Magnetic particle imaging reconstruction method based on multi-angle Cartesian scanning adaptive optimization

Abstract

The invention belongs to the technical field of magnetic particle imaging, relates to a magnetic particle imaging reconstruction method based on multi-angle Cartesian scanning adaptive optimization, and aims to solve the problems of easy information redundancy, low resolution and insufficient response signal representation under the condition of low gradient in free point scanning in the existing imaging technology. The method comprises the following steps: setting magnetic field generation related parameters and a measurement mode of a rotating Cartesian angle scanning track of a magnetic field free point; constructing a system function harmonic diagram under each rotation angle; constructing a system matrix under each rotation angle, analyzing the system matrix under each angle and the weight and contribution of each frequency component in the system matrix, and determining an optimal system matrix; constructing an observation vector sequence; and reconstructing a magnetic particle image. According to the method, image reconstruction is carried out through the Cartesian scanning track of the free point of the rotating magnetic field and the weighted multi-angle system matrix, the system matrix and particle response representation under the low gradient are enriched, and the imaging resolution is improved.

Description

Background Art

[0002] In recent years, the development of non-invasive in-vivo molecular imaging technology has been rapid, enabling effective observation of various biomolecules in the body. Among them, magnetic particle imaging technology, as an emerging technology based on the non-linear magnetization response of magnetic nanoparticles, has attracted much attention. Compared with traditional imaging methods, magnetic particle imaging technology has significant advantages. It has high sensitivity and high spatial resolution. When detecting small lesions, it can accurately identify subtle differences; it has no radiation hazard, is safe for the human body, and is suitable for multiple examinations; it has no background signal interference, the images are clear, avoiding interference with diagnosis; it has no imaging depth limitation, can penetrate deep into the body to obtain information, and has great application potential especially in the detection of deep small tumors and the monitoring of cardiovascular and cerebrovascular functions.

[0003] Magnetic particle imaging technology detects the concentration distribution of magnetic nanoparticle tracers through a combination of static and dynamic magnetic fields. The static magnetic field (gradient magnetic field) contains low magnetic field regions, such as elliptical dot-shaped magnetic field free points or linear magnetic field free lines. The dynamic field is a uniformly alternating excitation magnetic field, which is used to drive the low magnetic field region to move within the imaging field of view and excite the magnetic nanoparticles to generate non-linear characteristic signals, thereby realizing the spatial encoding of the magnetic nanoparticle concentration. Among them, although the magnetic field free point has a small low magnetic field range, it has accurate positioning, diverse scanning trajectories, and high imaging resolution, and occupies an important position in research and applications.

[0004] In the process of reconstructing magnetic particle imaging from signals to images, the system matrix method is a key technology. Its main process is to measure the signal spectrum of a unit sample at each pixel point in the imaging field of view, filter and then construct a system matrix in the frequency domain, quantify the signal of the object to be measured as an observation vector, and establish and solve a linear equation system to reconstruct the image. In the prior art, for a magnetic particle imaging system based on magnetic field free points, it is necessary to drive the magnetic field free point to cover the entire imaging field of view to accurately reconstruct; the traditional method uses a combination of an excitation magnetic field and a driving magnetic field to drive the magnetic field free point to cover the imaging field of view according to a preset trajectory (such as Cartesian, Lissajous scanning trajectories). However, these conventional scanning trajectories have many problems: in the case of low gradients, the magnetic field free point is large, the particle response signal is weak, and there is a lack of useful information in the reconstruction process, resulting in a decrease in resolution; moreover, the conventional preset scanning trajectories can only make the magnetic field free point scan each grid point in the field of view at one angle. In one scanning cycle, although the magnetic field free point passes through the same grid point multiple times, the obtained time-domain signals have high similarity and redundancy in information, lacking the signal representation generated by the magnetic particles under excitation at other angles, resulting in a low imaging effect.

[0005] Therefore, in order to improve the resolution and imaging effect of magnetic particle imaging, it is urgent to propose a more effective magnetic field free point imaging method based on the system matrix. Summary of the Invention

[0006] To solve the above problems in the prior art, that is, to solve the problems of information redundancy and low resolution in free-point scanning of magnetic particle imaging in the prior art, and insufficient characterization of particle concentration-response signals under low gradient conditions, the present invention proposes a magnetic particle imaging reconstruction method based on multi-angle Cartesian scanning adaptive optimization, which is applied to a three-dimensional dual-excitation channel symmetric magnetic particle imaging device based on a magnetic field free point. The image is reconstructed by rotating the Cartesian scanning trajectory of the magnetic field free point and weighting the system matrix at multiple angles. The method includes: Step S100: Set scanning parameters, and turn on and run the three-dimensional dual-excitation channel symmetric magnetic particle imaging device based on the magnetic field free point; Step S200: Take the low-frequency movement direction of the Cartesian scanning trajectory as the main movement direction of the magnetic field free point, take the angle between the main movement direction of the magnetic field free point and the perpendicular bisector of the receiving coil in the x direction as the scanning trajectory angle, and set the rotation angle and rotation order of the Cartesian scanning trajectory; Rotate the magnetic field free point according to the rotation angle and the rotation order; Step S300: Measure the signal of the unit sample at each pixel within the imaging field of view, and then construct the system function harmonic diagram at each rotation angle; Step S400: Calibrate the system matrix at different angles based on the system function harmonic diagram, perform frequency weighting selection at a single angle and multi-angle adaptive angle selection, determine the two-dimensional weight factor, and obtain the optimal system matrix at each rotation angle based on the two-dimensional weight factor; Step S500: Measure the response signal of the target object to be imaged at each rotation angle, and construct the observation vector sequence at each rotation angle based on the two-dimensional weight factor; Step S600: Combine the observation vector sequences and the optimal system matrices at each rotation angle by columns respectively. After combination, construct a linear equation system; Solve the linear equation system to obtain the reconstructed magnetic particle image corresponding to the target object to be imaged.

[0007] In some preferred embodiments, the method for setting the rotation angle of the Cartesian scanning trajectory is as follows: Assume that the basic trajectory of the Cartesian scanning trajectory is a grid movement along the x and y directions, and the specific expression is as follows: ; ; where, and respectively represent the driving magnetic fields in the x and y directions, A represents the magnetic field amplitude, and are the change frequencies of the trajectory in the x and y directions respectively; When Obtain a basic Cartesian trajectory excited along the x direction; Set the rotation parameters : ; Perform trigonometric function processing on the dual-axis excitation magnetic fields of the X-axis and Z-axis with respect to the angle, and perform a rotation transformation based on the basic Cartesian trajectory to obtain a rotational Cartesian trajectory with the main direction of the magnetic field free point being : ; Among them, is the rotation angle, and the rotational Cartesian trajectories at multiple angles are measured by setting a set of rotation angles .

[0008] In some preferred embodiments, the rotation order of the Cartesian scan trajectory is set, and the method is as follows: Taking the initial direction of the basic Cartesian trajectory as 0°, stepwise rotate at equal intervals or unequal intervals to ensure that the movement range of the magnetic field free point can cover the entire imaging field of view, and rotate a total of N angles; where N represents the set number.

[0009] In some preferred embodiments, the optimal system matrix at each rotation angle is obtained, and the method is as follows: Step S410: Obtain the system matrix at each rotation angle of the Cartesian scan trajectory , is the rotation angle, ; Step S420: Perform a selection operation on the system matrices at each rotation angle according to the constructed evaluation function to obtain the weights of each angle; The selection operation includes elimination, low weight value or merging; the evaluation function includes a contribution degree evaluation function and a redundancy evaluation function; Step S430: Analyze the weights and contributions of the respective frequency components of the system matrix at a single angle to determine the frequency weighting matrix at this rotation angle; Step S440: Multiply the weights of each angle by the corresponding frequency weighting matrix to obtain a two-dimensional weight factor , and globally weight each system matrix based on the two-dimensional weight factor to generate the optimal system matrix at each rotation angle .

[0010] In some preferred embodiments, the weights of each angle are obtained, and the method is as follows: Step S421: Perform a contribution degree evaluation on the system matrices at each rotation angle, and the contribution degree of each angle is: ; Among them, is the system matrix at an angle of , where , is the finally reconstructed image; Step S422: Evaluate the redundancy of the system matrices at each rotation angle. The redundancy of each angle is: ; Among them, are two different rotation angles, , are the system matrices at angles of , ; Step S423: Screen the system matrices at each angle according to the contribution degree and redundancy, and set dynamic thresholds , , and screen out the angles whose evaluation is higher than the predetermined threshold and assign weights; ; Then, the system matrix at each angle is: ; Among them, , is the weighted matrix for multi-angle screening, that is, the angle screening weight. The vector in the matrix is the weight of each angle, , .

[0011] In some preferred embodiments, a linear equation system is constructed: ; Among them, is the response signal of the target object to be imaged at each rotation angle; C represents the discrete vector form of the target object to be imaged.

[0012] Advantages of the present invention: The present invention reconstructs images by means of the Cartesian scanning trajectory of the free points of the rotating magnetic field and the system matrices under weighted multi-angles, reduces redundant information and computational complexity, while maintaining high sensitivity and image quality, optimizes the reconstruction efficiency, and the finally generated comprehensive system matrix is more comprehensive in terms of direction information coverage, and the reconstruction effect is better than the traditional method under low gradient fields, providing higher-quality image data for the application of magnetic particle imaging technology; By adaptively screening and weighting the angles and frequency components that have a great impact on the imaging quality, the resolution of magnetic particle imaging is significantly improved, and the spatial resolution under low gradient fields is significantly improved; The present invention introduces a multi - angle Cartesian trajectory detection method. By setting a rotational Cartesian angle scanning trajectory and related parameters based on the magnetic field free point, and rotating the Cartesian scanning trajectory of the magnetic field free point, it effectively enriches the system matrix and particle response characterization at low gradients, and solves the problem of insufficient characterization of particle concentration - response signal in existing methods at low gradients. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] Other features, objects, and advantages of the present application will become more apparent by reading the detailed description of the non - restrictive embodiments with reference to the following drawings: Figure 1 is a flowchart of the magnetic particle imaging reconstruction method based on multi - angle Cartesian scanning adaptive optimization of the present invention; Figure 2 is a schematic structural diagram of a three - dimensional dual - excitation channel symmetric magnetic particle imaging device based on the magnetic field free point of the present invention Figure 3 is a schematic diagram of the magnet structure of a three - dimensional dual - excitation channel symmetric magnetic particle imaging device based on the magnetic field free point of the present invention; Figure 4 is a time - domain diagram of the Cartesian scanning trajectory when the rotation angle is 0° in an embodiment of the present invention; Figure 5 is a time - domain diagram of the Cartesian scanning trajectory when the rotation angle is 30° in an embodiment of the present invention; Figure 6 is a time - domain diagram of the Cartesian scanning trajectory when the rotation angle is 60° in an embodiment of the present invention; Figure 7 is a time - domain diagram of the Cartesian scanning trajectory when the rotation angle is 90° in an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0014] The present application will be further described in detail below with reference to the drawings and embodiments. It can be understood that the specific embodiments described herein are only used to explain the related invention and do not limit the invention. Additionally, it should be noted that for the sake of description, only parts related to the relevant invention are shown in the drawings.

[0015] It should be noted that, without conflict, the embodiments in the present application and the features in the embodiments can be combined with each other. The present application will be described in detail below with reference to the drawings and embodiments.

[0016] The present invention provides a magnetic particle imaging reconstruction method based on adaptive optimization of multi-angle Cartesian scanning. In the existing magnetic particle imaging method, the system matrix has insufficient characterization of particle concentration-response signals during the measurement process at low gradients, severely limiting the imaging resolution of magnetic particle imaging devices. At the same time, in order to reduce redundant information and computational complexity and optimize the reconstruction efficiency, the present invention reconstructs images by rotating the Cartesian scanning trajectory of the magnetic field free point and weighting the system matrix at multiple angles, enriching the system matrix and particle response characterization at low gradients and improving the resolution of magnetic particle imaging.

[0017] A magnetic particle imaging reconstruction method based on adaptive optimization of multi-angle Cartesian scanning according to the present invention, the method comprising the following steps: Step S100, set scanning parameters, turn on and run the three-dimensional dual-excitation channel symmetric magnetic particle imaging device based on the magnetic field free point; Step S200, take the low-frequency movement direction of the Cartesian scanning trajectory as the main movement direction of the magnetic field free point, take the angle between the main movement direction of the magnetic field free point and the perpendicular bisector of the receiving coil in the x direction as the scanning trajectory angle, set the rotation angle and rotation order of the Cartesian scanning trajectory; rotate the magnetic field free point according to the rotation angle and the rotation order; Step S300, measure the signal of the unit sample at each pixel within the imaging field of view, and then construct the system function harmonic diagram at each rotation angle; Step S400, calibrate the system matrix at different angles based on the system function harmonic diagram, perform frequency weighting selection at a single angle and multi-angle adaptive angle selection, determine the two-dimensional weight factor, and obtain the optimal system matrix at each rotation angle based on the two-dimensional weight factor; Step S500, measure the response signal of the target object to be imaged at each rotation angle, and construct the observation vector sequence at each rotation angle based on the two-dimensional weight factor; Step S600, merge the observation vector sequences and the optimal system matrices at each rotation angle by columns respectively. After merging, construct a linear equation system; solve the linear equation system to obtain the reconstructed magnetic particle image corresponding to the target object to be imaged.

[0018] To more clearly illustrate the magnetic particle imaging reconstruction method based on adaptive optimization of multi-angle Cartesian scanning of the present invention, the following combines Figure 1 Expand and detail each step in the embodiment of the present invention.

[0019] The magnetic particle imaging reconstruction method based on adaptive optimization of multi-angle Cartesian scanning in the first embodiment of the present invention, the method comprising the following steps S100 to S600, and is applied to a three-dimensional dual-excitation channel symmetric magnetic particle imaging device based on the magnetic field free point.

[0020] The three-dimensional dual-excitation channel symmetric magnetic particle imaging device based on the magnetic field free point is as Figure 2 shown, and is composed of four parts: a gradient end, a drive and excitation end, a receiving end, and a compensation end; the gradient end is composed of a pair of Maxwell coils and two iron cores. The two coils work in an energization mode of independent power supply of low-frequency alternating current superimposed with direct current bias, and are responsible for generating the magnetic field free point (FFP) and driving the FFP to move along the y direction; the drive and excitation end can apply high-frequency and low-frequency alternating current simultaneously, and is responsible for the two pairs of Helmholtz coils that move the FFP and excite the magnetic particles to generate non-linear response signals; the receiving end is arranged in the x and z axis directions. The x-axis receiving coil is cylindrical, and the z-axis is a saddle-shaped coil that is symmetrically uniform up and down. Both are made of 0.4mm diameter Litz wire, and are responsible for receiving the response signals of the particles in these two directions respectively; the compensation coils are divided into an external excitation compensation in the x direction and an internal compensation coil in the z direction to eliminate the direct feedthrough signals generated by the excitation end at the receiving end.

[0021] As Figure 3 shown, a pair of coils in the gradient end responsible for generating the magnetic field free point (FFP) and driving the FFP to move along the y direction are respectively named Cy1 and Cy2 coils. The two pairs of coils in the drive and excitation end responsible for the two-dimensional movement of the FFP are respectively named Cx and Cz coils. The magnetic fields generated by the three pairs of coils can realize the three-dimensional movement of the FFP. Among them, the magnetic fields generated by the Cx and Cz coils, in addition to being responsible for driving the FFP to perform two-dimensional movement, can also perform high-frequency excitation to generate MPI excitation signals; the receiving coils are named Rx and Rz, and their directions are the same as those of Cx and Cz, and are responsible for inducing the MPI signals of the magnetic nanoparticles; the external excitation compensation coil in the x direction of the compensation end is named Sx, and the internal compensation coil in the z direction is named Sc, which is responsible for eliminating the direct feedthrough signals of the Cx and Cz excitation signals on Rx and Rz.

[0022] Step S100: Set the scanning parameters, and turn on and run the three-dimensional dual-excitation channel symmetric magnetic particle imaging device based on the magnetic field free point.

[0023] Preferably, in this embodiment, a fixed dual-channel lower excitation field is used, and the excitation field magnetic field is set to realize multi-angle rotation of the magnetic field free point for Cartesian measurement; the method for setting the rotation angle of the above-mentioned Cartesian scanning trajectory is as follows: First, define the basic Cartesian trajectory: Let the basic trajectory of the Cartesian scanning trajectory be a grid movement along the x and y directions, and the specific expression is as follows: ; ; Among them, and respectively represent the driving magnetic fields in the x and y directions, A represents the magnetic field amplitude, and are respectively the change frequencies of the trajectory in the x and y directions; when a basic Cartesian trajectory excited along the x direction is obtained; Set the rotation parameter and introduce: ; Perform trigonometric function processing on the dual-axis excitation magnetic fields of the X-axis and Z-axis with respect to angle, and perform a rotation transformation based on the basic Cartesian trajectory to obtain a rotation Cartesian trajectory with the main direction of the magnetic field free point being : ; wherein, is the rotation angle, and the rotation Cartesian trajectories at multiple angles are measured by setting a set of rotation angles for measurement.

[0024] Step S200: Take the low-frequency motion direction of the Cartesian scan trajectory as the main motion direction of the magnetic field free point, take the angle between the main motion direction of the magnetic field free point and the perpendicular bisector of the x-direction receiving coil as the scan trajectory angle, set the rotation angle and rotation order of the Cartesian scan trajectory; rotate the magnetic field free point according to the rotation angle and the rotation order.

[0025] Preferably, take the initial direction of the basic Cartesian trajectory as 0°, and stepwise rotate in an equal-spacing or unequal-spacing manner to ensure that the motion range of the magnetic field free point can cover the entire imaging field of view, and rotate a total of N angles; where N represents the set number. Refer to Figures 4 - 7 , in this embodiment, it is preferably rotated to 90°, N is 3, and each rotation is 30°.

[0026] Step S300: Measure the signal of the unit sample at each pixel within the imaging field of view, and then construct the system function harmonic diagram at each rotation angle. The method is as follows: Measure the time-domain signal of the unit sample at each pixel, and obtain the spectrum through fast Fourier transform; according to the signal-to-noise ratio of the magnetic particle imaging device based on the magnetic field free point, screen the harmonic signals with a signal-to-noise ratio greater than the set signal-to-noise ratio threshold, and use the screened harmonic signals as the first harmonic signals; based on the spectrum, synthesize the first harmonic signals of the same order at each pixel to obtain the system function harmonic diagrams of different orders.

[0027] Step S400: calibrating the system function harmonic graph to obtain system matrices at different angles, performing frequency weighted selection at a single angle and multi-angle adaptive angle selection, determining a two-dimensional weight factor, and obtaining the optimal system matrix at each rotation angle based on the two-dimensional weight factor; Preferably, the optimal system matrix at each rotation angle is obtained by: Step S410: Obtain the system matrix at each rotation angle of the Cartesian scanning trajectory , is the rotation angle, ; Step S420: performing a selection operation on the system matrix at each rotation angle according to the constructed evaluation function to obtain the weight of each angle; The selection operation includes elimination, low weight or merging; the evaluation function includes a contribution evaluation function and a redundancy evaluation function.

[0028] Preferably, the weight of each angle is obtained by: Step S421: Evaluate the contribution of the system matrix at each rotation angle. for: ; in, The angle is The system matrix of , is the final reconstructed image; Step S422: Redundancy evaluation of the system matrix at each rotation angle is performed. for: ; in, For two different rotation angles, 、 The angle is 、 The system matrix; Step S423: Screen the system matrix of each angle according to contribution and redundancy, and set the dynamic threshold 、 , filter out reviews Angles above a predetermined threshold are assigned weights; ; Then, the system matrix at each angle is: ; in, , is the weighted matrix for multi - angle screening, that is, the angle screening weight, and the vectors in the matrix are the weights of each angle, , .

[0029] Preferably, in this embodiment, the selection operation includes elimination, low weight value or combination.

[0030] Step S430: Analyze the weights and contributions of the frequency components of the system matrix under a single angle, and determine the frequency - weighted matrix at this rotation angle. The method is as follows: Step S431: Perform a fast Fourier transform on the received signal, extract the frequency components and their amplitude information, analyze the amplitudes of the signal at different frequencies, and determine the main harmonic components of the signal; Use a threshold value to screen the effective frequencies and determine the screening function : ; Wherein, is the dynamic threshold, adjusted based on the noise level, h is the signal amplitude; f 0, f 1 represents the driving - field frequency, k, k 1 represents the modulation coefficient of the modulation signal; Step S432: Extract the corresponding columns from the system matrix according to the selected frequency index, and calculate the weight matrix : ; Wherein, is the signal amplitude of the signal received at the angle of , and adjust the weight based on the signal amplitude or contribution degree, z represents the frequency components higher than the dynamic threshold; Step S433: Obtain the screened system matrix at a certain angle : ; Wherein, is the frequency - weighted matrix of a single angle , and the vectors in the matrix are the weights of each frequency at this angle; is the system matrix at the angle of , where ; Then the frequency - screening weight of the system matrix is: .

[0031] Step S440: Multiply the weights of each angle by the corresponding frequency weighting matrix to obtain a two-dimensional weight factor. Based on the two-dimensional weight factor, globally weight each system matrix to generate an optimal system matrix at each rotation angle. The method is as follows: Determine the two-dimensional weight factor based on the angle screening weight and the frequency screening weight , specifically: ; Based on the two-dimensional weight factor Globally weight the system matrix to generate an optimal system matrix at each rotation angle : ; Among them, .

[0032] By analyzing the system matrices at multiple angles, the signal-to-noise ratio of each frequency component, and the contribution to the final reconstruction quality, filter out the frequency components that have a greater impact on the imaging quality and provide a larger weight factor for the system matrix at high contribution angles, thereby reducing redundant information and computational complexity, and optimizing the reconstruction quality and efficiency; it can significantly improve the spatial resolution under low gradient fields while maintaining high sensitivity and image quality. In addition, the optimal system matrix has a more comprehensive coverage of direction information, and the reconstruction effect will be better than the traditional method under low gradient fields.

[0033] Preferably, optimize the angle screening weight and the frequency screening weight, and further optimize the optimal system matrix as: ; Among them, is the optimal system matrix after multi-angle and frequency screening weighting, y is the observed response signal, and x is the concentration distribution to be solved; is the regularization term, which is used to suppress noise or introduce prior information; is the regularization term of the angle screening weight, which is used to constrain the sparsity or smoothness of the screened angles; is the regularization term of the frequency screening weight, which is used to constrain the sparsity of the screened frequencies; is the angle screening weight, is the frequency screening weight; 、 、 are the regularization weights.

[0034] Further preferably, optimize the angle screening weight and the frequency screening weight. The method is as follows: A. Set up evaluation indicators as the reconstruction quality evaluation criteria. The evaluation indicators include PSNR, SSIM, time resolution, and robustness; B. Initial Angle Weight and Frequency Weight with the Initial Image ; C. Fixed Weight and , and optimize the image by solving the following equation: ; D. Fixed Image x , and optimize the weight by updating the following equation: ; ; E. Alternating Iterative Optimization x , , , until the error converges.

[0035] Step S500: Measure the response signals of the target object to be imaged at each rotation angle, and construct an observation vector sequence at each rotation angle based on the two-dimensional weight factor; Obtain the response signals at each rotation angle of the Cartesian scan trajectory: ; Construct an observation vector sequence at each rotation angle based on the two-dimensional weight factor: ; where, is the response signal received after scanning at each rotation angle, is the rotation angle, .

[0036] Step S600: Combine the observation vector sequences at each rotation angle and the optimal system matrix column by column. After combination, construct a linear equation system; Construct a linear equation system: ; where, , is the response signal of the target object to be imaged at each rotation angle; C represents the discrete vector form of the target object to be imaged; Solve the linear equation system to obtain the reconstructed magnetic particle image corresponding to the target object to be imaged.

[0037] In the above embodiments, although each step is described in the above order, those skilled in the art can understand that in order to achieve the effects of this embodiment, different steps do not have to be executed in such an order. They can be executed simultaneously (in parallel) or in a reversed order, and these simple changes are within the protection scope of the present invention.

[0038] An adaptive optimization magnetic particle imaging reconstruction system based on multi - angle Cartesian scanning according to the second embodiment of the present invention, the system includes a signal acquisition device and a central processing device; The signal acquisition device includes a three - dimensional dual - excitation channel symmetric magnetic particle imaging device based on magnetic field free points; The signal acquisition device is configured to set scanning parameters, the rotation angle and rotation order of the Cartesian scanning trajectory, and rotate the magnetic field free points according to the rotation angle and the rotation order; It is also configured to collect the signals of unit samples at each pixel within the imaging field of view, and collect the response signals of the target object to be imaged at each rotation angle; The central processing device includes a CPU and a GPU; The central processing device includes: A data processing module, configured to construct a system function harmonic diagram at each rotation angle based on the signals of unit samples at each pixel, and then obtain a system matrix at different angles; It is also configured to construct an observation vector sequence based on the above - mentioned response signals; A matrix weighting module, configured to perform frequency weighting selection at a single angle and multi - angle adaptive angle selection on the system matrices at different angles, and obtain an optimal system matrix at each rotation angle; An imaging module, configured to merge the observation vector sequence by columns, merge the optimal system matrices at each rotation angle by columns, and after merging, construct a linear equation system; Solve the linear equation system to obtain the reconstructed magnetic particle image corresponding to the target object to be imaged.

[0039] Those skilled in the art can clearly understand that for the convenience and brevity of description, the specific working process and related descriptions of the above - described system can refer to the corresponding process in the foregoing method embodiment, and will not be repeated here.

[0040] It should be noted that the above - described adaptive optimization magnetic particle imaging reconstruction system based on multi - angle Cartesian scanning is only illustrated by the above - mentioned division of each functional module. In practical applications, the above functions can be allocated to different functional modules according to needs, that is, the modules or steps in the embodiments of the present invention can be further decomposed or combined. For example, the modules in the above - mentioned embodiments can be merged into one module, or further split into multiple sub - modules to complete all or part of the functions described above. For the names of the modules and steps involved in the embodiments of the present invention, they are only used to distinguish each module or step, and are not regarded as an improper limitation of the present invention.

[0041] An electronic device according to the third embodiment of the present invention, includes: At least one processor; and A memory communicatively connected to at least one of the processors; wherein, The memory stores instructions executable by the processor, and the instructions are used to be executed by the processor to implement the above-mentioned magnetic particle imaging reconstruction method based on multi-angle Cartesian scan adaptive optimization.

[0042] A computer-readable storage medium according to a fourth embodiment of the present invention, the computer-readable storage medium stores computer instructions, and the computer instructions are used to be executed by a computer to implement the above-mentioned magnetic particle imaging reconstruction method based on multi-angle Cartesian scan adaptive optimization.

[0043] Those skilled in the art can clearly understand that for the convenience and brevity of description, the specific working processes and related descriptions of the above-described electronic device and computer-readable storage medium can refer to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0044] Those skilled in the art should be able to realize that the modules and method steps of each example described in combination with the embodiments disclosed herein can be implemented by electronic hardware, computer software, or a combination of the two. The programs corresponding to the software modules and method steps can be placed in a random access memory (RAM), internal memory, read-only memory (ROM), electrically programmable ROM, electrically erasable programmable ROM, registers, hard disk, removable disk, CD-ROM, or any other form of storage medium well-known in the technical field. To clearly illustrate the interchangeability of electronic hardware and software, the composition and steps of each example have been generally described according to functions in the above description. Whether these functions are executed in the form of electronic hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered to exceed the scope of the present invention.

[0045] Computer program code for performing the operations of the present application can be written in one or more programming languages or combinations thereof. The above-mentioned programming languages include object-oriented programming languages - such as Java, Smalltalk, C++, and also include conventional procedural programming languages - such as the "C" language or similar programming languages. The program code can be executed entirely on the user's computer, partially on the user's computer, executed as an independent software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In the case of a remote computer, the remote computer can be connected to the user's computer through any type of network - including a local area network (LAN) or a wide area network (WAN) - or can be connected to an external computer (for example, by using an Internet service provider to connect through the Internet).

[0046] The flowcharts and block diagrams in the accompanying drawings illustrate the possible architectures, functions, and operations of systems, methods, and computer program products according to various embodiments of the present application. In this regard, each block in the flowchart or block diagram may represent a module, a segment of a program, or a portion of code that contains one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions noted in the blocks may occur in a different order than that noted in the accompanying drawings. For example, two consecutive blocks shown may actually be executed substantially in parallel, or they may sometimes be executed in the reverse order, depending on the functions involved. It should also be noted that each block in the block diagram and / or flowchart, and combinations of blocks in the block diagram and / or flowchart, may be implemented by a dedicated hardware-based system that performs the specified functions or operations, or by a combination of dedicated hardware and computer instructions.

[0047] The terms "first", "second", etc. are used to distinguish similar objects and not to describe or indicate a particular order or sequence.

[0048] The term "comprising" or any other similar term is intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus / device that comprises a list of elements does not include only those elements but also other elements not expressly listed, or elements that are inherent to such process, method, article, or apparatus / device.

[0049] So far, the technical solution of the present invention has been described in connection with the preferred embodiments shown in the accompanying drawings. However, it is easily understood by those skilled in the art that the protection scope of the present invention is obviously not limited to these specific embodiments. Without departing from the principle of the present invention, those skilled in the art can make equivalent changes or substitutions to the relevant technical features, and the technical solutions after these changes or substitutions will fall within the protection scope of the present invention.

Claims

1. A magnetic particle imaging reconstruction method based on multi-angle Cartesian scan adaptive optimization, which is applied to a three-dimensional dual-excitation channel symmetric magnetic particle imaging device based on magnetic field free points, and is characterized in that The method includes the following steps: Step S100: Set scanning parameters, turn on and run the three-dimensional dual-excitation channel symmetric magnetic particle imaging device based on a magnetic field free point; Step S200: Take the low-frequency motion direction of the Cartesian scanning trajectory as the main direction of the magnetic field free point movement, take the angle between the main direction of the magnetic field free point movement and the perpendicular bisector of the x-direction receiving coil as the scanning trajectory angle, set the rotation angle and rotation order of the Cartesian scanning trajectory; Rotate the magnetic field free point according to the rotation angle and the rotation order; Step S300: Measure the signal of the unit sample at each pixel within the imaging field of view, and then construct the system function harmonic diagram at each rotation angle; Step S400: Calibrate the system matrix at different angles based on the system function harmonic diagram, perform frequency weighting selection at a single angle and multi-angle adaptive angle selection, determine the two-dimensional weight factor, and obtain the optimal system matrix at each rotation angle based on the two-dimensional weight factor; Step S500: Measure the response signals of the target object to be imaged at each rotation angle, and construct the observation vector sequence at each rotation angle based on the two-dimensional weight factor; Step S600: Combine the observation vector sequences and the optimal system matrices at each rotation angle by columns respectively. After combination, construct a linear equation system; Solve the linear equation system to obtain the reconstructed magnetic particle image corresponding to the target object to be imaged.

2. The magnetic particle imaging reconstruction method based on multi-angle Cartesian scan adaptive optimization according to claim 1, wherein Set the rotation angle of the Cartesian scanning trajectory, and the method is as follows: Let the basic trajectory of the Cartesian scanning trajectory be a grid movement along the x and y directions, and the specific expression is as follows: ； ； Among them, and respectively represent the driving magnetic fields in the x and y directions, A represents the magnetic field amplitude, and are respectively the change frequencies of the trajectory in the x and y directions; when a basic Cartesian trajectory excited along the x direction is obtained; Set rotation parameters : ； Perform trigonometric function processing on the two-axis excitation magnetic field of the X-axis and Z-axis with respect to the angle, and perform a rotation transformation based on the basic Cartesian trajectory to obtain a rotation Cartesian trajectory with the main direction of the magnetic field free point being : ； Among them, is the rotation angle, and the multi-angle rotation Cartesian trajectory is measured by setting a set of rotation angles for measurement.

3. The magnetic particle imaging reconstruction method based on multi-angle Cartesian scan adaptive optimization according to claim 2, wherein Set the rotation order of the Cartesian scanning trajectory, and the method is as follows: Take the initial direction of the basic Cartesian trajectory as 0°, and step and rotate in an equal-spacing or unequal-spacing manner to ensure that the movement range of the magnetic field free point can cover the entire imaging field of view, and rotate a total of N angles; where N represents the set number.

4. The magnetic particle imaging reconstruction method based on multi-angle Cartesian scan adaptive optimization according to claim 1, wherein Obtain the optimal system matrix at each rotation angle, and the method is as follows: Step S410: Obtain the system matrix at each rotation angle of the Cartesian scan trajectory , where θ is the rotation angle, ; Step S420: Perform a selection operation on the system matrices at each rotation angle according to the constructed evaluation function to obtain the weights at each angle; The selection operation includes elimination, low weight value or combination; The evaluation function includes a contribution degree evaluation function and a redundancy evaluation function; Step S430: Analyze the weights and contributions of the respective frequency components of the system matrix at a single angle to determine the frequency weighting matrix at this rotation angle; Step S440: Multiply the weights at each angle by the corresponding frequency weighting matrix to obtain the two-dimensional weight factor, and globally weight each system matrix based on the two-dimensional weight factor to generate the optimal system matrix at each rotation angle.

5. The magnetic particle imaging reconstruction method based on multi-angle Cartesian scan adaptive optimization according to claim 4, characterized in that Obtain the weights at each angle, and the method is as follows: Step S421: Evaluate the contribution of the system matrix at each rotation angle, and the contribution of each angle is as follows: ； Among them, is the system matrix with an angle of , where , is the finally reconstructed image; Step S422: Evaluate the redundancy of the system matrix at each rotation angle, and the redundancy of each angle is as follows: ； in, For two different rotation angles, 、 The angle is 、 The system matrix; Step S423: Screen the system matrices of each angle according to the contribution degree and redundancy, and set a dynamic threshold , , and screen out the angles whose evaluation is higher than the predetermined threshold, and assign weights; ； Then, the system matrices at each angle are as follows: ； Among them, , is the weighted matrix for multi-angle screening, that is, the angle screening weight. The vector in the matrix is the weight of each angle, , .

6. The magnetic particle imaging reconstruction method based on multi-angle Cartesian scan adaptive optimization according to claim 4, wherein Determine the frequency weighting matrix at this rotation angle, and the method is as follows: Step S431: Perform a fast Fourier transform on the received signal, extract the frequency components and their amplitude information, analyze the amplitudes of the signal at different frequencies, and determine the main multiple frequency components of the signal; Use a threshold value to screen for valid frequencies and determine the screening function : ； Among them, is a dynamic threshold, adjusted based on the noise level, h is the signal amplitude; f 0, f 1 represents the driving field frequency, k, k 1 represents the modulation coefficient of the modulation signal; Step S432: Extract the corresponding column from the system matrix according to the selected frequency index, and calculate the weight matrix : ； in, The angle is The signal amplitude of the signal received at the time is adjusted based on the signal amplitude or contribution, z Represents frequency components above the dynamic threshold; Step S433: Obtain an angle The system matrix after screening at the following: ； Among them, is the frequency weighting matrix for a single angle, and the vectors in the matrix are the weights of each frequency at this angle; is the system matrix with an angle of where ;​ Then the frequency screening weight of the system matrix is as follows: 。 7. The magnetic particle imaging reconstruction method based on multi-angle Cartesian scan adaptive optimization according to claim 5 or 6, characterized in that, Obtain the two-dimensional weight factor, and the method is as follows: Determine the two-dimensional weight factor based on the angle screening weight and the frequency screening weight , specifically as follows: ； Based on the two-dimensional weight factor perform global weighting on the system matrix to generate an optimal system matrix at each rotation angle : ； Among them, .

8. The magnetic particle imaging reconstruction method based on multi-angle Cartesian scan adaptive optimization according to claim 7, characterized in that Optimize the angle screening weight and the frequency screening weight, and then optimize the optimal system matrix as: ； Among them, is the optimal system matrix after multi-angle and frequency screening and weighting, y is the observed response signal, and x is the concentration distribution to be solved; is the regularization term, which is used to suppress noise or introduce prior information; is the regularization term of the angle screening weight, which is used to constrain the sparsity or smoothness of the screening angle; is the regularization term of the frequency screening weight, which is used to constrain the sparsity of the screening frequency; is the angle screening weight, is the frequency screening weight; 、 and are the regularization weights.

9. The magnetic particle imaging reconstruction method based on multi-angle Cartesian scan adaptive optimization according to claim 8, characterized in that Optimize the angle screening weight and the frequency screening weight, and the method is as follows: A. Set up evaluation metrics as the reconstruction quality evaluation criteria, and the evaluation metrics include PSNR, SSIM, temporal resolution, and robustness; B. Initial angle weight and frequency weight with the initial image ; C, Fixed weight and , and optimize by solving the image through the following equation: ； D. Fix the image x and optimize it by updating the weights through the following equation: ； ； E. Alternately and iteratively optimize x, , until the error converges.

10. The magnetic particle imaging reconstruction method based on multi-angle Cartesian scan adaptive optimization according to claim 7, wherein, Construct a system of linear equations: ； Among them, , is the response signal of the target object to be imaged at each rotation angle; C represents the discrete vector form of the target object to be imaged.

Images