A method for fat-suppressed coronary imaging based on steady-state free precession sequence

Abstract

The application discloses a kind of fat suppression coronary artery imaging methods based on steady-state free precession sequence, it is related to magnetic resonance imaging technical field;Including the following steps: reconstruct bSSFP sequence using short-time exponential decay variable flip angle radio frequency excitation mode;Determine the optimal parameter in exponential decay variable flip angle radio frequency pulse model by simulation experiment;Carry out phantom verification to determine the best zero-crossing point position;According to the optimal parameter and the best zero-crossing point position determined above, human body scanning imaging is carried out.By directly using the inherent characteristics of exponential decay variable flip angle mode, the signal is quickly smoothed into steady state, without using additional preparation pulses to let the magnetization vector into steady state, avoiding residual image artifacts caused by signal oscillation;In addition, the characteristic that the off-resonance magnetization vector exists zero-crossing point in this mode can be directly used to realize robust fat suppression effect.No additional design is needed for frequency-selective fat suppression pulse, and the problem of insufficient fat suppression effect is also avoided.

Description

Technical Field

[0001] This invention relates to the field of magnetic resonance imaging technology, specifically to a fat-suppressed coronary artery imaging method based on steady-state free precession sequences. Background Technology

[0002] In the field of magnetic resonance imaging (MRI), the steady-state free precession (bSSFP) sequence is a widely used fundamental sequence. Because this sequence can acquire images with high signal-to-noise ratio and high tissue contrast within a very short repetition time (TR) (typically <5ms), it is well-suited for applications where scan time is critical, such as coronary artery imaging. Unlike the traditional clinically used T1 / T2 weighting, the bSSFP sequence exhibits a special T2 / T1 weighting: the closer the tissue T2 / T1 ratio is to 1, the higher the signal (typically, tissue T2 values ​​are less than T1 values, and both are positive); conversely, the smaller the T2 / T1 ratio, the lower the signal.

[0003] Generally, blood within the heart's blood vessels exhibits both high T1 and high T2 values ​​(i.e., T2 / T1 is approximately 1), resulting in a strong signal. This contrasts sharply with the low signal intensity (T2 / T1 < 1) of the vessel wall tissue. Coronary artery imaging utilizes this high contrast between the vessels and their walls to assess and diagnose coronary artery disease. However, in actual imaging, fat also exhibits a high signal intensity because its T2 / T1 ratio is close to 1. Without appropriate strategies to suppress the fat signal, the bright signal from epicardial fat can easily interfere with the vascular signal, reducing the salience of the vessels in the image.

[0004] The main drawbacks of existing methods are as follows: 1) Theoretically, after the application of a frequency-selective fat-suppression pulse, the fat signal will saturate and then gradually recover over time. Generally, the fat suppression effect is best if the data acquisition phase begins immediately after the pulse application, and deteriorates thereafter. However, in traditional coronary artery imaging based on bSSFP sequences, due to the steady-state characteristics of bSSFP, several repetitive cycles without signal acquisition must be added between the fat-suppression pulse and data acquisition to allow the signal to enter a steady state. Although this method can greatly reduce image artifacts caused by signal oscillations, it also sacrifices some fat-suppression effect; 2) Under resonance conditions, the α / 2-TR / 2 preparation pulse selected in traditional methods can allow the magnetization vector to quickly enter a steady state. However, under partial resonance conditions, the signal generated by the magnetization vector will still have significant signal oscillations, leading to image artifacts. In actual cardiac scans, the magnetization vector in the imaging area (especially around distal vessels) inevitably undergoes partial resonance. Therefore, this preparation pulse mode often cannot completely avoid image artifacts.

[0005] Based on this, this invention proposes a fat suppression coronary imaging method based on a steady-state free precession sequence. An exponentially decaying variable deflection angle bSSFP sequence is designed, utilizing the inherent zero-crossing point that occurs during the transition of the partial resonance signal from FSE mode to bSSFP mode to achieve a more efficient and robust fat suppression effect. Furthermore, this bSSFP sequence has a very smooth signal evolution curve, which can avoid partial resonance signal oscillations caused by imperfect preparation pulses, thereby reducing the possibility of residual artifacts and improving image quality. Summary of the Invention

[0006] The purpose of this invention is to provide a fat suppression coronary imaging method based on a steady-state free precession sequence. This method combines the insensitivity of the free-sounding sequence (FSE) to off-resonance phenomena in bSSFP, causing the magnetization vector signal corresponding to adipose tissue to exhibit a zero-crossing point during the transition from FSE to bSSFP. This feature is then utilized for fat suppression. This technique can largely ensure a smooth signal evolution and prevent image artifacts caused by signal oscillations. Simultaneously, it avoids the drawbacks of other bSSFP fat suppression sequences, such as increased imaging time and limitations on TR selection.

[0007] The technical problem solved by this invention is: In the existing coronary artery imaging based on bSSFP sequences, a preparation pulse needs to be applied to quickly enter a steady state. However, under partial resonance conditions, the signal generated by the magnetization vector still exhibits significant signal oscillation, leading to image artifacts.

[0008] This invention can be achieved through the following technical solution: a fat-suppressed coronary artery imaging method based on steady-state free precession sequences, characterized by comprising the following steps:

[0009] Step 1: Reconstruct the bSSFP sequence using a short-time exponentially decaying radio frequency excitation mode with a variable flip angle;

[0010] Step 2: Determine the optimal parameters in the exponentially decaying variable flip angle RF pulse model through simulation experiments;

[0011] Step 3: Perform phantom verification to determine the optimal zero-crossing position;

[0012] Step 4: Design a segmented alternating phase encoding strategy for the coronary imaging process;

[0013] Step 5: Human trials for verification, adjusting to the optimal parameters for actual imaging.

[0014] A further technical improvement of the present invention is that the radio frequency excitation mode with exponential decay and variable flip angle in step one specifically includes:

[0015] A1: Apply a 90° radio frequency pulse to completely deflect the magnetization vector to the xoy plane;

[0016] A2: Immediately afterwards, a 180° radio frequency pulse is applied, causing the magnetization vector to flip 180°;

[0017] A3: Apply N TIDE A series of exponentially decaying pulses, the radio frequency pulse attenuation model is shown in Equation [1] and Equation [2];

[0018]

[0019] α(i)=θ(i)+θ(i+1) [2]

[0020] Where θ represents the angle of deviation between the magnetization vector and the Z-axis, or simply the deflection angle; α represents the actual applied RF pulse angle; n is an empirical value selected during the actual scanning process; α min α represents the magnitude of the deflection angle at the end of the decay. max This indicates the deflection angle at which decay begins.

[0021] A further technical improvement of the present invention is that: in step three, zero-phase gradient encoding is used for pre-scanning in the phantom verification, and the phantom verification does not require signal triggering.

[0022] A further technical improvement of the present invention is that: the pre-scan is performed using the optimal parameters determined in step two. After exciting the resonant and partial resonant protons, a one-dimensional inverse Fourier transform is performed on the received signal to obtain the projection signal evolution curves of the resonant and partial resonant protons, thereby obtaining the optimal zero-crossing position of the partial resonant signal evolution curve.

[0023] A further technical improvement of the present invention is that: in step four, the phase encoding gradient needs to be turned on during human body scanning, and a segmented alternating phase encoding method is formulated. The signal acquisition of each heartbeat cycle starts from the center position of K space and extends alternately to the edge of K space.

[0024] A further technical improvement of the present invention is that signal triggering is required in human body scanning imaging, and the triggering method is ECG triggering or PG triggering.

[0025] A further technical improvement of the present invention is that the bSSFP sequence acquires coronary artery scanning data at the optimal zero-crossing position, and performs pseudo-half Fourier acquisition alternately at intervals within each heartbeat cycle, thereby achieving efficient suppression of fat signals.

[0026] A further technical improvement of the present invention is that, in human experimental verification, human body scanning imaging is performed based on the optimal parameters and the best zero-crossing position determined in steps two and three, and the above-mentioned optimal parameters are fine-tuned according to the actual imaging situation.

[0027] Compared with the prior art, the present invention has the following beneficial effects:

[0028] 1. By employing an exponentially variable flip angle attenuation RF pulse mode, a smooth transition of the magnetization vector to a steady state is achieved, avoiding the partial resonance signal oscillation and image residual artifacts that may be caused by traditional preparation pulses. Compared to the linear variable flip angle attenuation mode used in traditional T IDE technology, the method of this invention can also, to some extent, avoid weak signal oscillations at the boundary between the end of attenuation and the formal signal acquisition; in addition, it can greatly reduce the frequency of using large flip angles and reduce SAR values.

[0029] 2. Using T IDE technology with exponential decay mode instead of frequency-selective fat suppression pulses to achieve fat inhibition avoids the problem of insufficient fat suppression caused by the necessary preparatory pulses after selective fat suppression pulses. Furthermore, this method is more robust and less sensitive to off-resonance phenomena.

[0030] 3. The redesigned phase encoding scheme is more suitable for triggered coronary artery imaging. Attached Figure Description

[0031] To facilitate understanding by those skilled in the art, the present invention will be further described below with reference to the accompanying drawings.

[0032] Figure 1 This is a flowchart illustrating the implementation steps of the present invention;

[0033] Figure 2 This is a schematic diagram illustrating the evolution of the radio frequency pulse flip angle according to the present invention;

[0034] Figure 3 This is a schematic diagram illustrating the evolution of the (partial) resonance signal in this invention;

[0035] Figure 4 This is a schematic diagram of the segmented alternating phase encoding of the present invention. Detailed Implementation

[0036] To further illustrate the technical means and effects of the present invention in achieving its intended purpose, the following detailed description of the specific implementation methods, structures, features, and effects of the present invention, in conjunction with the accompanying drawings and preferred embodiments, is provided.

[0037] Please see Figure 1-4 As shown, a fat-suppressed coronary artery imaging method based on steady-state free precession sequences is proposed. Its core includes two aspects: designing an exponentially decaying variable deflection angle radio frequency pulse to reconstruct the bSSFP sequence structure and constructing a segmented alternating phase coding strategy for coronary artery imaging.

[0038] Reconstruction of the bSSFP sequence structure:

[0039] As an important class of steady-state sequences, bSSFP sequences utilize the steady-state signals of both the transverse and longitudinal magnetization vectors, largely preserving usable signals and resulting in high signal-to-noise ratios in imaging. However, for bSSFP sequences without any preparation pulses, the signal often requires several hundred transcessive pulses (TRs) to reach steady state. This additional transient time undoubtedly sacrifices the rapid imaging efficiency of bSSFP.

[0040] To match the existing characteristics of bSSFP sequence signal acquisition, some preparatory pulse (or catalytic) techniques need to be used before the bSSFP sequence signal is formally acquired, so that the signal can enter a steady state within a dozen or so TRs.

[0041] A typical preparation pulse pattern is the α / 2-TR / 2-α-TR-α-TR-… pattern, where the first RF pulse deflection angle is α / 2, the interval between the first and second RF pulses is TR / 2, and the RF pulse intensity and repetition time remain constant at α and TR respectively.

[0042] Although adding a preparation pulse greatly reduces the transient time and allows resonant protons to quickly enter a steady state, its effect on off-resonant protons is limited; therefore, reconstructed images (especially in coronary artery imaging where off-resonance phenomena are more pronounced) often contain residual artifacts.

[0043] In traditional T IDE technology, a linearly attenuating RF excitation mode with varying flip angles is used to ensure smooth signal evolution. However, in this invention, to minimize the use of large flip angles (reducing SAR values ​​during scanning) and avoid weak signal oscillations at the attenuation tail of the linear mode, we need to design a set of exponentially attenuating RF pulse modes with varying flip angles to allow the signal to quickly and smoothly enter its final steady state.

[0044] like Figure 2 As shown, in this invention, an exponentially decaying radio frequency excitation mode with varying flip angle is used for signal excitation, specifically including:

[0045] A1: Apply a 90° radio frequency pulse (①) to completely deflect the magnetization vector to the xoy plane;

[0046] A2: Then apply a 180° radio frequency pulse (②, α) max This causes the magnetization vector to flip 180°.

[0047] A3: Apply N TIDE A series of exponentially decaying pulses (②-③), the radio frequency pulse attenuation model is shown in formula [1] and formula [2];

[0048]

[0049] α(i)=θ(i)+θ(i+1)[2]

[0050] Where θ represents the angle of deviation between the magnetization vector and the Z-axis, or simply the deflection angle; α represents the actual applied RF pulse angle; n is an empirical value selected during the actual scanning process; α min α represents the magnitude of the deflection angle at the end of the decay. max This indicates the deflection angle at which decay begins.

[0051] In this mode, the signal evolution caused by the large flip angle in the first half is similar to that of the FSE signal. Since the FSE signal is not sensitive to partial resonance, its evolution is smooth with almost no signal oscillation. Later, during the transition to the bSSFP signal, the resonant signal smoothly transitions to a steady state, while the partial resonance signal exhibits a distinct zero-crossing point, such as... Figure 3 As shown.

[0052] A segmented alternating phase coding strategy for coronary artery imaging:

[0053] Theoretically, if the data acquisition at the center of K-space is scheduled at the zero-crossing point during data acquisition, the tissue signal corresponding to the off-resonance protons is optimally suppressed. Utilizing this characteristic, and combined with the practicalities of coronary artery imaging, a pseudo-semi-Fourier acquisition method with segmented alternating phase encoding within K-space is constructed for data acquisition. The encoding method is as follows: Figure 4 As shown;

[0054] In the diagram, T DELAY The delay from triggering to end-diastole is indicated; TIDE indicates an exponentially decaying radio frequency pulse; ACQ* indicates data acquisition; the right side is a schematic diagram of K-space filling. Signal acquisition for each heartbeat cycle starts from the center of K-space and extends alternately to the edge of K-space. The arrows clearly indicate the acquisition direction and sequence. This illustration takes four heartbeat cycles as an example.

[0055] like Figure 1 As shown, combining the reconstructed bSSFP sequence and the segmented alternating phase-coded half-Fourier acquisition method in K-space, the fat-suppressed coronary artery imaging method based on steady-state free precession sequences includes the following steps:

[0056] Step 1: Determine the optimal parameters in the exponentially decaying variable flip angle RF pulse model through simulation experiments, that is, determine the optimal N in formula [1]. TIDE α max α min And the n parameter;

[0057] Step 2: Perform phantom verification to determine the optimal zero-crossing position (signal triggering is not required in this step).

[0058] S21. Use the optimal parameters determined in step one to perform a pre-scan to obtain the projection signal;

[0059] S22. Calculate the projection signal using one-dimensional inverse Fourier transform;

[0060] S23. Analyze the evolution curve of the (partial) resonance proton projection signal and extract the optimal zero-crossing position;

[0061] S24. Use the optimal zero-crossing position in the actual imaging sequence to test the fat-suppressing effect of the image.

[0062] Regarding the pre-scan, it should be noted that this process requires turning off the phase encoding gradient and continuously acquiring zero-phase encoded signals. However, the bSSFP sequence during the scan is completely consistent with that during the formal imaging scan, thereby acquiring a series of projection signals along the X-direction. The pre-scan requires simultaneously scanning the signal change trends of the water phantom and the fat phantom.

[0063] The evolution relationship diagram of water-fat signals is obtained by processing the projection signals, as shown in the figure. Figure 2 The relationship between the evolution of the resonance signal and the partial resonance signal is shown in the figure. As can be seen from the figure, the signal change trends of the resonance signal and the partial resonance signal are quite different. During conventional sequence scanning, there is a process of correcting the center frequency to the spin frequency of water protons. Due to the inherent frequency difference between water protons and fat protons, fat protons exhibit resonance shift at this point. Therefore, the zero-crossing positions generated by the partial resonance characteristics of adipose tissue can be used to achieve fat suppression.

[0064] More often, the determination of optimal sequence parameters and the pre-scanning process are preparation steps. Only one scan is needed, and the same parameters and zero-crossing results can be used in subsequent imaging processes. There is no need to re-determine them before each scan, thus avoiding extra work.

[0065] Step 3: Actual human body imaging verification (signal trigger required)

[0066] S31. Develop a segmented alternating phase coding strategy;

[0067] S32. Use the optimal parameters determined in step one to perform a scan, and then turn on the phase encoding gradient.

[0068] S33. Collect coronary artery scan data using the optimal zero-crossing position from step two;

[0069] S34. Reconstruct and test the lipid-suppressing effect in the human coronary arteries and output the ideal scan image.

[0070] The above signal triggering methods are ECG triggering or PG triggering; in order to obtain high-resolution coronary artery imaging, it is necessary to reasonably arrange the K-space phase encoding method to obtain high-resolution coronary artery images with efficient fat suppression.

[0071] Furthermore, the simulation results in step one above are based on ideal conditions. However, during actual scanning, differences in hardware and other conditions may cause fluctuations in the designed RF pulse mode. Therefore, after designing the exponentially variable inversion angle attenuation mode pulse, fine-tuning through actual scanning is still necessary.

[0072] Furthermore, the repetition time (TR) of the bSSFP sequence is relatively short, which greatly limits the selectivity of the RF pulse. Generally, when using a small deflection angle, the relationship between the layer selection profile and the RF pulse envelope can be directly approximated using Fourier transform. However, when using a large deflection angle, the layer selection profile obtained by Fourier approximation is not ideal, so techniques such as SLR are needed to assist in the design. In the bSSFP sequence, the TR time is generally short (2-5ms), thus limiting the RF excitation time. In this invention, a short-duration pulse with a large flip angle is redesigned for the actual bSSFP sequence to obtain an ideal layer selection profile.

[0073] Explanation of relevant terms in this invention:

[0074] Lateral and longitudinal magnetization vectors: Generally, when the object / human body being measured enters the scanning cavity (which has strong magnetism), it becomes "magnetized." This can be imagined as countless tiny magnets arranged in a single direction (the main magnetic field) within the tissue. When a radio frequency pulse of specific energy is applied, the orientation of these "tiny magnets" deviates from the direction of the main magnetic field. At this point, we generally call the component of the magnetization intensity along the direction of the main magnetic field the longitudinal magnetization vector, and the component of the magnetization intensity along the direction perpendicular to the main magnetic field the lateral magnetization vector.

[0075] T1 and T2 values: Inherent characteristic parameters of human tissues. When a radio frequency pulse of specific energy is applied, the direction of the magnetization vector will completely deflect to the direction perpendicular to the main magnetic field. Once the radio frequency excitation ends, the magnetization vector needs to return to the direction of the main magnetic field. This phenomenon, macroscopically speaking, manifests on the one hand as the recovery of the longitudinal magnetization vector, i.e., T1 relaxation, and on the other hand as the dissipation of the transverse magnetization vector, i.e., T2 relaxation. Generally, the time it takes for the longitudinal magnetization vector to recover to 63% of its initial state is the T1 value, and the time it takes for the transverse magnetization vector to dissipate to 37% of its initial state is the T2 value.

[0076] T1 / T2 weighted imaging: Two commonly used contrast modes for magnetic resonance imaging in clinical practice. Imaging is performed during the intermediate process of T1 recovery or T2 decay. By utilizing the differences in T1 or T2 values ​​between different tissues, different contrasts (i.e., different gray levels) are produced between different tissues, making it easier to distinguish between different tissue signals or abnormal and normal tissue signals.

[0077] Resonance: When the precession frequency of protons within the main magnetic field is consistent with the Larmor frequency of the main magnetic field, it is called a resonant state. When the precession frequency of protons is higher or lower than the Larmor frequency, it is called a partial resonant state. Generally, in actual experiments, we calibrate the Larmor frequency of the main magnetic field to the precession frequency of water protons. At this time, water protons are in a resonant state, while fat protons exhibit a partial resonant state (fat protons and water protons generally have a fixed frequency difference).

[0078] FSE sequence: A fast scan sequence based on spin echo. Multiple echo chains (90-180-180-…) are used to accelerate the scan and increase scan speed.

[0079] bSSFP sequence: A type of gradient echo sequence. Its main characteristic is that the gradient integrals in all three gradient directions are zero, ensuring that the gradient does not introduce additional phase within each TR. This sequence utilizes the steady-state signals of both the transverse and longitudinal magnetization vectors, achieving a high signal-to-noise ratio within a short TR and high signal utilization (in gradient echo TR, the time interval between the midpoints of two adjacent small-angle pulses).

[0080] K-space: Essentially, it contains the spatial frequency information of an image. Data acquired from a typical scan sequence is filled into the K-space according to a certain pattern, and then a Fourier transform is performed to generate a magnetic resonance imaging (MRI) image for clinical diagnosis.

[0081] Frequency coding and phase coding: In order to encode the actual position of tissue pixels in space into K-space, we generally use frequency coding to distinguish position information in one direction and phase coding to distinguish position information in another direction.

[0082] Projected signal: Theoretically, if we disable phase encoding, the acquired signal can only distinguish positional information in a single direction. In this case, if we perform a one-dimensional inverse Fourier transform along that direction, the signal at a single pixel position actually reflects the sum (or integral) of the signal distributed along another direction, which is what we call the projected signal.

[0083] Artifacts: Due to imperfections in the magnetic resonance imaging system or missing data, the imaging results may show phenomena such as tangles or black stripes.

[0084] Coronary artery imaging: Specific sequences (such as bSSFP used in this invention) are typically used to make the blood vessels appear as high-signal vessels in the scanned images, while the vessel walls appear as relatively low-signal vessels. This contrast advantage is then utilized to analyze and diagnose vascular lesions.

[0085] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention in any way. Although the present invention has been disclosed above with reference to preferred embodiments, it is not intended to limit the present invention. Any person skilled in the art can make some modifications or alterations to the above-disclosed technical content to create equivalent embodiments without departing from the scope of the present invention. Any simple modifications, equivalent changes and alterations made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the scope of the present invention.

Claims

1. A fat-suppressed coronary artery imaging method based on steady-state free precession sequences, characterized in that: Includes the following steps: Step 1: Reconstruct the bSSFP sequence using a short-time exponentially decaying radio frequency excitation mode with a variable flip angle; Step 2: Determine the optimal parameters in the exponentially decaying variable flip angle RF pulse model through simulation experiments; Step 3: Perform phantom verification to determine the optimal zero-crossing position; Step 4: Design a segmented alternating phase encoding strategy for the coronary imaging process; Step 5: Human trials for verification, adjusting to the optimal parameters for actual imaging; The exponentially decaying, flip-angle-dependent RF excitation mode described in step one specifically includes: A1: Apply a 90° radio frequency pulse to completely deflect the magnetization vector to the xoy plane; A2: Immediately afterwards, a 180° radio frequency pulse is applied, causing the magnetization vector to flip 180°; A3: Apply N TIDE A series of exponentially decaying pulses, the radio frequency pulse attenuation model is shown in Equation [1] and Equation [2]; [1] in, The angle between the magnetization vector and the Z-axis, also known as the deflection angle. This represents the actual angle of the applied radio frequency pulse; n is an empirical value selected during the actual scanning process. This indicates the magnitude of the deflection angle at the end of the decay. This indicates the magnitude of the deflection angle at the start of decay; In this mode, the signal evolution caused by the large flip angle in the first half is analogous to FSE. The FSE signal is not sensitive to the partial resonance phenomenon, so the signal evolution is smooth. Then, in the transition to the bSSFP signal, the resonance signal still smoothly transitions to the steady state, while the partial resonance signal will show a clear zero crossing position. In step four, the phase encoding gradient needs to be turned on during the human body scan, and a segmented alternating phase encoding method needs to be formulated. The signal acquisition for each heartbeat cycle starts from the center of the K space and extends alternately to the edge of the K space.

2. The method for fat-suppressed coronary artery imaging based on steady-state free precession sequences according to claim 1, characterized in that, In step three, the phantom verification uses zero-phase gradient encoding for pre-scanning, and the phantom verification does not require signal triggering.

3. The method for fat-suppressed coronary artery imaging based on steady-state free precession sequences according to claim 2, characterized in that, The pre-scan is performed using the optimal parameters determined in step two. After exciting the resonant and partial resonant protons, a one-dimensional inverse Fourier transform is performed on the received signal to obtain the projection signal evolution curves of the resonant and partial resonant protons, thereby obtaining the optimal zero-crossing position of the partial resonant signal evolution curve.

4. The method for fat-suppressed coronary artery imaging based on steady-state free precession sequences according to claim 1, characterized in that, Human body scanning imaging requires signal triggering, either ECG triggering or PG triggering.

5. The method for fat-suppressed coronary artery imaging based on steady-state free precession sequences according to claim 1, characterized in that, The bSSFP sequence acquires coronary artery scan data at the optimal zero-crossing position and performs pseudo-half Fourier acquisition at alternating intervals during each heartbeat cycle, thereby achieving efficient suppression of fat signals.

6. The method for fat-suppressed coronary artery imaging based on steady-state free precession sequences according to claim 1, characterized in that, In human trials, human body scanning and imaging were performed based on the optimal parameters and the best zero-crossing position determined in steps two and three, and the optimal parameters were fine-tuned based on the actual imaging results.

Images