Low-field magnetic resonance parallel imaging array design method and acquisition parameter optimization method

Abstract

This paper presents a design method and acquisition parameter optimization method for a low-field magnetic resonance parallel imaging array. The design method first acquires system parameters, then generates coil elements and adjustable parameters on a support model. These adjustable parameters are then constrained and optimized to ensure the coil elements meet multiple parallel accelerated imaging conditions. The design parameters are then output, and the receiving array is designed based on these parameters. The acquisition parameter optimization method first acquires relevant parameters of the coil elements, then determines parallel reconstruction parameters based on these parameters, and finally adjusts and optimizes these parameters. Since the coil elements can meet multiple parallel accelerated imaging conditions, parallel accelerated magnetic resonance imaging can be achieved in a low-field environment. Because the receiver sensitivity gradient is maximized after coil element optimization, a larger acceleration factor can be set, resulting in higher imaging efficiency. Therefore, this design can achieve parallel accelerated magnetic resonance imaging in a low-field environment with high imaging efficiency.

Description

Technical Field

[0001] This invention relates to a parallel imaging array design method and an acquisition parameter optimization method, belonging to the field of low-field MRI technology, and particularly to a low-field magnetic resonance parallel imaging array design method and acquisition parameter optimization method that combines hardware and software to reduce thermal noise. Background Technology

[0002] Magnetic resonance imaging (MRI), with its advantages of no ionizing radiation, excellent soft tissue contrast, and multi-parameter imaging, has become a core tool for early screening and precise diagnosis of neuroscience, oncology, and cardiovascular diseases. Currently, MRI technology is extending to primary healthcare and bedside monitoring; however, it is limited by the inherently low signal-to-noise ratio (SNR) caused by the reduced strength of the main magnetic field, and thermal noise is quite noticeable in the images. Furthermore, the introduction of environmental electromagnetic interference in unshielded environments easily leads to texture blurring and loss of detail, severely restricting the clarity of MRI images.

[0003] Chinese patent application number 202410528604.7, filed on April 29, 2024, discloses an ultra-low field magnetic resonance imaging system based on emission array spatial coding. The system includes an emission array spatial coding coil group, a multi-channel noise and signal synchronous acquisition spectrometer subsystem, and a computer. The computer is connected to the multi-channel noise and signal synchronous acquisition spectrometer subsystem to send pulse sequences and simultaneously receive MRI imaging signals and noise signals acquired by the spectrometer subsystem, removing electromagnetic interference signals through active noise reduction and deep learning noise reduction. The multi-channel noise and signal synchronous acquisition spectrometer subsystem is connected to the emission array spatial coding coil group, generating excitation signals based on the pulse sequences sent by the computer and outputting them to the emission array spatial coding coil group to drive the coils to generate a magnetic field. Although this design can remove electromagnetic interference signals through noise reduction to achieve clear imaging, it still has the following drawbacks:

[0004] The spatial coding scheme for the transmitter array disclosed in this design uses the amplitude and phase gradient distribution on the transmitter side as its spatial positioning information. Its coding mechanism is completely deviated from the traditional k-space linear sampling paradigm. That is, standard parallel reconstruction methods based on k-space local correlation, such as GRAPPA and SENSE, cannot be directly applied. Therefore, it loses the ability to further shorten the scanning time using parallel imaging technology, resulting in a shortcoming in clinical efficiency.

[0005] The information disclosed in this background section is intended only to enhance understanding of the overall background of this application and should not be construed as an admission or in any way implying that the information constitutes prior art known to those skilled in the art. Summary of the Invention

[0006] The purpose of this invention is to overcome the defects and problems of existing technologies, such as the inability to achieve parallel accelerated magnetic resonance imaging in low-field environments and poor imaging efficiency, and to provide a design method and acquisition parameter optimization method for a low-field magnetic resonance parallel imaging array that can achieve parallel accelerated magnetic resonance imaging in low-field environments and has high imaging efficiency.

[0007] To achieve the above objectives, the technical solution of the present invention is:

[0008] A method for designing a low-field magnetic resonance parallel imaging array, the method comprising the following steps:

[0009] Step 1: Obtain the system parameters of the target low-field magnetic resonance system, including the main magnetic field direction, Larmor frequency, target imaging area, phase encoding direction, number of receiving channels, and target parallel imaging acceleration factor;

[0010] Step 2: Establish a support surface model that matches the shape of the target imaging area, and generate multiple candidate coil units on the support surface model; wherein, the adjustable parameters of the candidate coil units include at least one or a combination of the following: number of coil units, shape of coil units, area of ​​coil units, position of coil units, rotation angle of coil units, number of turns of coil units, overlapping area of ​​adjacent coil units, and number of coil unit layers.

[0011] Step 3: Adjust the position of the candidate coil unit according to the direction of the main magnetic field so that the radio frequency receiving field component generated by the candidate coil unit in the target imaging area satisfies the orthogonal receiving condition with the direction of the main magnetic field;

[0012] Step 4: Based on the phase coding direction and the target parallel imaging acceleration factor, construct an array optimization objective function; the array optimization objective function includes at least two of the following: increasing the difference in receiving sensitivity between different coil units in the phase coding direction, improving the synthetic sensitivity coverage integrity of the target imaging area, reducing the coupling degree between coil units, and improving the single-channel signal-to-noise ratio;

[0013] Step 5: Using the array optimization objective function as the optimization objective, and with constraints such as the orthogonality of the radio frequency receiving field, the coverage integrity of the target imaging area, the decoupling performance between coil units, the manufacturability of coil units, and the number of receiving channels, the adjustable parameters of the candidate coil units are optimized to obtain a set of coil units that meet the requirements of parallel imaging.

[0014] Step 6: Output the receiving array design parameters according to the coil unit set. The receiving array design parameters include at least one or a combination of the following: number of coil units, shape of coil units, area of ​​coil units, position of coil units, rotation angle of coil units, number of turns of coil units, overlapping area of ​​adjacent coil units, number of coil unit layers, and tuning matching parameters corresponding to the Larmor frequency. Then design the receiving array according to the receiving array design parameters.

[0015] The fifth step, optimizing the adjustable parameters of the candidate coil units, includes an array level selection step:

[0016] The number of effective coil units arranged along the phase encoding direction is determined based on the size of the target imaging region in the phase encoding direction, the target parallel imaging acceleration factor, and the number of receiving channels.

[0017] When the number of receiving channels and the number of effective coil units can meet the requirements of target imaging area coverage integrity, phase coding direction receiving sensitivity difference and decoupling performance between coil units on the single-layer support curved surface, a single-layer sparse array structure is selected.

[0018] When the number of receiving channels or the target parallel imaging acceleration factor makes it impossible for a single-layer sparse array structure to simultaneously meet the requirements of target imaging area coverage integrity, phase coding direction receiving sensitivity difference, and decoupling performance between coil units, a multi-layer dense array structure is selected.

[0019] In the fifth step, the multi-layer dense array structure includes a first layer of coil units arranged along the phase encoding direction, and at least one supplementary coil unit offset along the direction perpendicular to the phase encoding direction or along the normal of the support surface, so as to improve the spatial sampling density of the target imaging area and reduce noise amplification in parallel imaging reconstruction.

[0020] In the sixth step, the tuning matching parameters corresponding to the Larmor frequency are obtained through a multi-channel joint tuning matching step, which includes:

[0021] (1) After the coil unit is fixed on the support surface model according to the optimized coil unit position, coil unit rotation angle and adjacent coil unit overlap area, the resonant frequency, input impedance and reflection parameters of each coil unit in the no-load state and the load state are obtained respectively.

[0022] (2) Based on the Larmor frequency of the target low-field magnetic resonance system and the resonant frequency of the coil unit under no-load conditions, adjust the series adjustable capacitor and / or parallel adjustable capacitor in each coil unit so that the resonant frequency of each coil unit under load conditions is consistent with the Larmor frequency.

[0023] (3) Connect each coil unit to the matching network and adjust the matching network so that the input impedance of each coil unit at the Larmor frequency is converted to the reference impedance range;

[0024] (4) After completing the tuning and matching of any coil unit, the transmission parameters between the coil unit and the adjacent coil unit are detected, and when the transmission parameters do not meet the preset decoupling threshold, the matching parameters of the coil unit and the overlapping area of ​​the adjacent coil units are jointly adjusted until the reflection parameters of the coil unit meet the reflection parameter constraints and the transmission parameters meet the decoupling parameter constraints.

[0025] (5) Traverse all coil units and repeat (2) to (4) to obtain the tuning capacitance value, matching capacitance value, reference impedance range, and corresponding reflection and transmission parameters of each coil unit, thus obtaining the tuning matching parameters corresponding to the Larmor frequency.

[0026] In the sixth step, the method further includes setting transmit protection detuning parameters for each coil unit, the step including:

[0027] Based on the tuning capacitor value, matching capacitor value, coil inductance value, and Larmor frequency of the target low-field magnetic resonance system of the coil unit, determine the active detuning branch parameters and passive detuning branch parameters corresponding to the coil unit.

[0028] A preamplifier, an active detuning branch, and a passive detuning branch are connected to each coil unit. The active detuning branch includes an active control switching device and a first resonant element, and the passive detuning branch includes a passive conduction device and a second resonant element.

[0029] When the target low-field magnetic resonance system is in the radio frequency transmission stage, the active control switching device is turned on under the action of the external bias control signal, and the first resonant element constructs a parallel resonant circuit, so that the active detuning branch forms a specified impedance state at the transmission frequency to block the radio frequency transmission energy from entering the corresponding coil unit and its preamplifier.

[0030] When the active detuning branch is not turned on or the radio frequency transmission energy exceeds a preset threshold, the passive turning device turns on under the action of the induced voltage at both ends of the coil unit. The second resonant element constructs a parallel resonant circuit, so that the passive detuning branch forms a specified impedance state at the transmission frequency, so as to provide redundant protection for the coil unit and its preamplifier.

[0031] When the target low-field magnetic resonance system is in the signal receiving stage, both the active detuning branch and the passive detuning branch are in a non-detuned state, so that the coil unit is maintained in a tuned matching state corresponding to the Larmor frequency, and the reflection parameter constraint and the decoupling parameter constraint between coil units are satisfied.

[0032] In the sixth step, the method further includes a preamplifier decoupling parameter determination step, which includes:

[0033] Based on the transmission parameters, mutual inductance parameters, and input impedance of each coil unit under load, determine the target range of the preamplifier input impedance corresponding to each coil unit.

[0034] The input impedance of the preamplifier is made lower than the equivalent loop impedance of the corresponding coil unit at the Larmor frequency, so that the crosstalk current generated by the coupling between the coil units flows preferentially into the input terminal of the corresponding preamplifier.

[0035] Within the target range of the input impedance of the preamplifier, the input impedance, noise figure, gain, and stability parameters of the preamplifier are jointly constrained to ensure that the signal-to-noise ratio of each coil unit meets the single-channel signal-to-noise ratio requirement and the transmission parameters between coil units meet the decoupling parameter requirements.

[0036] When the transmission parameters between any two coil units do not meet the preset decoupling threshold, the input impedance of the preamplifier, the matching network parameters, and the overlap area of ​​adjacent coil units are jointly adjusted until the reflection parameters of the coil units meet the reflection parameter constraints, the transmission parameters of the coil units meet the transmission parameter constraints, and the signal-to-noise ratio of the coil units meets the single-channel signal-to-noise ratio constraints.

[0037] A method for optimizing acquisition parameters in the design of a low-field magnetic resonance parallel imaging array, the method comprising the following steps:

[0038] Step 1: Obtain the receiving sensitivity distribution, single-channel signal-to-noise ratio, reflection parameters, and decoupling parameters between coil units in the receiving array within the target imaging area;

[0039] Step 2: Based primarily on the degree of difference in the receiving sensitivity distribution of each coil unit in the phase encoding direction, and secondarily on the single-channel signal-to-noise ratio, reflection parameters, and decoupling parameters between coil units, determine the parallel imaging acquisition parameters. The parallel imaging acquisition parameters include at least one or a combination of external reduction factor, central full sampling area range, phase encoding line sampling interval, and parameter calibration area range.

[0040] In the second step, determining the parallel imaging acquisition parameters includes:

[0041] The candidate range of the external reduction factor is determined based on the degree of difference in the receiving sensitivity distribution of different coil units in the phase coding direction;

[0042] When the difference in the receiving sensitivity distribution is lower than the preset sensitivity difference threshold, or the decoupling parameters between coil units do not meet the preset decoupling threshold, the external reduction factor is reduced or the range of the central full sampling area is increased.

[0043] When the degree of difference in the receiving sensitivity distribution meets the preset sensitivity difference threshold, and the decoupling parameters between coil units meet the preset decoupling threshold, the external reduction factor is increased or the range of the central full sampling area is reduced.

[0044] There is a third step after the second step:

[0045] Step 3: Based on the single-channel signal-to-noise ratio, reflection parameters, and decoupling parameters between coil units, determine the channel quality parameters of each coil unit; these channel quality parameters are used to determine the set of coil units participating in parallel imaging reconstruction, specifically including:

[0046] Each coil unit is scored based on its single-channel signal-to-noise ratio, reflection parameters, decoupling parameters between coil units, and receiver sensitivity coverage.

[0047] When the quality score of any coil unit is lower than the preset channel quality threshold, the weight of that coil unit in parallel imaging reconstruction is reduced, or the coil unit is removed from the set of coil units participating in parallel imaging reconstruction.

[0048] When the decoupling parameters of any coil unit and its adjacent coil units do not meet the preset decoupling threshold, the channel weights and acquisition parameters corresponding to that coil unit are adjusted to reduce the impact of channel crosstalk on parallel imaging reconstruction.

[0049] There is a fourth step following the third step:

[0050] Step 4: Based on the received sensitivity distribution, parallel imaging acquisition parameters, and channel quality parameters, determine the parallel imaging reconstruction parameters used for frequency space data completion; the determination of the parallel imaging reconstruction parameters includes:

[0051] Based on the central full sampling area range, the external reduction factor, the set of coil units participating in parallel imaging reconstruction, the channel quality parameters, and the degree of difference in the receiving sensitivity distribution of each coil unit in the phase coding direction, the neighborhood size, the number of participating channels, and the channel weights of the parallel imaging reconstruction model are determined.

[0052] When the channel quality parameter is lower than the preset channel quality threshold, increase the parameter calibration area range, reduce the channel weight of the corresponding coil unit, or adjust the neighborhood size of the parallel imaging reconstruction model.

[0053] When the image quality evaluation parameters after parallel imaging reconstruction do not meet the preset image quality threshold, the external reduction factor, the range of the central full sampling area, the number of participating channels or the channel weights are adjusted, and the parallel imaging reconstruction parameters are redefined.

[0054] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0055] 1. In the low-field magnetic resonance parallel imaging array design method and acquisition parameter optimization method of the present invention, the low-field magnetic resonance parallel imaging array design method includes the following steps: Step 1: Obtain the system parameters of the target low-field magnetic resonance system; Step 2: Establish a support surface model based on the target imaging region, and generate multiple candidate coil units and adjustable parameters on the model; Step 3: Adjust the position of the candidate coil units so that the radio frequency receiving field component generated by the candidate coil units in the target imaging region satisfies the orthogonal receiving condition with the direction of the main magnetic field; Step 4: Construct an array optimization objective function based on the system parameters, the array optimization objective function including at least: increasing the number of different coil units in the phase encoding direction. The fifth step involves optimizing the array with the objective function as the optimization target and with constraints such as the orthogonality of the RF receiving field, the coverage integrity of the target imaging area, the decoupling performance between coil units, the manufacturability of coil units, and the number of receiving channels, to obtain a set of coil units that meets the requirements of parallel imaging. The sixth step involves outputting the receiving array design parameters based on the set of coil units, and then designing the receiving array based on the receiving array design parameters. The advantages of this invention also include:

[0056] First, the radio frequency receiving field component generated by the coil unit in the target imaging area satisfies the orthogonal receiving condition with the direction of the main magnetic field. Therefore, the receiving array designed based on this invention can be applied to the MRI system to receive weak radio frequency signals.

[0057] Secondly, this invention optimizes the position of the coil units through an array optimization objective function, maximizing the receiving sensitivity gradient of each coil unit in the phase coding direction, thus providing sufficient coding information for subsequent parallel imaging. During image reconstruction, cross-validation of multi-channel coding information can more accurately complete the missing k-space data, reducing reconstruction artifacts caused by insufficient sampling or noise interference. Because sufficient coding information is generated, it supports setting a higher external reduction factor, which can significantly shorten the scanning time.

[0058] Thirdly, the array optimization objective function improves the synthetic sensitivity coverage integrity of the target imaging area, that is, it expands the spatial coverage range and reduces the signal blind zone; at the same time, the array optimization objective function reduces the mutual inductance between coil units and suppresses crosstalk noise between coil units by optimizing the overlapping area between coil units.

[0059] Fourthly, through the rigorous verification in the third step, we ensure that the RF receiving field direction of all coil units is perpendicular to the main magnetic field direction, which conforms to Faraday's law of electromagnetic induction, maximizes the induction efficiency of the magnetic resonance signal, avoids signal attenuation caused by field direction mismatch, and improves the signal-to-noise ratio (SNR) from the source.

[0060] Fifthly, compared to existing technologies, the pure receiving device does not require complex transmission links and related control mechanisms. Compared to the spatial coding scheme of the transmission array, the hardware complexity is greatly reduced, and the power consumption and heat dissipation requirements are significantly reduced, making it more suitable for the low-cost and easy-to-deploy requirements of portable / grassroots medical scenarios.

[0061] Sixth point: In the prior art, the effectiveness of the nonlinear coding field depends on the long-term stability of the amplitude and phase control of each transmission channel. Changes in the field pattern caused by factors such as temperature drift and power amplifier aging will directly destroy the coding orthogonality. The system lacks an adaptive correction mechanism, resulting in poor imaging quality. However, the present invention replaces the nonlinear coding at the transmitting end with the difference in receiving sensitivity between different coil units in the phase coding direction. It does not need to rely on the long-term stability of the amplitude and phase control of the transmission channel, thereby avoiding the destruction of coding orthogonality by the aforementioned factors and ensuring stable and excellent imaging quality.

[0062] Seventh point: The optimized array design can directly reuse the traditional gradient coding and parallel imaging reconstruction process without modifying the gradient subsystem of the existing MRI system, thus lowering the threshold for clinical implementation.

[0063] Therefore, the present invention can realize parallel accelerated magnetic resonance imaging in a low-field environment with high imaging efficiency.

[0064] 2. In the low-field magnetic resonance parallel imaging array design method and acquisition parameter optimization method of the present invention, the fifth step of optimizing the adjustable parameters of the candidate coil units includes an array layer selection step: determining the number of effective coil units arranged along the phase encoding direction based on the size of the target imaging region in the phase encoding direction, the target parallel imaging acceleration factor, and the number of receiving channels; when the number of receiving channels and the number of effective coil units can meet the requirements of target imaging region coverage integrity, phase encoding direction receiving sensitivity difference, and decoupling performance between coil units on the single-layer support surface, a single-layer sparse array structure is selected; when the aforementioned requirements cannot be met simultaneously, a multi-layer dense array structure is selected. In application, the single-layer sparse array structure aims to expand the coverage of a single channel, thereby obtaining a higher single-channel signal-to-noise ratio (SNR), while constructing a significantly different sensitivity distribution in the acceleration direction; compared with the single-layer sparse array structure, the dense array structure adds multi-layer stacking along the direction perpendicular to the acceleration direction (such as the axial / interlayer direction) to form two-dimensional or three-dimensional spatial coverage, increasing the spatial sampling density, reducing the geometric factor, and further enhancing the acceleration capability of parallel imaging. Therefore, this invention achieves a precise match between coil resources and imaging requirements.

[0065] 3. In the design method and acquisition parameter optimization method of the low-field magnetic resonance parallel imaging array of the present invention, in the sixth step, the tuning matching parameter corresponding to the Larmor frequency is obtained through a multi-channel joint tuning matching step, the steps of which include: (1) obtaining the resonant frequency, input impedance and reflection parameters of each coil unit under no-load and load conditions respectively; (2) adjusting the series adjustable capacitor and / or parallel adjustable capacitor in each coil unit according to the Larmor frequency of the target low-field magnetic resonance system so that the resonant frequency of each coil unit under load conditions is consistent with the Larmor frequency; (3) connecting each coil unit to the matching network and adjusting the matching network so that the input impedance of each coil unit at the Larmor frequency is converted to the reference impedance range; (4) after completing the tuning matching of any coil unit, detecting the transmission parameters between the coil unit and the adjacent coil units, and when the transmission parameters do not meet the preset decoupling threshold, jointly adjusting the matching of the coil unit. The parameters and the overlapping area of ​​adjacent coil units are calculated until the reflection parameters of the coil unit satisfy the reflection parameter constraints and the transmission parameters satisfy the decoupling parameter constraints; (5) Traverse all coil units and repeat (2) to (4) to obtain the tuning capacitor value, matching capacitor value, reference impedance range and corresponding reflection parameters and transmission parameters of each coil unit, that is, obtain the tuning matching parameters corresponding to the Larmor frequency. In application, this multi-channel joint tuning matching step is the key to ensure the efficient operation of the coil unit. First, adjust the resonant frequency of the circuit so that the coil unit can receive the MR signal of the specific frequency most effectively. Then, clamp the coil impedance strictly within the reference impedance range to eliminate the signal reflection caused by impedance mismatch, ensure that the signal can be output from the preamplifier without loss, and also reduce the noise of the preamplifier to the minimum, greatly reducing the noise of the output image. At the same time, the decoupling effect of the coil unit is ensured by the detection step. In actual operation, iterative adjustments are required to achieve the optimal effect. Therefore, the signal transmission effect of the present invention is better.

[0066] 4. In the sixth step of the present invention, a design method and acquisition parameter optimization method for a low-field magnetic resonance parallel imaging array, the method further includes setting transmission protection detuning parameters for each coil unit, including: determining the active detuning branch parameters and passive detuning branch parameters corresponding to the coil unit based on the relevant parameters of the coil unit and the Larmor frequency of the target low-field magnetic resonance system; connecting a preamplifier, an active detuning branch, and a passive detuning branch to each coil unit; when the target low-field magnetic resonance system is in the radio frequency transmission stage, the active detuning branch forms a specified impedance state at the transmission frequency to block the radio frequency transmission energy from entering the corresponding coil unit and its preamplifier; when the active detuning branch is not conducting or the radio frequency transmission energy exceeds a preset threshold, the passive detuning branch... A specified impedance state is established at a given frequency to provide redundant protection for the coil unit and its preamplifier. When the target low-field magnetic resonance system is in the signal reception phase, both the active and passive detuning branches are in a non-detuned state, maintaining the coil unit in a tuned-matched state corresponding to the Larmor frequency. In application, this mechanism constitutes a safety barrier of "dynamic impedance switching" and "multi-level redundancy protection." During the RF transmission phase, the active detuning branch protects the coil unit. In extreme cases where the transmission power exceeds the limit, the passive detuning branch provides redundant protection for the coil unit. During the signal reception phase, both the active and passive detuning branches return to a non-detuned state, ensuring that the coil unit is strictly maintained in a tuned-matched state, allowing the weak magnetic resonance signal to be transmitted to the preamplifier without loss. Therefore, this invention has a dual safety guarantee mechanism.

[0067] 5. In the low-field magnetic resonance parallel imaging array design method and acquisition parameter optimization method of the present invention, the sixth step further includes a preamplifier decoupling parameter determination step, which includes: determining the target range of the preamplifier input impedance corresponding to each coil unit based on the relevant parameters between each coil unit; ensuring that the input impedance of the preamplifier is lower than the equivalent loop impedance of the corresponding coil unit at the Larmor frequency, so that the crosstalk current generated by the coupling between coil units preferentially flows into the input terminal of the corresponding preamplifier; and then jointly constraining the relevant parameters to ensure that each coil unit meets the single-channel signal-to-noise ratio requirement and the decoupling parameter requirement. If any two coil units cannot meet the aforementioned requirements, then the preamplifier... The amplifier, coil units, and related parameters are adjusted until the aforementioned requirements are met. In application, when adjacent coil units generate electromagnetic coupling due to spatial proximity, the induced crosstalk current will preferentially flow into the lower impedance preamplifier circuit, rather than being transmitted through the mutual inductance path between coils. This low-impedance shunt mechanism effectively cuts off the noise coupling path between channels, significantly improving the isolation of the multi-channel array. This ensures that the k-space data acquired by each channel maintains a high degree of independence and orthogonality, avoiding signal aliasing caused by channel crosstalk from the source, and providing a clean data foundation for subsequent noise reduction and reconstruction based on Kalman filtering. At the same time, parameter constraint steps ensure the stability of the coil unit during operation. Therefore, the data of this invention is pure.

[0068] 6. In the low-field magnetic resonance parallel imaging array design method and acquisition parameter optimization method of the present invention, the acquisition parameter optimization method includes the following steps: First step: Obtain the receiving sensitivity distribution, single-channel signal-to-noise ratio, reflection parameters, and decoupling parameters between coil units in the target imaging area of ​​each coil unit in the receiving array; Second step: Determine the parallel imaging acquisition parameters based on the parameter information in the first step; Third step: Adjust the weight of the coil unit in parallel imaging reconstruction according to the relevant parameters; Fourth step: Determine the parallel imaging reconstruction parameters according to the relevant parameters. In application, the acquisition parameter optimization method achieves the best balance between scanning efficiency and image quality in a low-field environment through a closed-loop process of "sensitivity-driven—quality screening—reconstruction adaptation". When the sensitivity difference is significant and the channel decoupling performance is excellent, it improves... An acceleration factor is used to shorten scanning time; conversely, the acceleration factor is reduced to ensure the image is artifact-free. This data-driven sampling strategy avoids the risk of mismatch between fixed parameters under complex human body loads. The coil unit is quality-scored by combining single-channel signal-to-noise ratio, reflection parameters, and decoupling parameters, dynamically eliminating channels whose performance is degraded due to poor contact or load effects. By reducing the weight of low-quality channels or directly excluding them from the reconstruction set, systematic noise is effectively avoided, improving the robustness of parallel imaging reconstruction from the source. Based on channel quality parameters and sensitivity distribution, the optimal neighborhood size and channel weights of the parallel imaging reconstruction model are determined. When the reconstructed image quality is found to be unsatisfactory, the external reduction factor, central region range, or number of participating channels is automatically adjusted, forming an iterative optimization closed loop. Therefore, the parallel imaging reconstruction effect of this invention is good. Attached Figure Description

[0069] Figure 1 This is a schematic diagram of the structure of the present invention.

[0070] Figure 2 yes Figure 1 The main view.

[0071] Figure 3 This is a schematic diagram illustrating the application of the present invention.

[0072] Figure 4 This is a schematic diagram of the circuit board connected when the present invention is applied.

[0073] Figure 5 yes Figure 1 A schematic diagram of the circuit structure connecting the intermediate coil unit.

[0074] Figure 6 This is a schematic diagram of the structure of Example 2.

[0075] Figure 7 This is a schematic diagram of the acceleration factor in Example 7.

[0076] Figure 8This is a structural schematic diagram of Example 7.

[0077] Figure 9 This is a schematic diagram showing the application results of the present invention compared to the control group. Detailed Implementation

[0078] The present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0079] Please see Figure 1 — Figure 9 A method for designing a low-field magnetic resonance parallel imaging array, the method comprising the following steps:

[0080] Step 1: Obtain the system parameters of the target low-field magnetic resonance system, including the main magnetic field direction, Larmor frequency, target imaging area, phase encoding direction, number of receiving channels, and target parallel imaging acceleration factor;

[0081] Step 2: Establish a support surface model that matches the shape of the target imaging area, and generate multiple candidate coil units on the support surface model; wherein, the adjustable parameters of the candidate coil units include at least one or a combination of the following: number of coil units, shape of coil units, area of ​​coil units, position of coil units, rotation angle of coil units, number of turns of coil units, overlapping area of ​​adjacent coil units, and number of coil unit layers.

[0082] Step 3: Adjust the position of the candidate coil unit according to the direction of the main magnetic field so that the radio frequency receiving field component generated by the candidate coil unit in the target imaging area satisfies the orthogonal receiving condition with the direction of the main magnetic field;

[0083] Step 4: Based on the phase coding direction and the target parallel imaging acceleration factor, construct an array optimization objective function; the array optimization objective function includes at least two of the following: increasing the difference in receiving sensitivity between different coil units in the phase coding direction, improving the synthetic sensitivity coverage integrity of the target imaging area, reducing the coupling degree between coil units, and improving the single-channel signal-to-noise ratio;

[0084] Step 5: Using the array optimization objective function as the optimization objective, and with constraints such as the orthogonality of the radio frequency receiving field, the coverage integrity of the target imaging area, the decoupling performance between coil units, the manufacturability of coil units, and the number of receiving channels, the adjustable parameters of the candidate coil units are optimized to obtain a set of coil units that meet the requirements of parallel imaging.

[0085] Step 6: Output the receiving array design parameters according to the coil unit set. The receiving array design parameters include at least one or a combination of the following: number of coil units, shape of coil units, area of ​​coil units, position of coil units, rotation angle of coil units, number of turns of coil units, overlapping area of ​​adjacent coil units, number of coil unit layers, and tuning matching parameters corresponding to the Larmor frequency. Then design the receiving array according to the receiving array design parameters.

[0086] The fifth step, optimizing the adjustable parameters of the candidate coil units, includes an array level selection step:

[0087] The number of effective coil units arranged along the phase encoding direction is determined based on the size of the target imaging region in the phase encoding direction, the target parallel imaging acceleration factor, and the number of receiving channels.

[0088] When the number of receiving channels and the number of effective coil units can meet the requirements of target imaging area coverage integrity, phase coding direction receiving sensitivity difference and decoupling performance between coil units on the single-layer support curved surface, a single-layer sparse array structure is selected.

[0089] When the number of receiving channels or the target parallel imaging acceleration factor makes it impossible for a single-layer sparse array structure to simultaneously meet the requirements of target imaging area coverage integrity, phase coding direction receiving sensitivity difference, and decoupling performance between coil units, a multi-layer dense array structure is selected.

[0090] In the fifth step, the multi-layer dense array structure includes a first layer of coil units arranged along the phase encoding direction, and at least one supplementary coil unit offset along the direction perpendicular to the phase encoding direction or along the normal of the support surface, so as to improve the spatial sampling density of the target imaging area and reduce noise amplification in parallel imaging reconstruction.

[0091] In the sixth step, the tuning matching parameters corresponding to the Larmor frequency are obtained through a multi-channel joint tuning matching step, which includes:

[0092] (1) After the coil unit is fixed on the support surface model according to the optimized coil unit position, coil unit rotation angle and adjacent coil unit overlap area, the resonant frequency, input impedance and reflection parameters of each coil unit in the no-load state and the load state are obtained respectively.

[0093] (2) Based on the Larmor frequency of the target low-field magnetic resonance system and the resonant frequency of the coil unit under no-load conditions, adjust the series adjustable capacitor and / or parallel adjustable capacitor in each coil unit so that the resonant frequency of each coil unit under load conditions is consistent with the Larmor frequency.

[0094] (3) Connect each coil unit to the matching network and adjust the matching network so that the input impedance of each coil unit at the Larmor frequency is converted to the reference impedance range;

[0095] (4) After completing the tuning and matching of any coil unit, the transmission parameters between the coil unit and the adjacent coil unit are detected, and when the transmission parameters do not meet the preset decoupling threshold, the matching parameters of the coil unit and the overlapping area of ​​the adjacent coil units are jointly adjusted until the reflection parameters of the coil unit meet the reflection parameter constraints and the transmission parameters meet the decoupling parameter constraints.

[0096] (5) Traverse all coil units and repeat (2) to (4) to obtain the tuning capacitance value, matching capacitance value, reference impedance range, and corresponding reflection and transmission parameters of each coil unit, thus obtaining the tuning matching parameters corresponding to the Larmor frequency.

[0097] In the sixth step, the method further includes setting transmit protection detuning parameters for each coil unit, the step including:

[0098] Based on the tuning capacitor value, matching capacitor value, coil inductance value, and Larmor frequency of the target low-field magnetic resonance system of the coil unit, determine the active detuning branch parameters and passive detuning branch parameters corresponding to the coil unit.

[0099] A preamplifier, an active detuning branch, and a passive detuning branch are connected to each coil unit. The active detuning branch includes an active control switching device and a first resonant element, and the passive detuning branch includes a passive conduction device and a second resonant element.

[0100] When the target low-field magnetic resonance system is in the radio frequency transmission stage, the active control switching device is turned on under the action of the external bias control signal, and the first resonant element constructs a parallel resonant circuit, so that the active detuning branch forms a specified impedance state at the transmission frequency to block the radio frequency transmission energy from entering the corresponding coil unit and its preamplifier.

[0101] When the active detuning branch is not turned on or the radio frequency transmission energy exceeds a preset threshold, the passive turning device turns on under the action of the induced voltage at both ends of the coil unit. The second resonant element constructs a parallel resonant circuit, so that the passive detuning branch forms a specified impedance state at the transmission frequency, so as to provide redundant protection for the coil unit and its preamplifier.

[0102] When the target low-field magnetic resonance system is in the signal receiving stage, both the active detuning branch and the passive detuning branch are in a non-detuned state, so that the coil unit is maintained in a tuned matching state corresponding to the Larmor frequency, and the reflection parameter constraint and the decoupling parameter constraint between coil units are satisfied.

[0103] In the sixth step, the method further includes a preamplifier decoupling parameter determination step, which includes:

[0104] Based on the transmission parameters, mutual inductance parameters, and input impedance of each coil unit under load, determine the target range of the preamplifier input impedance corresponding to each coil unit.

[0105] The input impedance of the preamplifier is made lower than the equivalent loop impedance of the corresponding coil unit at the Larmor frequency, so that the crosstalk current generated by the coupling between the coil units flows preferentially into the input terminal of the corresponding preamplifier.

[0106] Within the target range of the input impedance of the preamplifier, the input impedance, noise figure, gain, and stability parameters of the preamplifier are jointly constrained to ensure that the signal-to-noise ratio of each coil unit meets the single-channel signal-to-noise ratio requirement and the transmission parameters between coil units meet the decoupling parameter requirements.

[0107] When the transmission parameters between any two coil units do not meet the preset decoupling threshold, the input impedance of the preamplifier, the matching network parameters, and the overlap area of ​​adjacent coil units are jointly adjusted until the reflection parameters of the coil units meet the reflection parameter constraints, the transmission parameters of the coil units meet the transmission parameter constraints, and the signal-to-noise ratio of the coil units meets the single-channel signal-to-noise ratio constraints.

[0108] A method for optimizing acquisition parameters in the design of a low-field magnetic resonance parallel imaging array, the method comprising the following steps:

[0109] Step 1: Obtain the receiving sensitivity distribution, single-channel signal-to-noise ratio, reflection parameters, and decoupling parameters between coil units in the receiving array within the target imaging area;

[0110] Step 2: Based primarily on the degree of difference in the receiving sensitivity distribution of each coil unit in the phase encoding direction, and secondarily on the single-channel signal-to-noise ratio, reflection parameters, and decoupling parameters between coil units, determine the parallel imaging acquisition parameters. The parallel imaging acquisition parameters include at least one or a combination of external reduction factor, central full sampling area range, phase encoding line sampling interval, and parameter calibration area range.

[0111] In the second step, determining the parallel imaging acquisition parameters includes:

[0112] The candidate range of the external reduction factor is determined based on the degree of difference in the receiving sensitivity distribution of different coil units in the phase coding direction;

[0113] When the difference in the receiving sensitivity distribution is lower than the preset sensitivity difference threshold, or the decoupling parameters between coil units do not meet the preset decoupling threshold, the external reduction factor is reduced or the range of the central full sampling area is increased.

[0114] When the degree of difference in the receiving sensitivity distribution meets the preset sensitivity difference threshold, and the decoupling parameters between coil units meet the preset decoupling threshold, the external reduction factor is increased or the range of the central full sampling area is reduced.

[0115] There is a third step after the second step:

[0116] Step 3: Based on the single-channel signal-to-noise ratio, reflection parameters, and decoupling parameters between coil units, determine the channel quality parameters of each coil unit; these channel quality parameters are used to determine the set of coil units participating in parallel imaging reconstruction, specifically including:

[0117] Each coil unit is scored based on its single-channel signal-to-noise ratio, reflection parameters, decoupling parameters between coil units, and receiver sensitivity coverage.

[0118] When the quality score of any coil unit is lower than the preset channel quality threshold, the weight of that coil unit in parallel imaging reconstruction is reduced, or the coil unit is removed from the set of coil units participating in parallel imaging reconstruction.

[0119] When the decoupling parameters of any coil unit and its adjacent coil units do not meet the preset decoupling threshold, the channel weights and acquisition parameters corresponding to that coil unit are adjusted to reduce the impact of channel crosstalk on parallel imaging reconstruction.

[0120] There is a fourth step following the third step:

[0121] Step 4: Based on the received sensitivity distribution, parallel imaging acquisition parameters, and channel quality parameters, determine the parallel imaging reconstruction parameters used for frequency space data completion; the determination of the parallel imaging reconstruction parameters includes:

[0122] Based on the central full sampling area range, the external reduction factor, the set of coil units participating in parallel imaging reconstruction, the channel quality parameters, and the degree of difference in the receiving sensitivity distribution of each coil unit in the phase coding direction, the neighborhood size, the number of participating channels, and the channel weights of the parallel imaging reconstruction model are determined.

[0123] When the channel quality parameter is lower than the preset channel quality threshold, increase the parameter calibration area range, reduce the channel weight of the corresponding coil unit, or adjust the neighborhood size of the parallel imaging reconstruction model.

[0124] When the image quality evaluation parameters after parallel imaging reconstruction do not meet the preset image quality threshold, the external reduction factor, the range of the central full sampling area, the number of participating channels or the channel weights are adjusted, and the parallel imaging reconstruction parameters are redefined.

[0125] The following are supplementary descriptions of the present invention:

[0126] The coil unit described in this invention is a coil unit of a receiving array. Each coil unit is a surface coil, and the initial area of ​​each coil unit is set to be similar to ensure that the signal-to-noise ratio (SNR) benchmark of each channel is balanced. Then, iterative optimization is performed using electromagnetic simulation software (HFSS, etc.) to make the coil unit fit tightly against the human head on the helmet-shaped support to increase the fill factor and thus improve the SNR. The shape of the coil unit can be rectangular, circular, etc. Circular coils can be used in multi-channel designs, while rectangular coils are generally used in low-channel designs. The number of coil turns is determined based on the quality factor formula, the resonant frequency of the coil unit, and the joint evaluation of simulation and experiment.

[0127] Example 1:

[0128] Please see Figure 1 — Figure 9 A method for designing a low-field magnetic resonance parallel imaging array, the method comprising the following steps:

[0129] Step 1: Obtain the system parameters of the target low-field magnetic resonance system, including the main magnetic field direction, Larmor frequency, target imaging area, phase encoding direction, number of receiving channels, and target parallel imaging acceleration factor;

[0130] Step 2: Establish a support surface model that matches the shape of the target imaging area, and generate multiple candidate coil units on the support surface model; wherein, the adjustable parameters of the candidate coil units include at least one or a combination of the following: number of coil units, shape of coil units, area of ​​coil units, position of coil units, rotation angle of coil units, number of turns of coil units, overlapping area of ​​adjacent coil units, and number of coil unit layers.

[0131] Step 3: Adjust the position of the candidate coil unit according to the direction of the main magnetic field so that the radio frequency receiving field component generated by the candidate coil unit in the target imaging area satisfies the orthogonal receiving condition with the direction of the main magnetic field;

[0132] Step 4: Based on the phase coding direction and the target parallel imaging acceleration factor, construct an array optimization objective function; the array optimization objective function includes at least two of the following: increasing the difference in receiving sensitivity between different coil units in the phase coding direction, improving the synthetic sensitivity coverage integrity of the target imaging area, reducing the coupling degree between coil units, and improving the single-channel signal-to-noise ratio;

[0133] Step 5: Using the array optimization objective function as the optimization objective, and with constraints such as the orthogonality of the radio frequency receiving field, the coverage integrity of the target imaging area, the decoupling performance between coil units, the manufacturability of coil units, and the number of receiving channels, the adjustable parameters of the candidate coil units are optimized to obtain a set of coil units that meet the requirements of parallel imaging.

[0134] Step 6: Output the receiving array design parameters according to the coil unit set. The receiving array design parameters include at least one or a combination of the following: number of coil units, shape of coil units, area of ​​coil units, position of coil units, rotation angle of coil units, number of turns of coil units, overlapping area of ​​adjacent coil units, number of coil unit layers, and tuning matching parameters corresponding to the Larmor frequency. Then design the receiving array according to the receiving array design parameters.

[0135] Example 2:

[0136] The basic content is the same as in Example 1, except that:

[0137] Please see Figure 1 — Figure 6In the fifth step, optimizing the adjustable parameters of the candidate coil units includes an array layer selection step: determining the number of effective coil units arranged along the phase encoding direction based on the size of the target imaging region in the phase encoding direction, the target parallel imaging acceleration factor, and the number of receiving channels; selecting a single-layer sparse array structure when the number of receiving channels and the number of effective coil units can meet the requirements of target imaging region coverage integrity, phase encoding direction receiving sensitivity difference, and decoupling performance between coil units on a single-layer support surface; selecting a multi-layer dense array structure when the number of receiving channels or the target parallel imaging acceleration factor prevents the single-layer sparse array structure from simultaneously meeting the requirements of target imaging region coverage integrity, phase encoding direction receiving sensitivity difference, and decoupling performance between coil units. In the fifth step, the multi-layer dense array structure includes a first layer of coil units arranged along the phase encoding direction, and at least one supplementary coil unit offset along a direction perpendicular to the phase encoding direction or along the normal of the support surface, to improve the spatial sampling density of the target imaging region and reduce noise amplification in parallel imaging reconstruction. Preferably, the single-layer sparse array structure includes multiple coil units disposed on the curved surface of the support, the multiple coil units being distributed non-coplanarly along the circumference of the target imaging region; wherein, at least some coil units are disposed on both sides of the target imaging region in the phase encoding direction to form a differentiated receiving sensitivity distribution along the phase encoding direction; at least some coil units are disposed in one or more supplementary regions on the top, bottom, front, or rear side of the target imaging region to compensate for insufficient receiving sensitivity of the coil units on both sides in local areas; an overlapping region satisfying the decoupling requirements is provided between adjacent coil units, so that the single-layer sparse array structure can reduce the coupling between coil units and improve the channel sensitivity difference required for parallel imaging while satisfying the coverage integrity of the target imaging region.

[0138] In application, the single-layer sparse array structure achieves an excellent balance between imaging performance and system complexity in scenarios with limited channel counts through a simplified layout of "circumferential non-coplanar encirclement + local overlap decoupling." Multiple coil units are non-coplanarly distributed circumferentially along the target imaging area (e.g., the head). Units arranged on both sides of the phase encoding direction form a differentiated sensitivity distribution along the acceleration direction, while supplementary units at the top, bottom, or front and rear compensate for insufficient sensitivity in local areas, ensuring complete coverage of the entire field of view. The controlled overlap between adjacent units effectively suppresses mutual coupling noise through geometric decoupling. Combined with electronic decoupling by the subsequent low-input-impedance preamplifier, excellent channel isolation is maintained while simplifying the array structure. Simultaneously, this structure improves the signal-to-noise ratio of a single channel through a high fill factor, making it particularly suitable for fast scanning scenarios with 8 channels or less and small to medium field of view. Multi-layer dense array structure By employing an "axial / interlayer stacking arrangement," the physical limitations of single-layer sparse arrays in terms of channel count and spatial coverage are effectively overcome. While arranging the first layer of coil units in the phase encoding direction to ensure basic sensitivity differences, at least one supplementary coil unit is added along the normal direction of the support surface (such as the axial or vertical direction), significantly improving the spatial sampling density of the target imaging area. This high-density sampling strategy allows the system to maintain a high acceleration factor while effectively suppressing the noise amplification effect of geometric factors in parallel imaging reconstruction through redundant information from supplementary channels, avoiding image artifacts and texture blurring caused by undersampling. Simultaneously, the multi-layer structure allows adjacent interlayer coil units to maintain a large overlap area for geometric decoupling, and, in conjunction with a low-input-impedance preamplifier, suppresses crosstalk between non-adjacent channels, ensuring that even in a compact three-dimensional arrangement, each channel maintains a high degree of independence and orthogonality. Figure 6 As shown, the equivalent model of two adjacent coil units reveals the physical root of crosstalk between channels: coil one ( ) and coil two ( Mutual induction exists between them. This represents magnetic coupling caused by spatial proximity. Magnetic coupling can lead to mutual signal leakage, impedance mismatch, and a decrease in signal-to-noise ratio. This invention employs precise adjustment of the relative spatial position, rotation angle, and overlap area of ​​the coil units, utilizing the principle of destructive interference of magnetic fields to reduce mutual inductance. Approaching zero, this blocks noise coupling paths between adjacent channels, ensuring that each coil unit operates independently at the physical level, providing a high-purity independent data source for subsequent parallel imaging. The size of the overlap area between adjacent coil units is mainly determined through experimental testing during the design phase. Specifically, a debugging method based on a vector network analyzer can be used: by monitoring the S21 transmission parameters between adjacent coils, the relative position and overlap of the units are dynamically adjusted until the S21 spectrum peak shape remains intact without splitting or deformation, at which point the optimal decoupling state is achieved. The design of this overlap area effectively suppresses mutual coupling noise between channels through geometric decoupling, and ensures that the sensitivity coverage in the phase encoding direction has continuous and smooth gradient differences. This differentiation provides stable and independent spatial encoding information for subsequent parallel imaging algorithms, enabling the system to accurately complete missing data through algorithms even when scanning with a higher acceleration factor, thereby effectively suppressing the noise amplification effect caused by geometric factors and significantly improving the signal-to-noise ratio and clarity of the reconstructed image.

[0139] Example 3:

[0140] The basic content is the same as in Example 1, except that:

[0141] Please see Figure 1 — Figure 5 In the sixth step, the tuning matching parameters corresponding to the Larmor frequency are obtained through a multi-channel joint tuning matching step, which includes: (1) after the coil unit is fixed on the support surface model according to the optimized coil unit position, coil unit rotation angle and overlapping area of ​​adjacent coil units, the resonant frequency, input impedance and reflection parameters of each coil unit in the no-load state and the load state are obtained respectively; (2) according to the Larmor frequency of the target low-field magnetic resonance system and the resonant frequency of the coil unit in the no-load state, the series adjustable capacitor and / or parallel adjustable capacitor in each coil unit are adjusted so that the resonant frequency of each coil unit in the load state is consistent with the Larmor frequency; (3) each coil unit is matched with the tuning matching parameter. (3) Connect the network and adjust the matching network so that the input impedance of each coil unit at the Larmor frequency is converted to the reference impedance range; (4) After completing the tuning matching of any coil unit, detect the transmission parameters between the coil unit and the adjacent coil unit, and when the transmission parameters do not meet the preset decoupling threshold, jointly adjust the matching parameters of the coil unit and the overlapping area of ​​the adjacent coil units until the reflection parameters of the coil unit meet the reflection parameter constraints and the transmission parameters meet the decoupling parameter constraints; (5) Traverse all coil units and repeat (2) to (4) to obtain the tuning capacitance value, matching capacitance value, reference impedance range and corresponding reflection parameters and transmission parameters of each coil unit, that is, obtain the tuning matching parameters corresponding to the Larmor frequency.

[0142] In application, this multi-channel joint tuning and matching mechanism achieves precise calibration and environmental adaptation of coil unit performance at the physical level, ensuring high fidelity and high stability of signal transmission in low-field environments. Firstly, by comparing the resonant frequencies under no-load and load conditions, the impact of the load on coil characteristics is precisely quantified. Based on this, series and / or parallel adjustable capacitors are adjusted to ensure that each coil unit can accurately lock onto the Larmor frequency (2.32MHz in this system) under load, achieving optimal signal sensing efficiency. Secondly, the introduction of the matching network tightly clamps the input impedance within the reference impedance range (e.g., 50 ohms), with an imaginary part of 0, effectively eliminating signal reflection and transmission line standing wave effects, achieving maximum power transfer from the coil to the preamplifier. Crucially, the joint adjustment mechanism in step (4) further detects crosstalk between channels after tuning. By fine-tuning the matching parameters and the coil overlap area, it ensures that the single-channel reflection parameters meet the standards while forcibly satisfying the preset decoupling threshold. The matching parameters include the resonant frequency deviation, the real part of the input impedance obtained under the no-load state of the coil unit, the imaginary part of the input impedance obtained under the no-load state of the coil unit, the reflection parameter S11, the transmission parameter S21 of the adjacent coil, the tuning capacitor value, and the matching capacitor value. Through the closed-loop optimization process of "single-channel tuning - multi-channel decoupling", the consistency of the entire array is ensured after traversing all units, which greatly reduces the system noise floor from the source and provides a high-purity data foundation for subsequent parallel imaging reconstruction based on sensitivity differences.

[0143] Example 4:

[0144] The basic content is the same as in Example 1, except that:

[0145] Please see Figure 1 — Figure 5In the sixth step, the method further includes setting transmit protection detuning parameters for each coil unit. This step includes: determining the active detuning branch parameters and passive detuning branch parameters corresponding to the coil unit based on the tuning capacitor value, matching capacitor value, coil inductance value, and the Larmor frequency of the target low-field magnetic resonance system; connecting a preamplifier, an active detuning branch, and a passive detuning branch to each coil unit. The active detuning branch includes an active control switching device and a first resonant element, and the passive detuning branch includes a passive conduction device and a second resonant element. When the target low-field magnetic resonance system is in the RF transmission phase, the active control switching device is turned on under the action of an external bias control signal, and the first resonant element constructs a parallel resonant circuit. The active detuning branch forms a specified impedance state at the transmission frequency to block the radio frequency transmission energy from entering the corresponding coil unit and its preamplifier. When the active detuning branch is not conducting or the radio frequency transmission energy exceeds a preset threshold, the passive conducting device conducts under the action of the induced voltage across the coil unit. The second resonant element constructs a parallel resonant circuit, causing the passive detuning branch to form a specified impedance state at the transmission frequency to provide redundant protection for the coil unit and its preamplifier. When the target low-field magnetic resonance system is in the signal receiving stage, both the active and passive detuning branches are in a non-detuned state, keeping the coil unit in a tuned matching state corresponding to the Larmor frequency and satisfying the reflection parameter constraints and the decoupling parameter constraints between coil units.

[0146] In application, during the RF transmission phase, the active control switching device in the active detuning branch conducts under the action of an external bias control signal. The first resonant element constructs a parallel resonant circuit, causing the branch to present extremely high impedance at the transmission frequency (Ramohr frequency), forcibly blocking the backflow of high-power transmission energy to the coil unit and the fragile preamplifier, preventing coil overheating or component burnout. If a system control anomaly occurs (such as bias signal failure) or the transmission power exceeds the limit and the RF transmission energy exceeds a preset threshold, the passive conduction device will automatically conduct based on the extremely high RF voltage induced across the coil. The second resonant element similarly constructs a parallel resonant circuit. This creates pure hardware-level redundancy protection, further ensuring the reliability of energy blocking. When the system switches to the signal receiving stage, both the active and passive branches return to a non-detuned state, and the coil unit is strictly maintained in a tuned matching state (the resonant frequency is locked at the Larmor frequency, and the input impedance is matched to the reference impedance range), allowing the weak magnetic resonance signal to be transmitted to the preamplifier without loss. This intelligent switching mechanism of "transmit blocking - receive conduction" not only ensures safety during high-power transmission but also maintains high sensitivity during low-field weak signal reception, perfectly adapting to the imaging requirements of complex electromagnetic interference and low signal-to-noise ratio in unshielded environments.

[0147] Example 5:

[0148] The basic content is the same as in Example 1, except that:

[0149] Please see Figure 1 — Figure 6 In the sixth step, the method further includes a preamplifier decoupling parameter determination step, which includes: determining the target range of the preamplifier input impedance corresponding to each coil unit based on the transmission parameters, mutual inductance parameters, and input impedance of each coil unit under load; ensuring that the input impedance of the preamplifier is lower than the equivalent loop impedance of the corresponding coil unit at the Larmor frequency, so that the crosstalk current generated by the coupling between coil units preferentially flows into the input terminal of the corresponding preamplifier; within the target range of the preamplifier input impedance, jointly constraining the input impedance, noise figure, gain, and stability parameters of the preamplifier, so that the signal-to-noise ratio of each coil unit meets the single-channel signal-to-noise ratio requirement and the transmission parameters between coil units meet the decoupling parameter requirements; when the transmission parameters between any two coil units do not meet the preset decoupling threshold, jointly adjusting the preamplifier input impedance, matching network parameters, and overlapping area of ​​adjacent coil units until the reflection parameters of the coil unit meet the reflection parameter constraints, the transmission parameters of the coil unit meet the transmission parameter constraints, and the signal-to-noise ratio of the coil unit meets the single-channel signal-to-noise ratio constraints.

[0150] In application, this preamplifier decoupling mechanism forms the end current discharge path of the multi-channel receiving array. Based on the virtual ground effect, it blocks the noise coupling path between channels from the source. The preamplifier is precisely configured within an extremely low input impedance target range (typically below 5Ω), significantly lower than the equivalent loop impedance of the coil unit at the Larmor frequency. This allows crosstalk current generated by mutual inductance between adjacent coil units to preferentially flow into the lower impedance input of the preamplifier, rather than forming a harmful closed loop through the coil circuit. In the process of jointly constraining input impedance, noise figure, gain, and stability parameters, not only is the high signal-to-noise ratio receiving performance of a single channel ensured, but also crosstalk between channels is suppressed below the preset decoupling threshold through a low-impedance shunt mechanism at the physical level. Even in a compact multi-layer array arrangement, this dual guarantee of "impedance matching optimization + active noise suppression" can effectively avoid signal aliasing and artifacts.

[0151] Example 6:

[0152] The basic content is the same as in Example 1, except that:

[0153] Please see Figure 1 — Figure 4A receiving array device obtained by a low-field magnetic resonance parallel imaging array design method is disclosed. The device includes a support surface model, multiple coil units, a tuning matching network, a transmit protection detuning circuit, and multiple preamplifiers. The support surface model has a non-planar curved surface structure adapted to the shape of the target imaging region. The multiple coil units are arranged on the support surface model according to the receiving array design parameters and form a non-coplanar surrounding distribution on the support surface model. At least some of the multiple coil units are distributed along the phase encoding direction to form mutually differentiated receiving sensitivity distributions in the phase encoding direction. There is an overlap area between adjacent coil units determined by the receiving array design parameters, so that adjacent coil units meet the preset decoupling parameter requirements. Each coil unit is connected to the tuning matching network, the transmit protection detuning circuit, and the corresponding preamplifier, so that each coil unit meets the tuning matching requirements at the Larmor frequency during the signal reception stage and enters the transmit protection state during the radio frequency transmission stage.

[0154] In application, the device is first fitted onto the area to be tested (e.g., the head). When the MRI system emits a high-power radio frequency pulse, the transmission protection detuning circuit is activated to protect the coil unit. When the radio frequency pulse is emitted and the area to be tested releases a weak magnetic resonance signal, the coil unit senses the magnetic resonance signal. The tuning matching network accurately converts the coil impedance to a reference impedance (e.g., 50Ω), eliminating signal reflection and standing wave effects to achieve maximum power transmission. After matching, the signal is transmitted to the preamplifier, where low input impedance (virtual ground effect) is used to suppress inter-channel crosstalk, while amplifying the weak signal and outputting it to the system receiving link. The multi-coil units distributed along the phase encoding direction provide differentiated sensitivity distribution, and adjacent units achieve geometric decoupling through overlapping areas, ensuring that the signals of each channel are independent and providing high-purity k-space data for subsequent parallel reconstruction.

[0155] Example 7:

[0156] The basic content is the same as in Example 1, except that:

[0157] Please see Figure 1 — Figure 9 A method for optimizing acquisition parameters in the design of a low-field magnetic resonance parallel imaging array, the method comprising the following steps:

[0158] Step 1: Obtain the receiving sensitivity distribution, single-channel signal-to-noise ratio, reflection parameters, and decoupling parameters between coil units in the receiving array within the target imaging area;

[0159] Step 2: Based primarily on the degree of difference in the receiving sensitivity distribution of each coil unit in the phase encoding direction, and secondarily on the single-channel signal-to-noise ratio, reflection parameters, and decoupling parameters between coil units, determine the parallel imaging acquisition parameters. The parallel imaging acquisition parameters include at least one or a combination of external reduction factor, central full sampling area range, phase encoding line sampling interval, and parameter calibration area range.

[0160] Example 8:

[0161] The basic content is the same as in Example 1, except that:

[0162] Please see Figure 1 — Figure 9 In the second step, determining the parallel imaging acquisition parameters includes: determining the candidate range of the external reduction factor based on the degree of difference in the receiving sensitivity distribution of different coil units in the phase encoding direction; when the degree of difference in the receiving sensitivity distribution is lower than a preset sensitivity difference threshold, or the decoupling parameters between coil units do not meet a preset decoupling threshold, decreasing the external reduction factor or increasing the range of the central full sampling area; when the degree of difference in the receiving sensitivity distribution meets a preset sensitivity difference threshold, and the decoupling parameters between coil units meet a preset decoupling threshold, increasing the external reduction factor or decreasing the range of the central full sampling area.

[0163] In application, the determination of parallel imaging acquisition parameters completes the "adaptive trade-off between acceleration factor and image quality" before scanning. First, the sensitivity distribution difference of each coil unit in the phase encoding direction is evaluated. If the difference is significant and the channel decoupling performance is excellent, the external reduction factor is increased to shorten the scanning time and meet the speed requirements of the application. Conversely, if the sensitivity difference is insufficient or there is a risk of channel crosstalk, the acceleration factor is reduced or the central full sampling area is increased to suppress geometric factor noise amplification and ensure that the image is free of artifacts. This dynamic sampling strategy avoids the risk of mismatch of fixed parameters under complex human body load, enabling the low-field MRI system to maximize scanning efficiency while maintaining a high signal-to-noise ratio, perfectly adapting to the diversified clinical needs from rapid screening to fine diagnosis.

[0164] Example 9:

[0165] The basic content is the same as in Example 1, except that:

[0166] Please see Figure 1 — Figure 9Following the second step, there is a third step: The third step involves determining the channel quality parameters of each coil unit based on the single-channel signal-to-noise ratio, reflection parameters, and decoupling parameters between coil units. These channel quality parameters are used to determine the set of coil units participating in parallel imaging reconstruction. Specifically, this includes: scoring each coil unit based on its single-channel signal-to-noise ratio, reflection parameters, decoupling parameters between coil units, and receiver sensitivity coverage; reducing the weight of any coil unit in parallel imaging reconstruction or removing it from the set of coil units participating in parallel imaging reconstruction when its quality score is lower than a preset channel quality threshold; and adjusting the channel weight, acquisition parameters, or reconstruction parameters corresponding to any coil unit when its decoupling parameters with adjacent coil units do not meet a preset decoupling threshold to reduce the impact of channel crosstalk on parallel imaging reconstruction.

[0167] In application, a comprehensive quality score is performed on each coil unit based on the single-channel signal-to-noise ratio, reflection parameters, decoupling parameters, and receiver sensitivity coverage. If a channel's performance degrades due to load effects or poor contact (e.g., excessively low signal-to-noise ratio, excessive reflection, or excessive crosstalk), its weight in parallel imaging reconstruction is reduced or it is directly excluded from the reconstruction set. This effectively avoids the risk of systematic noise and artifacts introduced by the degraded channel. At the same time, for channels whose decoupling parameters do not meet the standards, their channel weights and acquisition parameters are adjusted jointly (e.g., reducing the acceleration factor or increasing the central sampling area) to further suppress the negative impact of inter-channel crosstalk on image quality. This "selective use and dynamic weighting" strategy not only significantly improves the robustness and stability of parallel imaging reconstruction, but also ensures that even with fluctuations in the performance of some channels, high-quality clinical images with high signal-to-noise ratio and no aliasing can still be output in low-field environments.

[0168] Example 10:

[0169] The basic content is the same as in Example 1, except that:

[0170] Please see Figure 1 — Figure 9Following the third step, there is a fourth step: The fourth step involves determining the parallel imaging reconstruction parameters for frequency space data completion based on the receiver sensitivity distribution, parallel imaging acquisition parameters, and channel quality parameters. The determination of the parallel imaging reconstruction parameters includes: determining the neighborhood size, number of participating channels, and channel weights of the parallel imaging reconstruction model based on the central full sampling area range, the outer reduction factor, the set of coil units participating in parallel imaging reconstruction, the channel quality parameters, and the degree of difference in receiver sensitivity distribution of each coil unit in the phase coding direction; when the channel quality parameters are lower than a preset channel quality threshold, increasing the parameter calibration area range, decreasing the channel weight of the corresponding coil unit, or adjusting the neighborhood size of the parallel imaging reconstruction model; when the image quality evaluation parameters after parallel imaging reconstruction do not meet the preset image quality threshold, adjusting the outer reduction factor, the central full sampling area range, the number of participating channels, or the channel weights, and re-determining the parallel imaging reconstruction parameters.

[0171] In application, the neighborhood size, number of participating channels, and channel weights of the parallel imaging reconstruction model are dynamically determined based on the central full-sampling area, the external reduction factor, the set of coil units participating in parallel imaging reconstruction, channel quality parameters, and the degree of difference in the receiving sensitivity distribution of each coil unit in the phase encoding direction. When channel quality parameters are detected to be lower than preset thresholds (e.g., low single-channel signal-to-noise ratio, poor reflection parameters, or unmet decoupling parameters) or the image quality evaluation parameters after parallel imaging reconstruction are not up to standard, the external reduction factor is reduced or the central full-sampling area is increased, at the cost of some scanning speed, to suppress geometric factor noise amplification and ensure that the image is free of artifacts and has clear texture. If the image quality meets the standard, the channel quality is good (e.g., complete sensitivity coverage, decoupling parameters meet the threshold), and it is desirable to further shorten the scanning time, the external reduction factor is increased and the number of phase encoding line samples is reduced to improve scanning efficiency while ensuring image quality. This "sensitivity-driven—quality feedback—parameter iteration" strategy enables the ultra-low field MRI system to output high signal-to-noise ratio and high-definition clinical diagnostic images in a very short scanning time.

[0172] Example 11:

[0173] The basic content is the same as in Example 1, except that:

[0174] like Figure 9As shown in the figure, this figure is a comparison of the application of the present invention and the control group. Based on the overall comparison figure (a) and the local magnified figure (b), the image obtained by the present invention has higher edge sharpness and clearer texture details. In the control group image, the undersampled image shows "artifact patches" in some areas, and the image details after normal sampling + Kalman filtering are still covered by noise, resulting in blurring. The two images using the GRAAPPA algorithm or the Nonlinear GRAAPPA algorithm have lower local detail restoration than the present invention.

[0175] Quantitative comparison of Peak Signal-to-Noise Ratio (PSNR) and Structural Similarity Index (SSIM): PSNR reflects the relative relationship between signal strength and noise in an image. A higher PSNR indicates stronger noise suppression and higher signal fidelity. In this invention, PSNR increased from 31.38 dB in the control group to 33.13 dB (an increase of 1.75 dB), indicating a significant enhancement in signal purity after noise reduction, further suppressing noise interference in the image. SSIM reflects the overall similarity of image brightness, contrast, and structure. A value closer to 1 indicates stronger structural consistency with the "real image (normal sampling)." In this invention, SSIM increased from 0.86 in the control group to 0.91, indicating more accurate reconstruction of the image's anatomical structure, and no significant "structural distortion" (such as artifacts, blurring, or blocky effects) occurred during parallel acceleration.

[0176] The above description is only a preferred embodiment of the present invention. The scope of protection of the present invention is not limited to the above embodiments. Any equivalent modifications or changes made by those skilled in the art based on the content disclosed in the present invention should be included within the scope of protection set forth in the claims.

Claims

1. A design method for a low-field magnetic resonance parallel imaging array, characterized in that: The method includes the following steps: Step 1: Obtain the system parameters of the target low-field magnetic resonance system, including the main magnetic field direction, Larmor frequency, target imaging area, phase encoding direction, number of receiving channels, and target parallel imaging acceleration factor; Step 2: Establish a support surface model that matches the shape of the target imaging area, and generate multiple candidate coil units on the support surface model; wherein, the adjustable parameters of the candidate coil units include at least one or a combination of the following: number of coil units, shape of coil units, area of ​​coil units, position of coil units, rotation angle of coil units, number of turns of coil units, overlapping area of ​​adjacent coil units, and number of coil unit layers. Step 3: Adjust the position of the candidate coil unit according to the direction of the main magnetic field so that the radio frequency receiving field component generated by the candidate coil unit in the target imaging area satisfies the orthogonal receiving condition with the direction of the main magnetic field; Step 4: Based on the phase coding direction and the target parallel imaging acceleration factor, construct an array optimization objective function; the array optimization objective function includes at least two of the following: increasing the difference in receiving sensitivity between different coil units in the phase coding direction, improving the synthetic sensitivity coverage integrity of the target imaging area, reducing the coupling degree between coil units, and improving the single-channel signal-to-noise ratio; Step 5: Using the array optimization objective function as the optimization objective, and with constraints such as the orthogonality of the radio frequency receiving field, the coverage integrity of the target imaging area, the decoupling performance between coil units, the manufacturability of coil units, and the number of receiving channels, the adjustable parameters of the candidate coil units are optimized to obtain a set of coil units that meet the requirements of parallel imaging. Step 6: Output the receiving array design parameters according to the coil unit set. The receiving array design parameters include at least one or a combination of the following: number of coil units, shape of coil units, area of ​​coil units, position of coil units, rotation angle of coil units, number of turns of coil units, overlapping area of ​​adjacent coil units, number of coil unit layers, and tuning matching parameters corresponding to the Larmor frequency. Then design the receiving array according to the receiving array design parameters.

2. The design method for a low-field magnetic resonance parallel imaging array according to claim 1, characterized in that: The fifth step, optimizing the adjustable parameters of the candidate coil units, includes an array level selection step: The number of effective coil units arranged along the phase encoding direction is determined based on the size of the target imaging region in the phase encoding direction, the target parallel imaging acceleration factor, and the number of receiving channels. When the number of receiving channels and the number of effective coil units can meet the requirements of target imaging area coverage integrity, phase coding direction receiving sensitivity difference and decoupling performance between coil units on the single-layer support curved surface, a single-layer sparse array structure is selected. When the number of receiving channels or the target parallel imaging acceleration factor makes it impossible for a single-layer sparse array structure to simultaneously meet the requirements of target imaging area coverage integrity, phase coding direction receiving sensitivity difference, and decoupling performance between coil units, a multi-layer dense array structure is selected.

3. The low-field magnetic resonance parallel imaging array design method according to claim 2, characterized in that: In the fifth step, the multi-layer dense array structure includes a first layer of coil units arranged along the phase encoding direction, and at least one supplementary coil unit offset along the direction perpendicular to the phase encoding direction or along the normal of the support surface, so as to improve the spatial sampling density of the target imaging area and reduce noise amplification in parallel imaging reconstruction.

4. A method for designing a low-field magnetic resonance parallel imaging array according to claim 1 or 2, characterized in that: In the sixth step, the tuning matching parameters corresponding to the Larmor frequency are obtained through a multi-channel joint tuning matching step, which includes: (1) After the coil unit is fixed on the support surface model according to the optimized coil unit position, coil unit rotation angle and adjacent coil unit overlap area, the resonant frequency, input impedance and reflection parameters of each coil unit in the no-load state and the load state are obtained respectively. (2) Based on the Larmor frequency of the target low-field magnetic resonance system and the resonant frequency of the coil unit under no-load conditions, adjust the series adjustable capacitor and / or parallel adjustable capacitor in each coil unit so that the resonant frequency of each coil unit under load conditions is consistent with the Larmor frequency. (3) Connect each coil unit to the matching network and adjust the matching network so that the input impedance of each coil unit at the Larmor frequency is converted to the reference impedance range; (4) After completing the tuning and matching of any coil unit, the transmission parameters between the coil unit and the adjacent coil unit are detected, and when the transmission parameters do not meet the preset decoupling threshold, the matching parameters of the coil unit and the overlapping area of ​​the adjacent coil units are jointly adjusted until the reflection parameters of the coil unit meet the reflection parameter constraints and the transmission parameters meet the decoupling parameter constraints. (5) Traverse all coil units and repeat (2) to (4) to obtain the tuning capacitance value, matching capacitance value, reference impedance range, and corresponding reflection and transmission parameters of each coil unit, thus obtaining the tuning matching parameters corresponding to the Larmor frequency.

5. The design method for a low-field magnetic resonance parallel imaging array according to claim 4, characterized in that: In the sixth step, the method further includes setting transmit protection detuning parameters for each coil unit, the step including: Based on the tuning capacitor value, matching capacitor value, coil inductance value, and Larmor frequency of the target low-field magnetic resonance system of the coil unit, determine the active detuning branch parameters and passive detuning branch parameters corresponding to the coil unit. A preamplifier, an active detuning branch, and a passive detuning branch are connected to each coil unit. The active detuning branch includes an active control switching device and a first resonant element, and the passive detuning branch includes a passive conduction device and a second resonant element. When the target low-field magnetic resonance system is in the radio frequency transmission stage, the active control switching device is turned on under the action of the external bias control signal, and the first resonant element constructs a parallel resonant circuit, so that the active detuning branch forms a specified impedance state at the transmission frequency to block the radio frequency transmission energy from entering the corresponding coil unit and its preamplifier. When the active detuning branch is not turned on or the radio frequency transmission energy exceeds a preset threshold, the passive turning device turns on under the action of the induced voltage at both ends of the coil unit. The second resonant element constructs a parallel resonant circuit, so that the passive detuning branch forms a specified impedance state at the transmission frequency, so as to provide redundant protection for the coil unit and its preamplifier. When the target low-field magnetic resonance system is in the signal receiving stage, both the active detuning branch and the passive detuning branch are in a non-detuned state, so that the coil unit is maintained in a tuned matching state corresponding to the Larmor frequency, and the reflection parameter constraint and the decoupling parameter constraint between coil units are satisfied.

6. The design method for a low-field magnetic resonance parallel imaging array according to claim 5, characterized in that: In the sixth step, the method further includes a preamplifier decoupling parameter determination step, which includes: Based on the transmission parameters, mutual inductance parameters, and input impedance of each coil unit under load, determine the target range of the preamplifier input impedance corresponding to each coil unit. The input impedance of the preamplifier is made lower than the equivalent loop impedance of the corresponding coil unit at the Larmor frequency, so that the crosstalk current generated by the coupling between the coil units flows preferentially into the input terminal of the corresponding preamplifier. Within the target range of the input impedance of the preamplifier, the input impedance, noise figure, gain, and stability parameters of the preamplifier are jointly constrained to ensure that the signal-to-noise ratio of each coil unit meets the single-channel signal-to-noise ratio requirement and the transmission parameters between coil units meet the decoupling parameter requirements. When the transmission parameters between any two coil units do not meet the preset decoupling threshold, the input impedance of the preamplifier, the matching network parameters, and the overlap area of ​​adjacent coil units are jointly adjusted until the reflection parameters of the coil units meet the reflection parameter constraints, the transmission parameters of the coil units meet the transmission parameter constraints, and the signal-to-noise ratio of the coil units meets the single-channel signal-to-noise ratio constraints.

7. A method for optimizing acquisition parameters in the design method of a low-field magnetic resonance parallel imaging array as described in claim 1, characterized in that: The method includes the following steps: Step 1: Obtain the receiving sensitivity distribution, single-channel signal-to-noise ratio, reflection parameters, and decoupling parameters between coil units in the receiving array within the target imaging area; Step 2: Based primarily on the degree of difference in the receiving sensitivity distribution of each coil unit in the phase encoding direction, and secondarily on the single-channel signal-to-noise ratio, reflection parameters, and decoupling parameters between coil units, determine the parallel imaging acquisition parameters. The parallel imaging acquisition parameters include at least one or a combination of external reduction factor, central full sampling area range, phase encoding line sampling interval, and parameter calibration area range.

8. The method for optimizing acquisition parameters in the design of a low-field magnetic resonance parallel imaging array according to claim 7, characterized in that: In the second step, determining the parallel imaging acquisition parameters includes: The candidate range of the external reduction factor is determined based on the degree of difference in the receiving sensitivity distribution of different coil units in the phase coding direction; When the difference in the receiving sensitivity distribution is lower than the preset sensitivity difference threshold, or the decoupling parameters between coil units do not meet the preset decoupling threshold, the external reduction factor is reduced or the range of the central full sampling area is increased. When the degree of difference in the receiving sensitivity distribution meets the preset sensitivity difference threshold, and the decoupling parameters between coil units meet the preset decoupling threshold, the external reduction factor is increased or the range of the central full sampling area is reduced.

9. The method for optimizing acquisition parameters in the design method of a low-field magnetic resonance parallel imaging array according to claim 7, characterized in that: There is a third step after the second step: Step 3: Based on the single-channel signal-to-noise ratio, reflection parameters, and decoupling parameters between coil units, determine the channel quality parameters of each coil unit; these channel quality parameters are used to determine the set of coil units participating in parallel imaging reconstruction, specifically including: Each coil unit is scored based on its single-channel signal-to-noise ratio, reflection parameters, decoupling parameters between coil units, and receiver sensitivity coverage. When the quality score of any coil unit is lower than the preset channel quality threshold, the weight of that coil unit in parallel imaging reconstruction is reduced, or the coil unit is removed from the set of coil units participating in parallel imaging reconstruction. When the decoupling parameters of any coil unit and its adjacent coil units do not meet the preset decoupling threshold, the channel weights and acquisition parameters corresponding to that coil unit are adjusted to reduce the impact of channel crosstalk on parallel imaging reconstruction.

10. The method for optimizing acquisition parameters in a low-field magnetic resonance parallel imaging array design method according to claim 9, characterized in that: There is a fourth step following the third step: Step 4: Based on the received sensitivity distribution, parallel imaging acquisition parameters, and channel quality parameters, determine the parallel imaging reconstruction parameters used for frequency space data completion; the determination of the parallel imaging reconstruction parameters includes: Based on the central full sampling area range, the external reduction factor, the set of coil units participating in parallel imaging reconstruction, the channel quality parameters, and the degree of difference in the receiving sensitivity distribution of each coil unit in the phase coding direction, the neighborhood size, the number of participating channels, and the channel weights of the parallel imaging reconstruction model are determined. When the channel quality parameter is lower than the preset channel quality threshold, increase the parameter calibration area range, reduce the channel weight of the corresponding coil unit, or adjust the neighborhood size of the parallel imaging reconstruction model. When the image quality evaluation parameters after parallel imaging reconstruction do not meet the preset image quality threshold, the external reduction factor, the range of the central full sampling area, the number of participating channels or the channel weights are adjusted, and the parallel imaging reconstruction parameters are redefined.

Images