Magnetic particle imaging signal enhancement method based on annular multi-channel receiving coil

Abstract

The invention belongs to the field of magnetic particle imaging, particularly relates to a magnetic particle imaging signal enhancement method based on an annular multi-channel receiving coil, and aims to solve the problem of how to deeply combine a multi-channel acquisition technology with a graph neural network algorithm. The method comprises the following steps: constructing a dynamic graph structure based on positions of multiple pairs of receiving coils of an annular receiving mechanism; signal coding is carried out on response signals generated by the multiple pairs of receiving coils of the annular receiving structure; encoding the spatial representation of the response signal into the node features of the dynamic graph structure, and performing information fusion based on the node features by using a graph neural network to obtain the output features of each pair of receiving coils in each convolutional layer of the graph neural network; and mapping the output feature of the last convolutional layer of the graph neural network back to the time domain space by using a one-dimensional linear layer and inverse Fourier transform to obtain a final enhanced signal. According to the invention, based on the receiving coil, multi-channel signal complementary fusion is realized by using a graph model, and finally, the MPI rapid scanning imaging capability is improved.

Description

Background Art

[0002] Magnetic particle imaging (MPI) is an emerging medical imaging technology that has been widely used in various biomedical fields. There is a mutually restrictive relationship between the scanning speed and signal quality of MPI. Although MPI can theoretically achieve a high scanning speed, in the actual imaging process, short-time scanning is often accompanied by more noise, resulting in a decrease in signal quality. Therefore, MPI usually needs to reduce noise by extending the scanning time, which makes it difficult to achieve the expected high-speed imaging effect, limiting the application of MPI in vascular imaging, functional imaging and other applications that require fast imaging. How to improve the quality of acquired signals while shortening the scanning time and achieve an optimal balance between scanning speed and signal quality is a key challenge in accelerating MPI imaging. Multi-channel parallel acquisition technology has significantly improved the fast scanning capabilities of various medical imaging modalities, and intelligent algorithms based on graph neural networks have shown unique advantages in multi-channel signal relationship modeling and complex data structure processing, especially in sparse data reconstruction, fast imaging and other tasks. These advances provide new solutions for further improving the fast imaging capabilities of MPI. During the scanning process of MPI, single-channel sparse and fast scanning signals often cannot fully characterize the spatial distribution information of magnetic particles. However, graph neural networks, with their powerful graph structure modeling capabilities, can effectively capture the complex relationship between multi-channel signals, and use the joint modeling of global and local signals to compensate for information loss and further improve signal quality and robustness. Therefore, how to deeply combine multi-channel acquisition technology with graph neural network algorithms is the core idea to improve the scanning speed and quality of MPI. Summary of the invention

[0003] In order to solve the above-mentioned problem in the prior art, that is, how to deeply combine the multi-channel acquisition technology with the graph neural network algorithm, the present invention provides a magnetic particle imaging signal enhancement method based on a toroidal multi-channel receiving coil, the method comprising: Based on the positions of multiple pairs of receiving coils of the toroidal receiving mechanism, a dynamic graph structure including multiple nodes is constructed; Signal encoding is performed on response signals generated by multiple pairs of receiving coils of the toroidal receiving structure to obtain a spatial representation of the response signals; The spatial representation of the response signal is encoded into the node features of the dynamic graph structure, and the graph neural network is used to fuse information based on the node features to obtain the output features of each pair of receiving coils in each convolutional layer of the graph neural network; The output features of the last convolutional layer of the graph neural network are mapped back to the time domain space using a one-dimensional linear layer and inverse Fourier transform to obtain the final output enhanced signal.

[0004] In a preferred embodiment, the circumferential receiving mechanism is: a plurality of pairs of receiving coils are stacked and placed in a direction surrounding the aperture to form a circumferential receiving structure with a plurality of different directions.

[0005] In a preferred embodiment, based on the positions of multiple pairs of receiving coils of the circumferential receiving mechanism, constructing a dynamic graph structure including multiple nodes includes: Each node of the dynamic graph structure represents a pair of receiving coils; Calculating the connection strength between any two pairs of receiving coils based on the offset angle between any two pairs of receiving coils; Calculate the cosine similarity of any two pairs of receiving coils; The final dynamic graph structure is determined based on the cosine similarity and connection strength of any two pairs of receiving coils.

[0006] In a preferred embodiment, calculating the connection strength between any two pairs of receiving coils includes: ; in, is the offset angle of the kth pair of coils, is the offset angle of the jth pair of coils, is the connection strength between the kth pair of coils and the jth pair of coils; k and j are the numbers of the coil pairs, For connection strength.

[0007] In a preferred embodiment, calculating the cosine similarity of any two pairs of receiving coils includes: ; in, is the cosine similarity between the kth pair of coils and the jth pair of coils, is the signal representation of the kth pair of coils, is the signal representation of the jth pair of coils, is the transpose of the signal representation of the j-th pair of coils, is the 2-norm of the signal representation of the k-th pair of coils, is the 2-norm of the signal representation of the j-th pair of coils, is the cosine similarity.

[0008] In a preferred embodiment, the final dynamic graph structure is: ; in, is the connection weight between the kth pair of coils and the jth pair of coils, To reconcile the weights.

[0009] In a preferred embodiment, encoding the response signals generated by the multiple pairs of receiving coils to obtain the spatial representation of the response signals includes: ; in, is the spatial representation of the response signal, is the convolution operator in the time domain, is the convolution operator in the frequency domain, is the activation function, represents Fourier transform.

[0010] In a preferred embodiment, the output features of each pair of receiving coils at each convolutional layer of the graph neural network obtained by using the graph neural network to perform information fusion based on node features include: The spatial representation of the response signal is used as the input feature of the first convolutional layer; For the kth pair of receiving coils, the output features of the oth convolutional layer are: ; in, is the output feature of the kth pair of receiving coils at the oth convolutional layer, W and are the parameters of the convolutional layer. is the connection weight between the kth pair of coils and the jth pair of coils, is the input feature of the kth pair of receiving coils in the oth convolutional layer, is the input feature of the j-th pair of receiving coils in the o-th convolutional layer.

[0011] In a preferred embodiment, the output features of the convolutional layer of the last layer of the graph neural network are mapped back to the time domain space using a one-dimensional linear layer and an inverse Fourier transform to obtain the final output enhanced signal, which includes: ; in, and are all trainable parameters. is the final output enhanced signal, is the time domain representation of the output feature of the last layer, is the frequency domain representation of the output features of the last layer, is the inverse Fourier transform.

[0012] In a second aspect of the present invention, a magnetic particle imaging signal enhancement system based on a toroidal multi-channel receiving coil is proposed, the system comprising: A graph structure building module, used for building a dynamic graph structure containing multiple nodes based on the positions of multiple pairs of receiving coils of the circumferential receiving mechanism; A signal encoding module, used for encoding the response signals generated by the multiple pairs of receiving coils of the toroidal receiving structure to obtain the spatial representation of the response signals; An information fusion module is used to encode the spatial representation of the response signal into the node features of the dynamic graph structure, and use the graph neural network to perform information fusion based on the node features to obtain the output features of each pair of receiving coils at each convolution layer of the graph neural network; The signal mapping module is used to map the output features of the convolutional layer of the last layer of the graph neural network back to the time domain space using a one-dimensional linear layer and inverse Fourier transform to obtain the final output enhanced signal.

[0013] Beneficial effects of the present invention: (1) The present invention acquires more signal information in the same time by constructing multi-channel parallel receiving coils, realizes joint analysis of multi-channel signals based on graph neural network, and improves the scanning speed of magnetic particle imaging on the basis of ensuring signal quality. At the same time, based on the construction of circumferential parallel receiving coils, the graph model is used to realize the complementary fusion of multi-channel signals, and the scanning time is reduced on the basis of ensuring high-quality signals, ultimately achieving the goal of improving the rapid scanning imaging capability of MPI; (2) The present invention proposes a surrounding stacked receiving coil structure, which reduces the number of turns of the opposing coils and stacks multiple pairs of receiving coils in the surrounding aperture direction; this receiving coil structure improves space utilization without changing the overall thickness and enriches the directional information of the magnetic particle response signal; (3) The present invention implements an algorithm to improve the quality of sparse fast signals and constructs a toroidal multi-channel receiving coil. The receiving coils are stacked in the direction surrounding the aperture to form a plurality of toroidal receiving structures in different directions. After acquiring the receiving signals in different directions at the same time, the graph neural network is used to perform joint complementary analysis to obtain high-quality signals. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] Other features, objects and advantages of the present application will become more apparent by reading the detailed description of non-limiting embodiments made with reference to the following drawings: Figure 1 is a schematic diagram of a magnetic particle imaging signal enhancement method based on a toroidal multi-channel receiving coil according to an embodiment of the present invention; Figure 2 is a schematic diagram of the design concept of multiple receiving coils of the circumferential receiving mechanism of an embodiment of the present invention; Figure 3 is a schematic diagram of a magnetic particle imaging signal enhancement method based on a toroidal multi-channel receiving coil using a graph neural network according to an embodiment of the present invention; Figure 4It is a structural diagram of a computer system of a server for implementing the method, system, and device embodiments of the present application. DETAILED DESCRIPTION

[0015] The present application will be further described in detail below in conjunction with the accompanying drawings and embodiments. It is to be understood that the specific embodiments described herein are only used to explain the relevant invention, rather than to limit the invention. It is also necessary to explain that, for ease of description, only the parts related to the relevant invention are shown in the accompanying drawings.

[0016] It should be noted that, in the absence of conflict, the embodiments and features in the embodiments of the present application can be combined with each other. The present application will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.

[0017] The present invention provides a method for enhancing magnetic particle imaging signals based on a toroidal multi-channel receiving coil, which specifically includes: Based on the positions of multiple pairs of receiving coils of the toroidal receiving mechanism, a dynamic graph structure including multiple nodes is constructed; Signal encoding is performed on response signals generated by multiple pairs of receiving coils of the toroidal receiving structure to obtain a spatial representation of the response signals; The spatial representation of the response signal is encoded into the node features of the dynamic graph structure, and the graph neural network is used to fuse information based on the node features to obtain the output features of each pair of receiving coils in each convolutional layer of the graph neural network; The output features of the last convolutional layer of the graph neural network are mapped back to the time domain space using a one-dimensional linear layer and inverse Fourier transform to obtain the final output enhanced signal.

[0018] In order to more clearly explain the magnetic particle imaging signal enhancement method based on the toroidal multi-channel receiving coil of the present invention, the following is combined with Figure 1 Each step in the embodiment of the present invention is described in detail.

[0019] The steps of the magnetic particle imaging signal enhancement method based on the toroidal multi-channel receiving coil according to the first embodiment of the present invention are described in detail as follows: Based on the positions of multiple pairs of receiving coils of the toroidal receiving mechanism, a dynamic graph structure including multiple nodes is constructed; In the present invention, the receiving coils appear in pairs at the MPI. Figure 2, which is a schematic diagram of the design concept of multiple receiving coils of the circumferential receiving mechanism of an embodiment of the present invention; in this embodiment, the circumferential receiving mechanism is: multiple pairs of receiving coils are stacked and placed in the direction of the circumferential aperture to form a circumferential receiving structure with multiple different directions. The receiving coil reduces the number of turns of the opposing coils, and multiple pairs of receiving coils are stacked and placed in the direction of the circumferential aperture. This receiving coil structure improves the space utilization without changing the overall thickness, and enriches the direction information of the magnetic particle response signal. The number of receiving coils can be initially set to 8 to 12 pairs according to actual conditions.

[0020] In this embodiment, based on the positions of multiple pairs of receiving coils of the circumferential receiving mechanism, constructing a dynamic graph structure including multiple nodes includes: Each node of the dynamic graph structure represents a pair of receiving coils; In this embodiment, the purpose of constructing a dynamic graph structure is to construct the correlation between signals of different channels, and to perform complementary enhancement of the signals between channels based on this correlation. The idea of ​​constructing the graph structure is as follows: The offset angle of the coil is , build a file containing N The graph structure of nodes, each node represents a pair of receiving coils, and any two pairs of receiving coils and The strength of the connection between them is defined by the angle difference between them.

[0021] The connection strength between any two pairs of receiving coils is calculated based on the offset angle between any two pairs of receiving coils. In this embodiment, the connection strength between any two pairs of receiving coils is calculated including: ; in, is the offset angle of the kth pair of coils, is the offset angle of the jth pair of coils, is the connection strength between the kth pair of coils and the jth pair of coils; k and j are the numbers of the coil pairs, For connection strength.

[0022] The graph structure constructed by the offset angle is fixed and cannot be dynamically adjusted according to the feature transformation during the model training process. Therefore, a dynamic graph structure is constructed based on the cosine similarity of the spatial representation of the response signal of each pair of coils, and the cosine similarity of any two pairs of receiving coils is calculated; In this embodiment, calculating the cosine similarity of any two pairs of receiving coils includes: ; in, is the cosine similarity between the kth pair of coils and the jth pair of coils, is the signal representation of the kth pair of coils, is the signal representation of the jth pair of coils, is the transpose of the signal representation of the j-th pair of coils, is the 2-norm of the signal representation of the k-th pair of coils, is the 2-norm of the signal representation of the j-th pair of coils, is the cosine similarity.

[0023] The final dynamic graph structure is determined based on the cosine similarity and connection strength of any two pairs of receiving coils.

[0024] In this embodiment, the final graph structure is determined by a linear combination of the connection strength of the two and the cosine similarity of the two: ; in, is the connection weight between the kth pair of coils and the jth pair of coils, To reconcile the weight, it can be initially set to 0.5. That is, the connection weight between the kth node and the jth node.

[0025] Signal encoding is performed on response signals generated by multiple pairs of receiving coils of the toroidal receiving structure to obtain a spatial representation of the response signals; In this embodiment, the response signal generated by each pair of coils is recorded as, which is essentially a one-dimensional time series signal. The present invention performs dimensional compression in the frequency domain and the time domain respectively through a one-dimensional convolution operation, and then splices them into a vector representation of the signal. Then, the spatial representation of the response signal generated by multiple pairs of receiving coils is obtained by signal encoding, including: ; in, is the spatial representation of the response signal, is the convolution operator in the time domain, is the convolution operator in the frequency domain, is the activation function, represents Fourier transform.

[0026] The spatial representation of the response signal is encoded into the node features of the dynamic graph structure, and the graph neural network is used to fuse information based on the node features to obtain the output features of each pair of receiving coils in each convolutional layer of the graph neural network; Graph neural network is a new model that performs convolution operations on non-Euclidean data and can simultaneously utilize the structural information and node feature information of the graph. Figure 3As shown, it is a schematic diagram of a magnetic particle imaging signal enhancement method based on a circumferential multi-channel receiving coil using a graph neural network according to an embodiment of the present invention. The present invention encodes hardware information into a graph structure, encodes response signal information into node features, and realizes hardware and software information fusion through a graph neural network. In this embodiment, the output features of each pair of receiving coils at each convolutional layer of the graph neural network obtained by information fusion based on node features using a graph neural network include: The spatial representation of the response signal is used as the input feature of the first convolutional layer; For the kth pair of receiving coils, the output features of the oth convolutional layer are: ; in, is the output feature of the kth pair of receiving coils at the oth convolutional layer, and are the parameters of the convolutional layer. is the connection weight between the kth pair of coils and the jth pair of coils, is the input feature of the kth pair of receiving coils in the oth convolutional layer, is the input feature of the j-th pair of receiving coils in the o-th convolutional layer.

[0027] Specifically, the spatial representation of each pair of coils is affected by the signals of the remaining coils, and the degree of influence is dynamically adjusted by the connection strength between the two. This allows the signals of each pair of coils to be jointly complemented with other signals during training, reducing the impact of undersampling on data quality.

[0028] The output features of the last convolutional layer of the graph neural network are mapped back to the time domain space using a one-dimensional linear layer and inverse Fourier transform to obtain the final output enhanced signal.

[0029] In this embodiment, the dimension of the received signal remains unchanged after multiple layers of convolution, and the output feature of the last layer is It can be expressed as: ; In this embodiment, the output features of the convolutional layer of the last layer of the graph neural network are mapped back to the time domain space using a one-dimensional linear layer and an inverse Fourier transform to obtain the final output enhanced signal including: ; in, and are all trainable parameters. is the final output enhanced signal, is the time domain representation of the output feature of the last layer, is the frequency domain representation of the output features of the last layer, is the inverse Fourier transform.

[0030] Although the various steps in the above embodiment are described in the above-mentioned order, those skilled in the art can understand that in order to achieve the effect 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 reverse order. These simple changes are within the scope of protection of the present invention.

[0031] A magnetic particle imaging signal enhancement system based on a toroidal multi-channel receiving coil according to a second embodiment of the present invention comprises: A graph structure building module, used for building a dynamic graph structure containing multiple nodes based on the positions of multiple pairs of receiving coils of the circumferential receiving mechanism; A signal encoding module, used for encoding the response signals generated by the multiple pairs of receiving coils of the toroidal receiving structure to obtain the spatial representation of the response signals; An information fusion module is used to encode the spatial representation of the response signal into the node features of the dynamic graph structure, and use the graph neural network to perform information fusion based on the node features to obtain the output features of each pair of receiving coils at each convolution layer of the graph neural network; The signal mapping module is used to map the output features of the convolutional layer of the last layer of the graph neural network back to the time domain space using a one-dimensional linear layer and inverse Fourier transform to obtain the final output enhanced signal.

[0032] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working process and related instructions of the system described above can refer to the corresponding process in the aforementioned method embodiment, and will not be repeated here.

[0033] It should be noted that the magnetic particle imaging signal enhancement system based on the toroidal multi-channel receiving coil provided in the above embodiment is only illustrated by the division of the above functional modules. In practical applications, the above functions can be assigned to different functional modules as needed, that is, the modules or steps in the embodiments of the present invention can be decomposed or combined. For example, the modules in the above embodiment can be combined into one module, or further divided into multiple sub-modules to complete all or part of the functions described above. The names of the modules and steps involved in the embodiments of the present invention are only for distinguishing the modules or steps, and are not regarded as improper limitations of the present invention.

[0034] An electronic device according to the third embodiment of the present invention comprises: 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 signal enhancement method based on the toroidal multi-channel receiving coil.

[0035] A computer-readable storage medium according to a fourth embodiment of the present invention stores computer instructions, and the computer instructions are used to be executed by the computer to implement the above-mentioned magnetic particle imaging signal enhancement method based on the toroidal multi-channel receiving coil.

[0036] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working process and related instructions of the storage device and processing device described above can refer to the corresponding process in the aforementioned method embodiment and will not be repeated here.

[0037] Those skilled in the art should be able to appreciate that the modules and method steps of each example described in conjunction with the embodiments disclosed herein can be implemented with electronic hardware, computer software, or a combination of the two, and the programs corresponding to the software modules and method steps can be placed in random access memory (RAM), memory, read-only memory (ROM), electrically programmable ROM, electrically erasable programmable ROM, registers, hard disks, removable disks, CD-ROMs, or any other form of storage medium known in the technical field. In order to clearly illustrate the interchangeability of electronic hardware and software, the composition and steps of each example have been generally described in the above description according to the function. Whether these functions are performed in 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.

[0038] Reference below Figure 4 , which shows a schematic diagram of the structure of a computer system of a server for implementing the method, system, and device embodiments of the present application. Figure 4 The server shown is merely an example and should not bring any limitation to the functions and scope of use of the embodiments of the present application.

[0039] like Figure 4 As shown, the computer system includes a central processing unit (CPU) 601, which can perform various appropriate actions and processes according to the program stored in the read-only memory (ROM) 602 or the program loaded from the storage part 608 to the random access memory (RAM) 603. In the RAM 603, various programs and data required for system operation are also stored. The CPU 601, ROM 602 and RAM 603 are connected to each other through a bus 604. An input / output (I / O) interface 605 is also connected to the bus 604.

[0040] The following components are connected to the I / O interface 605: an input section 606 including a keyboard, a mouse, etc.; an output section 607 including a cathode ray tube (CRT), a liquid crystal display (LCD), etc., and a speaker; a storage section 608 including a hard disk, etc.; and a communication section 609 including a network interface card such as a LAN (Local Area Network) card, a modem, etc. The communication section 609 performs communication processing via a network such as the Internet. A drive 610 is also connected to the I / O interface 605 as needed. A removable medium 611, such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, etc., is installed on the drive 610 as needed so that a computer program read therefrom is installed into the storage section 608 as needed.

[0041] In particular, according to an embodiment of the present disclosure, the process described above with reference to the flowchart can be implemented as a computer software program. For example, an embodiment of the present disclosure includes a computer program product, which includes a computer program carried on a computer-readable medium, and the computer program includes a program code for executing the method shown in the flowchart. In such an embodiment, the computer program can be downloaded and installed from the network through the communication part 609, and / or installed from the removable medium 611. When the computer program is executed by the central processing unit (CPU) 601, the above-mentioned functions defined in the method of the present application are executed. It should be noted that the above-mentioned computer-readable medium of the present application can be a computer-readable signal medium or a computer-readable storage medium or any combination of the above two. The computer-readable storage medium can be, for example, - but not limited to - an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or device, or any combination of the above. More specific examples of computer-readable storage media may include, but are not limited to, an electrical connection with one or more conductors, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In the present application, a computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in combination with an instruction execution system, an apparatus or a device. In the present application, a computer-readable signal medium may include a data signal propagated in a baseband or as part of a carrier wave, which carries a computer-readable program code. Such propagated data signals may take a variety of forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above. A computer-readable signal medium may also be any computer-readable medium other than a computer-readable storage medium, which may send, propagate, or transmit a program for use by or in combination with an instruction execution system, an apparatus or a device. The program code contained on the computer-readable medium may be transmitted using any appropriate medium, including but not limited to: wireless, wire, optical cable, RF, etc., or any suitable combination of the above.

[0042] Computer program code for performing the operations of the present application may be written in one or more programming languages ​​or a combination thereof, including object-oriented programming languages, such as Java, Smalltalk, C++, and conventional procedural programming languages, such as "C" or similar programming languages. The program code may be executed entirely on the user's computer, partially on the user's computer, as a separate software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving a remote computer, the remote computer may be connected to the user's computer via any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0043] The flow chart and block diagram in the accompanying drawings illustrate the possible architecture, function and operation of the system, method and computer program product according to various embodiments of the present application. In this regard, each square box in the flow chart or block diagram can represent a module, a program segment or a part of a code, and the module, the program segment or a part of the code contains one or more executable instructions for realizing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the square box can also occur in a sequence different from that marked in the accompanying drawings. For example, two square boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each square box in the block diagram and / or flow chart, and the combination of the square boxes in the block diagram and / or flow chart can be implemented with a dedicated hardware-based system that performs a specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions.

[0044] The terms "first", "second", etc. are used to distinguish similar objects rather than to describe or indicate a particular order or sequence.

[0045] The term "comprise" or any other similar term is intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus / device that includes a list of elements includes not only those elements but also other elements not expressly listed, or also includes elements inherent to such process, method, article, or apparatus / device.

[0046] So far, the technical solutions of the present invention have been described in conjunction with the preferred embodiments shown in the accompanying drawings. However, it is easy for those skilled in the art to understand 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 method for enhancing magnetic particle imaging signals based on a toroidal multi-channel receiving coil, characterized in that: The method comprises: Based on the positions of multiple pairs of receiving coils of the toroidal receiving mechanism, a dynamic graph structure including multiple nodes is constructed; Signal encoding is performed on response signals generated by multiple pairs of receiving coils of the toroidal receiving structure to obtain a spatial representation of the response signals; The spatial representation of the response signal is encoded into the node features of the dynamic graph structure, and the graph neural network is used to fuse information based on the node features to obtain the output features of each pair of receiving coils in each convolutional layer of the graph neural network; The output features of the last convolutional layer of the graph neural network are mapped back to the time domain space using a one-dimensional linear layer and inverse Fourier transform to obtain the final output enhanced signal.

2. The method for enhancing magnetic particle imaging signals based on a toroidal multi-channel receiving coil according to claim 1, characterized in that: The circumferential receiving mechanism is: a plurality of pairs of receiving coils are stacked and placed in a direction surrounding the aperture to form a circumferential receiving structure with a plurality of different directions.

3. The method for enhancing magnetic particle imaging signals based on a toroidal multi-channel receiving coil according to claim 2, characterized in that: Based on the positions of multiple pairs of receiving coils of the toroidal receiving mechanism, a dynamic graph structure containing multiple nodes is constructed including: Each node of the dynamic graph structure represents a pair of receiving coils; Calculating the connection strength between any two pairs of receiving coils based on the offset angle between any two pairs of receiving coils; Calculate the cosine similarity of any two pairs of receiving coils; The final dynamic graph structure is determined based on the cosine similarity and connection strength of any two pairs of receiving coils.

4. The method for enhancing magnetic particle imaging signals based on a toroidal multi-channel receiving coil according to claim 3, characterized in that: Calculating the connection strength between any two pairs of receiving coils involves: ； in, is the offset angle of the kth pair of coils, is the offset angle of the jth pair of coils, is the connection strength between the kth pair of coils and the jth pair of coils; k and j are the numbers of the coil pairs, For connection strength.

5. The method for enhancing magnetic particle imaging signals based on a toroidal multi-channel receiving coil according to claim 4, characterized in that: Calculating the cosine similarity of any two pairs of receiving coils involves: ； in, is the cosine similarity between the kth pair of coils and the jth pair of coils, is the signal representation of the kth pair of coils, is the signal representation of the jth pair of coils, is the transpose of the signal representation of the j-th pair of coils, is the 2-norm of the signal representation of the k-th pair of coils, is the 2-norm of the signal representation of the j-th pair of coils, is the cosine similarity.

6. The method for enhancing magnetic particle imaging signals based on a toroidal multi-channel receiving coil according to claim 5, characterized in that: The final dynamic graph structure is: ； in, is the connection weight between the kth pair of coils and the jth pair of coils, To reconcile the weights.

7. The method for enhancing magnetic particle imaging signals based on a toroidal multi-channel receiving coil according to claim 6, characterized in that: Signal encoding of response signals generated by multiple pairs of receiving coils to obtain spatial representation of response signals includes: ； in, is the spatial representation of the response signal, is the convolution operator in the time domain, is the convolution operator in the frequency domain, is the activation function, represents Fourier transform.

8. The method for enhancing magnetic particle imaging signals based on a toroidal multi-channel receiving coil according to claim 7, characterized in that: The output features of each pair of receiving coils in each convolutional layer of the graph neural network are obtained by using the graph neural network to fuse information based on node features, including: The spatial representation of the response signal is used as the input feature of the first convolutional layer; For the kth pair of receiving coils, the output features of the oth convolutional layer are: ； in, is the output feature of the kth pair of receiving coils at the oth convolutional layer, W and are the parameters of the convolutional layer. is the connection weight between the kth pair of coils and the jth pair of coils, is the input feature of the kth pair of receiving coils in the oth convolutional layer, is the input feature of the j-th pair of receiving coils in the o-th convolutional layer.

9. The method for enhancing magnetic particle imaging signals based on a toroidal multi-channel receiving coil according to claim 8, characterized in that: The output features of the convolutional layer of the last layer of the graph neural network are mapped back to the time domain space using a one-dimensional linear layer and inverse Fourier transform to obtain the final output enhanced signal including: ； in, and are all trainable parameters. is the final output enhanced signal, is the time domain representation of the output feature of the last layer, is the frequency domain representation of the output features of the last layer, is the inverse Fourier transform.

10. A magnetic particle imaging signal enhancement system based on a toroidal multi-channel receiving coil, characterized in that: The system comprises: A graph structure building module, used for building a dynamic graph structure containing multiple nodes based on the positions of multiple pairs of receiving coils of the circumferential receiving mechanism; A signal encoding module, used for encoding the response signals generated by the multiple pairs of receiving coils of the toroidal receiving structure to obtain the spatial representation of the response signals; An information fusion module is used to encode the spatial representation of the response signal into the node features of the dynamic graph structure, and use the graph neural network to perform information fusion based on the node features to obtain the output features of each pair of receiving coils at each convolution layer of the graph neural network; The signal mapping module is used to map the output features of the convolutional layer of the last layer of the graph neural network back to the time domain space using a one-dimensional linear layer and inverse Fourier transform to obtain the final output enhanced signal.

Images