteration includes one execution of all steps E, SS, P, and A) provides data for a single radial spoke in 2D k-space. For this purpose, the process is performed with multiple readouts (G ro The process is repeated in the direction of the sample selection to collect the corresponding number of radial spokes in the k-space, where the signal contribution is derived only from selected 2D slices of the sample. Mathematical tools are then used to reconstruct the sample image from the acquired data. Without loss of generality, x can be chosen as the direction of the slice selection gradient, such that, for example, only the 2D slices of the sample at x=0 remain coherently magnetized. The process is then repeated with the readout gradient G applied to the yz plane. ro When applied together, the k-space was obtained. y -k z The radial spokes in the image are then used. Mathematical tools are then able to reconstruct the image of the selected 2D slice in the yz direction.

[0047] In a first aspect, the present invention is characterized by an excitation step similar to that of standard ZTE technology, i.e., hard rf excitation occurs only in the gradient field (G roAfter activation, pulsed magnetization is performed. The RF pulse in the excitation step must coordinately rotate the magnetization 90° from the longitudinal direction (z) to the axis in the transverse plane for use in the subsequent slice selection (SS) step. Therefore, it is not necessary to determine the flip angle θ. x This is explained below in the description of the hold step (P). It is worth emphasizing an important difference between SS-ZTE and regular ZTE sequences: during the excitation step (E), the slice selection gradient field (G) is... SS The direction of the slice ultimately determines the slice selected by the former, rather than the readout direction of the latter.

[0048] The purpose of the slice selection (SS) step is to selectively lock / disrupt sample magnetization, thereby leaving only 2D sample slices with selectable coherence.

[0049] In a preferred embodiment of the invention, a slice selection (SS) step is performed on the bare (standard) spin-locked (SL) pulse, such as... Figure 2 As shown in Figure A, the magnetization in the FoV slice where the SL pulse resonates will be locked. The spins in the remaining FoVs are excited non-resonantly and thus undergo dephase at a rate T2* due to the slice selection gradient field (G). SS The presence of ) accelerates the process. Without loss of generality, considering that the excitation step results in magnetization along the -y' direction, the frequency (ω) SS ) and Rabi frequency (Ω) SS ) bare SL pulses around -y' (see Figure 2 A) Locking in a Larmor frequency ω L (pass From ω SS The detuned spin simultaneously disrupts (i.e., dephases and diffuses in the transverse plane) the remaining spins.

[0050] In another preferred embodiment of the invention, such as Figure 2 As shown in B, the slice selection (SS) step involves using a rotating echo SL pulse, where the phase changes from the first half to the second half to move the rotation axis from the -y' direction to the y' direction. This makes the SS step insensitive to the inhomogeneity of the rf field (B1), i.e., for the rf field amplitude and therefore Ω. SS It is insensitive to spatial variations. This also partially corrects for the incomplete correlation with ω. SS Spin dephase in resonant 2D slices improves slice quality.

[0051] In a particularly advantageous embodiment of the invention, the excitation (E) and slice selection (SS) steps are combined into a unique homonuclear dipole decoupling type of DANTE selection-excitation sequence. This technique allows WAHUHA or other decoupling pulse sequences to alternate between DANTE excitations of only 2D resonant slices of the sample.

[0052] Meanwhile, a major challenge overcome by SS-ZTE is how to switch from a gradient configuration (which allows slice selection (out-of-plane)) to an orthogonal configuration (where k-space radial spokes can be sampled (in-plane)) while preserving the magnetization coherence of short T2 samples. This task is performed through a hold (P) step, which ensures that the magnetization and coherence of the selected slices are unaffected by the reconfiguration of the magnetic field gradient. At its broadest extent, the hold (P) step involves a combination of rf and gradient pulses, which ensures that only the selected 2D slices of the sample will contribute to the detected FID signal.

[0053] In a preferred embodiment, the hold (P) step begins with a hard rf pulse that rotates the transverse component of the magnetization toward the longitudinal axis z. This MS pulse can be a simple 90° rotation about -x' (90°...). -x’ Once magnetized longitudinally, the slice selects the gradient (G). SS The gradient (G) was reduced to zero without dephasing the spins within the slice, but the remaining coherence of the spins outside the slice was dephased. ro (relative to G) SS The magnetization is reduced in the vertical (image encoding) direction to allow for subsequent acquisition in the corresponding step. Afterward, an excitation RF pulse rotates the magnetization toward the transverse plane by an angle θ, selectable other than 90°. x (Flip angle). This final step is exactly the same as a regular ZTE.

[0054] Another preferred embodiment of the present invention ( Figure 3 The RT for the SS-ZTE sequence (shown) includes a hold (P) step, which includes a gradient perturbation pulse (G). ro,spoil The gradient perturbation pulse is selected in the slice gradient pulse (G). SS Termination and reading gradient (G) ro The alternation between these phases effectively further perturbs (dephases) the spin contributions outside the sample that have not yet completely lost their quantum coherence during the long SL pulse of the slice selection (SS) step. This can occur, for example, when the sample comprises a combination of tissues with both short and long coherence times, or, for example, when the SL pulse is intentionally shortened to limit the rf absorption of the sample.

[0055] In another preferred embodiment, the key factor in the hold (P) step is the combination of the Hahn and solid-state echo (CHASE) sequences; wherein GSS Start and G ro The process stops, and both follow a trajectory in such a way that, after they have switched, the contribution of their time integrals to the non-uniform extension term in the Hamiltonian is exactly the same before and after the 180° pulse in the CHASE sub-step; that is, the time integral must cancel out for each spin operator. In this case, the 180° pulse suppresses the phase loss otherwise introduced by the dynamic gradient field, and the 90° pulse train eliminates the contribution of the dipole interaction. All 180° and 90° pulses must be quasi-instantaneous (much shorter than the duration of T2), and the final flip angle of CHASE is 90°.

[0056] In another preferred embodiment of the invention, such as Figure 4 As summarized, the CHASE sequence is a single CHASE-5 pulse train, and G SS It decreases linearly before the first 90° pulse and increases after the last pulse, while G ro First it decreases and then it rises. Here, region A... SS and A ro It must be the same in terms of both decline and rise, but A SS Not necessarily equivalent to A ro .

[0057] Finally, after the gradient selection in the gradient start and hold (P) step ends, FID signal detection and data acquisition are performed in the acquisition (A) step. Figure 3 and Figure 4 (The text is incomplete and appears to be a fragment of a larger document. A more accurate translation would require the full context.) Figure 3 In the CHASE embodiment, this td coincides with the start of the acquisition (A) step, thus encoding directly from k=0 in the k-space.

[0058] In a preferred embodiment of the invention, the holding (P) step is... Figure 3 The MS pulse implementation in the process involves a dead time (td) long enough to switch the RF electronics from transmit to receive mode, leaving an unsampled gap at the center of the k-space; if td is small, then a ZTE-coded sequence can be used. If it is too large, then the acquisition (A) step may include point-based k-space center sampling (similar to the standard PETRA technique).

[0059] Another particularly advantageous embodiment of the hold (P) step is implemented via a CHASE pulse, such as Figure 4 As shown, this allows td to be incorporated into the waiting time between pulses (τ), thus leaving no gap at the center of the k-space.

[0060] The following paragraphs describe some experimental results obtained through different embodiments of the present invention.

[0061] To demonstrate the slice selection capability of SS-ZTE sequences, Figure 5 An experimental demonstration was shown, in which the test tube was filled with a solution of copper sulfate as a sample, and the reference magnetic field B0≈0.3T and Larmor frequency ω were set. L ≈2π·14MHz. If the acquisition (A) step is performed immediately after the excitation (E) step (where the reading is along the slice selection direction (G)). ro =G SS If approximately 50 mT / m is executed, then the reconstructed one-dimensional (1D) profile (measured as spin density) will represent the full width of the test tube (≈9 mm, see [reference]). Figure 5 (The broken line A in the diagram). However, considering Ω SS A bare SL pulse of approximately 2π·2.5kHz and a duration equivalent to the 500μs included in the slice selection (SS) step shrinks the 1D line profile to a width of 1mm (see [link]). Figure 5 (B in the broken line).

[0062] at last, Figure 6 Another example of an SS-ZTE application is shown here. (Repeated here) Figure 5 Previous experiments, but with rotating echo SL pulses instead of bare spin-locked pulses. Similarly, Figure 6 The broken line A in the diagram represents the full width of the test tube, while the 1D line profile boundary of the selected slice ( Figure 6 The broken line (B1) in the diagram becomes clearer in this case because the effect of the inhomogeneity of the applied rf field (B1) is suppressed by the rotating echo configuration.

[0063] Additional advantages and modifications will readily emerge for those skilled in the art. Therefore, the invention is not limited in its broader aspects to the specific details and representative embodiments shown and described herein. Consequently, various modifications may be made without departing from the spirit or scope of the general inventive concept and its equivalents as defined by the appended claims.

Claims

1. A magnetic resonance imaging (MRI) method, comprising placing a sample along a longitudinal axis in a magnetic field B0, such that the magnetization of the sample is initially parallel to the direction of the magnetic field. in, The method includes performing an imaging scan of a sample based on a ZTE sequence, wherein the imaging scan includes the following steps: a) applying a slice selection gradient pulse G SS and applying at least one radio frequency excitation pulse, which generates a transverse magnetization of the sample with respect to the B0 direction; b) Apply a slice-select pulse sequence, the slice-select pulse sequence comprising at least one rf pulse Ω. SS To selectively lock the sample magnetization corresponding to a single 2D slice of the sample, wherein the slice selection pulse sequence includes at least one rotating echo SL pulse; c) Apply a hold pulse sequence comprising at least one rf pulse to maintain the magnetization coherence of the selected 2D slice, and additionally include G... SS Termination and reading gradient pulse G ro The beginning; d) Execution by G ro Encoding, reading, and acquisition of the FID signal of the radial spokes of the defined sample; e) Repeat steps a) to d) along multiple reading directions to collect the corresponding number of radial spokes in the k-space plane corresponding to the 2D slices of the sample; f) Process the FID signals collected in steps a) to e) and reconstruct images of 2D slices of the sample.

2. The MRI method of claim 1, wherein the hold pulse sequence begins with a hard rf pulse that rotates the transverse component of the magnetization to the longitudinal axis, and once magnetized longitudinally, G... SS The gradient is reduced to zero and the gradient G is read. ro The image is lowered in the vertical image encoding direction; then an RF pulse is excited to magnetize and rotate the sample toward the lateral plane.

3. The MRI method according to claim 2, wherein the hold pulse sequence includes G SS Termination and G ro An additional perturbation gradient pulse alternates between the start and end points; the additional perturbation gradient rises and then falls to further disrupt the magnetization coherence outside the selected 2D slice.

4. The MRI method according to claim 1, wherein the hold pulse sequence comprises a CHASE sequence.

5. The MRI method of claim 4, wherein the CHASE sequence is a single CHASE-5 pulse train; and furthermore, G SS The gradient is linearly decreased before the first 90° pulse, and a short gradient pulse is applied after the last pulse, while G... ro First, a short gradient pulse is applied, then it rises linearly.

6. The MRI method according to claim 2, comprising a dead time td prior to acquisition; and an acquisition procedure following the ZTE protocol.

7. The MRI method of claim 4, wherein the dead time td is incorporated into the waiting time between pulses of the hold pulse sequence; and the acquisition steps follow the ZTE scheme.

8. The MRI method according to any one of the preceding claims, wherein the sample comprises hard tissue with a T2 of less than 1 ms.

9. The MRI method according to claim 8, wherein the hard tissue includes bone tissue, dental tissue and / or tendon tissue.

10. An MRI device, comprising: A magnet, operable to provide a magnetic field; An RF transmitter is configured to emit an RF field toward a sample placed in the magnetic field. An RF receiver is configured to receive magnetic resonance signals; The data acquisition unit records the magnetic resonance signal; and An instrument for processing information provided by the data acquisition unit; The rf transmitter is configured to generate a ZTE sequence in accordance with any one of claims 1 to 9.

11. The MRI device of claim 10, wherein the RF transmitter and the RF receiver are combined into the same physical element.

12. The MRI device according to any one of claims 10 to 11, for imaging of dental bone, hard tissue, archaeological objects, mineralogical imaging analysis, gem authentication imaging, soil exploration imaging, and imaging analysis of solid chemical composition.

Images