Method, device and equipment for improving magnetic particle imaging spatial resolution

Abstract

The invention relates to the technical field of magnetic particle imaging, and discloses a method, device and equipment for improving the spatial resolution of magnetic particle imaging, and the method comprises the steps: obtaining SM component data, converting the SM component data from a complex number domain to an RGB domain, and obtaining SM sample data; white noise is added to the SM sample data to obtain SM noise-added sample data, and the SM sample data and the corresponding SM noise-added sample data form matched data pairs; training a preset SM de-noising model by using the matched data pair to obtain an SM de-noising model; and de-noising the actual SM data by using the SM de-noising model, and reconstructing a magnetic particle image by using the de-noised actual SM data. According to the method, the SM component is denoised by training the SM denoising model, noise information of the reconstructed magnetic particle image is reduced, the spatial resolution of magnetic particle imaging is improved, a high-resolution imaging result is obtained, and doctors are better assisted in clinical image interpretation and diagnosis.

Description

Technical Field

[0001] The present invention relates to the technical field of magnetic particle imaging, and in particular to a method, a device and equipment for improving the spatial resolution of magnetic particle imaging. Background Art

[0002] Magnetic particle imaging (MPI) is a new type of molecular imaging technology. Its working principle is to use a magnetic field to stimulate magnetic nanoparticles (MNPs) to generate magnetization signals, which are received by a receiving coil and then the MNPs concentration distribution map is reconstructed. MPI has the advantages of no imaging depth limitation, high sensitivity, no background signal, and no ionizing radiation. It has shown great advantages in applications such as cell tracing, angiography and drug delivery.

[0003] High spatial resolution means that the image can provide clear and accurate details, textures and small-size visual feature information, thus better assisting doctors in clinical image interpretation. However, since the MPI system is interfered by noise and harmonics, the calibrated system matrices (SM) contain noise information, resulting in noise in the imaging space and reduced resolution. Summary of the invention

[0004] In view of this, the present invention provides a method, a device and an apparatus for improving the spatial resolution of magnetic particle imaging, so as to solve the problem of low imaging spatial resolution of existing magnetic particle imaging systems.

[0005] In a first aspect, the present invention provides a method for improving the spatial resolution of magnetic particle imaging, the method comprising: Obtain SM component data, and convert the SM component data from the complex domain to the RGB domain to obtain SM sample data; Add white noise to the SM sample data to obtain SM noisy sample data, and the SM sample data and the corresponding SM noisy sample data constitute a matching data pair; The preset SM denoising model is trained by using the matching data to obtain the SM denoising model; The actual SM data is denoised using the SM denoising model to obtain denoised actual SM data, and the denoised actual SM data is used to reconstruct a magnetic particle image.

[0006] The method for improving the spatial resolution of magnetic particle imaging provided by the present invention realizes denoising of SM components by training the SM denoising model, reduces the noise information of reconstructing the magnetic particle image, thereby improving the spatial resolution of MPI, obtaining high-resolution imaging results, and better assisting doctors in clinical image interpretation and diagnosis.

[0007] In an optional implementation manner, obtaining SM component data includes: Setting the lower threshold of the SM component according to the effectiveness of the SM component; The SM component calibrated by magnetic particle imaging is compared with the lower threshold, and the SM component not less than the lower threshold is used as SM component data.

[0008] The method for improving the spatial resolution of magnetic particle imaging provided by the present invention screens out effective SM components by setting the lower threshold of the SM component, avoids further operation on invalid data, saves computing resources, and improves the processing efficiency of SM component data.

[0009] In an optional implementation, the SM component data is converted from the complex domain to the RGB domain to obtain SM sample data, including: Using the HSV color model, convert the SM component data into an HSV vector; Use the preset HSV-RGB conversion standard to convert the HSV vector into an RGB vector; The SM sample data in the RGB domain is obtained by linear amplitude scaling.

[0010] The method for improving the spatial resolution of magnetic particle imaging provided by the present invention is conducive to accurately representing and processing color data by converting HSV vectors into RGB vectors, thereby improving the accuracy and efficiency of color processing and facilitating magnetic particle image reconstruction using SM component data.

[0011] In an optional implementation, the preset SM denoising model includes: a feature extraction unit, a multi-scale information mining unit, and a multi-scale information fusion unit, wherein: A feature extraction unit, used to extract feature information from noisy SM component data; A multi-scale information mining unit, used to extract useful feature information of each preset scale from feature information according to multiple preset scales; The multi-scale information fusion unit is used to fuse the useful feature information of each preset scale to obtain the useful information in the noisy SM component.

[0012] The method for improving the spatial resolution of magnetic particle imaging provided by the present invention obtains a deep learning model by training a preset SM denoising model, and uses the deep learning model to denoise the SM component to support the acquisition of high spatial resolution imaging results, thereby providing clearer and more accurate details, textures and small-size visual information for the imaging results.

[0013] In an optional implementation, a preset SM denoising model is trained using matching data pairs to obtain the SM denoising model, including: Divide the matching data pairs into training set, test set and validation set according to the preset ratio; The preset SM denoising model is trained using the training set and the validation set, and the trained model is evaluated using the test set to obtain the SM denoising model.

[0014] The method for improving the spatial resolution of magnetic particle imaging provided by the present invention obtains an SM denoising model through deep learning model training, denoises SM data containing noise, and then reconstructs the image using the denoised SM data, thereby improving the spatial resolution of MPI and reducing the hardware design and task complexity of improving the spatial resolution only by hardware improvement.

[0015] In an optional implementation, a preset SM denoising model is trained using a training set and a validation set, and the trained model is validated using a test set to obtain a SM denoising model, including: Randomly initialize the parameters of the preset SM denoising model; The SM noisy sample data in the training set and the validation set are used as model input, and the corresponding SM sample data are used as labels, and are input into the preset SM denoising model together, and the parameters of the preset SM denoising model are updated to obtain the SM denoising training model; The test set is used to evaluate whether the performance of the SM denoising training model meets the preset performance requirements, and the model that meets the preset performance requirements is used as the SM denoising model.

[0016] The method for improving the spatial resolution of magnetic particle imaging provided by the present invention, the SM denoising model is a deep learning model, which can automatically extract the global and local information of the SM component, and master the ability of noise removal from the training data pair, and directly use it for the already collected SM component to improve the spatial resolution of MPI.

[0017] In a second aspect, the present invention provides a device for improving the spatial resolution of magnetic particle imaging, the device comprising: A sample acquisition module is used to acquire SM component data and convert the SM component data from a complex domain to an RGB domain to obtain SM sample data; A data noise adding module is used to add white noise to the SM sample data to obtain SM noise adding sample data, and the SM sample data and the corresponding SM noise adding sample data constitute a matching data pair; A model training module is used to train a preset SM denoising model using matching data pairs to obtain a SM denoising model; The SM denoising module is used to denoise the actual SM data using the SM denoising model to obtain the denoised actual SM data, and reconstruct the magnetic particle image using the denoised actual SM data.

[0018] In a third aspect, the present invention provides a computer device, comprising: a memory and a processor, the memory and the processor being communicatively connected to each other, the memory storing computer instructions, and the processor executing the method of the first aspect or any corresponding embodiment thereof by executing the computer instructions.

[0019] In a fourth aspect, the present invention provides a computer-readable storage medium having computer instructions stored thereon, the computer instructions being used to cause a computer to execute the method of the first aspect or any corresponding embodiment thereof.

[0020] In a fifth aspect, the present invention provides a computer program product, comprising computer instructions for causing a computer to execute the method of the first aspect or any corresponding embodiment thereof. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] In order to more clearly illustrate the specific implementation methods of the present invention or the technical solutions in the prior art, the drawings required for use in the specific implementation methods or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some implementation methods of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0022] Figure 1 is a schematic flow chart of a method for improving the spatial resolution of magnetic particle imaging according to an embodiment of the present invention; Figure 2 is a flow chart of another method for improving the spatial resolution of magnetic particle imaging according to an embodiment of the present invention; Figure 3 is a structural schematic diagram of an SM denoising model in a method for improving the spatial resolution of magnetic particle imaging according to an embodiment of the present invention; Figure 4 is a schematic diagram of the process of training an SM denoising model in a method for improving the spatial resolution of magnetic particle imaging according to an embodiment of the present invention; Figure 5 is a schematic diagram of input and output of an SM denoising model using a test set as input in a method for improving the spatial resolution of magnetic particle imaging according to an embodiment of the present invention; Figure 6 is a schematic diagram for comparing image reconstruction results obtained by imaging using the denoised SM component in a method for improving the spatial resolution of magnetic particle imaging according to an embodiment of the present invention; Figure 7 is a structural block diagram of a device for improving the spatial resolution of magnetic particle imaging according to an embodiment of the present invention; Figure 8 It is a schematic diagram of the hardware structure of a computer device according to an embodiment of the present invention. DETAILED DESCRIPTION

[0023] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present invention.

[0024] The embodiment of the present invention provides a method for improving the spatial resolution of magnetic particle imaging, which reduces the noise information of the reconstructed image by removing the noise in the SM, so as to achieve the effect of improving the spatial resolution of the MPI.

[0025] According to an embodiment of the present invention, a method embodiment for improving the spatial resolution of magnetic particle imaging is provided. It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and although a logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0026] In this embodiment, a method for improving the spatial resolution of magnetic particle imaging is provided, which can be used in the above-mentioned computer system. Figure 1 is a flow chart of a method for improving the spatial resolution of magnetic particle imaging according to an embodiment of the present invention. Figure 1 As shown, the process includes the following steps: Step S101, obtaining SM component data, and converting the SM component data from the complex domain to the RGB domain to obtain SM sample data.

[0027] Specifically, in the image reconstruction process based on the system matrix, it is first necessary to discretize the imaging field of view, construct the system matrix by scanning point by point using a small-size sample, and use the linear mapping relationship between the system matrix and the object to be measured to reverse the distribution of magnetic particles in space. However, due to the encoding method of the magnetic field free line, the magnetic particles along the direction of the magnetic field free line cannot be decoded, so in practice, it is necessary to measure the magnetic field free line at multiple angles to achieve the position decoding of the magnetic particles. Accordingly, in the system matrix reconstruction method, it is necessary to measure the system matrix at multiple angles and use it after splicing. Therefore, the system matrix directly affects the imaging effect of the reconstructed image. For multiple system matrices, the component data of each system matrix is ​​obtained, and the SM component data is converted from the complex domain to the RGB domain, and the SM component in the RGB domain is obtained as the SM sample data, which is used to train the preset SM denoising model.

[0028] Step S102, adding white noise to the SM sample data to obtain SM noisy sample data, and the SM sample data and the corresponding SM noisy sample data constitute a matching data pair.

[0029] Specifically, the embodiment of the present invention uses the norm function of the scipy package to add white noise to the SM sample data to obtain SM noisy sample data, and the SM sample data and the corresponding SM noisy sample data after noisy addition constitute a matching data pair. The process of adding white noise to the sample data is a mature technology and will not be repeated here.

[0030] Step S103: training a preset SM denoising model using the matching data to obtain a SM denoising model.

[0031] Specifically, the embodiment of the present invention uses the train_test_split function in scipy to divide all SM components into a training set and a test set, where the training set and the test set account for and , and then randomly select from the training set As a validation set. Use the training set and test set to train the model, and use the validation set to evaluate the model, and finally get the optimal SM denoising model that meets the preset requirements.

[0032] Step S104, denoising the actual SM data using the SM denoising model to obtain denoised actual SM data, and reconstructing the magnetic particle image using the denoised actual SM data.

[0033] Specifically, in the actual magnetic particle imaging space, actual SM data is acquired, and denoised using the SM denoising model to obtain denoised actual SM data, and the denoised actual SM data is used to reconstruct the magnetic particle image. This process is a mature technical solution and will not be described in detail here. Since the SM data has been denoised in advance, the magnetic particle image reconstructed using the denoised SM data has a higher resolution.

[0034] The method for improving the spatial resolution of magnetic particle imaging provided in this embodiment realizes denoising of the SM component by training the SM denoising model, thereby reducing the noise information of the reconstructed magnetic particle image, thereby improving the spatial resolution of MPI, obtaining high-resolution imaging results, and better assisting doctors in clinical image interpretation and diagnosis.

[0035] In this embodiment, a method for improving the spatial resolution of magnetic particle imaging is provided, which can be used in the above-mentioned computer system. Figure 2 is a flow chart of a method for improving the spatial resolution of magnetic particle imaging according to an embodiment of the present invention. Figure 2 As shown, the process includes the following steps: Step S201, obtaining SM component data, and converting the SM component data from the complex domain to the RGB domain to obtain SM sample data.

[0036] Specifically, the above step S201 includes: Step S2011, setting a lower threshold of the SM component according to the validity of the SM component.

[0037] Specifically, a is set as the lower threshold of the valid SM component, that is, SM components greater than or equal to a can be used and retained; SM components less than a are unusable and screened out. The size of a is set according to actual conditions, and the specific setting standard is not limited here.

[0038] Step S2012, comparing the SM component calibrated by magnetic particle imaging with a lower threshold, and taking the SM component not less than the lower threshold as SM component data.

[0039] Specifically, the signal to interference plus noise ratio (SNR) is screened out from the calibrated SM. SM component It can be used as a valid SM component, and the remaining SM components smaller than a are unusable and are screened out.

[0040] The method for improving the spatial resolution of magnetic particle imaging provided in this embodiment sets the lower threshold of the SM component to screen out valid SM components, avoids further operations on invalid data, saves computing resources, and improves the processing efficiency of SM component data.

[0041] Step S2013, using the HSV color model, convert the SM component data into an HSV vector.

[0042] Specifically, the HSV (Hue, Saturation, Value) color model is used to represent the phase ,Right now It should be noted that each value of the SM component is in the form of a+bj complex number, which can be converted into a vector To represent, then each value of the SM component is converted to ( ) format, where the 0th bit represents the phase angle, the 1st bit is 1, and the 2nd bit is the absolute value of the variable modulus.

[0043] Step S2014: using a preset HSV-RGB conversion standard, convert the HSV vector into an RGB vector.

[0044] Specifically, the HSV vector is converted into an RGB vector using the preset HSV-RGB conversion standard. The specific conversion process is a mature existing technology and will not be described here.

[0045] Step S2015, obtain SM sample data in the RGB domain by using amplitude linear scaling conversion.

[0046] Specifically, the SM sample data is the SM component finally converted into the RGB domain, expressed as .

[0047] The method for improving the spatial resolution of magnetic particle imaging provided in this embodiment, by converting HSV vectors into RGB vectors, is conducive to accurately representing and processing color data, improving the accuracy and efficiency of color processing, and facilitating the reconstruction of magnetic particle images using SM component data.

[0048] Step S202, add white noise to the SM sample data to obtain SM noisy sample data, and the SM sample data and the corresponding SM noisy sample data form a matching data pair. Figure 1 Step S102 of the illustrated embodiment will not be described in detail here.

[0049] Step S203: training a preset SM denoising model using the matching data to obtain a SM denoising model.

[0050] In some optional implementations, a preset SM denoising model based on deep learning, such as Figure 3 As shown, including: The feature extraction unit is used to extract feature information from the noisy SM component data.

[0051] Specifically, using the CNN-based backbone, the noisy SM component input into the preset SM denoising model is subjected to convolution, pooling, and activation layers to obtain the feature map of the noisy SM. The channel attention mechanism is used to calculate the attention weights of each feature map in the high- and low-frequency feature channels respectively, and multiply them with the corresponding feature maps. The results are used as high- and low-frequency information, that is, feature information, respectively. The feature information contains useful information and noise information.

[0052] The multi-scale information mining unit is used to extract useful feature information of each preset scale from the feature information according to multiple preset scales.

[0053] Specifically, useful feature information of different scales, such as overall information and detail information, is extracted from the feature map of the noisy SM through receptive fields of different sizes: a large receptive field can extract the overall information in the noisy SM, while a small receptive field focuses on abstracting the detail information in the noisy SM. Each information extraction part contains a convolution layer, a pooling layer, and a LeakyReLU activation layer, where the convolution layer further abstracts the feature map, and the other two layers fuse the information of different receptive fields and perform nonlinear output. The output of each layer is the input of the next layer. Each information extraction part also contains a transition layer at the end, which is used to process the feature map into the same size.

[0054] The multi-scale information fusion unit is used to fuse the useful feature information of each preset scale to obtain the useful information in the noisy SM component.

[0055] Specifically, after obtaining multiple global and local information of SM, the useful feature information of each preset scale extracted under different receptive fields is globally and locally fused, and then the useful information is selected from the noisy SM component and the noise information is removed.

[0056] The method for improving the spatial resolution of magnetic particle imaging provided in this embodiment is to obtain a deep learning model by training a preset SM denoising model, and to use the deep learning model to denoise the SM component to support obtaining high spatial resolution imaging results, thereby providing clearer and more accurate details, textures, and small-size visual information for the imaging results.

[0057] Specifically, the above step S203 includes: Step S2031, dividing the matching data pairs into a training set, a test set, and a validation set according to a preset ratio.

[0058] Specifically, the train_test_split function in scipy is used to divide all matching data pairs into training sets and test sets, where the proportion of training sets and test sets is and , and then randomly select from the training set as a validation set.

[0059] Step S2032: train the preset SM denoising model using the training set and the validation set, and evaluate the trained model using the test set to obtain the SM denoising model.

[0060] Specifically, the SM noisy sample data with noise added in the training set is used as the network input, and the SM sample data without noise added is used as the label to supervise the training of the preset SM denoising model. The optimization algorithm used in the model supervised training is stochastic gradient descent (SGD), and the loss function uses MSE, which is only used as an example, but not limited to this.

[0061] The method for improving the spatial resolution of magnetic particle imaging provided in this embodiment obtains an SM denoising model through deep learning model training, denoises SM data containing noise, and then reconstructs the image using the denoised SM data, thereby improving the spatial resolution of MPI and reducing the hardware design and task complexity of improving the spatial resolution only through hardware improvement.

[0062] In some optional embodiments, such as Figure 4 As shown, the process of model training in step S2032 includes: Step a1, randomly initialize the parameters of the preset SM denoising model.

[0063] Specifically, the preset SM denoising model to be trained is loaded, and its parameters are randomly initialized.

[0064] Step a2: The SM noisy sample data in the training set and the validation set are used as model input, and the corresponding SM sample data are used as labels, and are input into the preset SM denoising model together, and the parameters of the preset SM denoising model are updated to obtain the SM denoising training model.

[0065] Specifically, before each training, the SM in the training set is preprocessed by random 90° rotation and random inversion, and the preprocessed data is input into the preset SM denoising model for forward propagation to obtain the processing results.

[0066] For the processing results, select mean square error as the loss function and back-propagate the training error, and select Adam optimizer to adjust the model parameters.

[0067] If the number of training times does not reach the preset number of training times, this step needs to be repeated until the number of training times reaches the preset number of training times and the training ends.

[0068] The validation set is used to validate the model trained within the training times, and the model that meets the evaluation index is used as the prediction model, that is, the SM denoising training model. The evaluation index can be selected as nRMSE and signal-noise ratio (SNR), which are only used as examples but not limited to this.

[0069] Step a3: Use the test set to evaluate whether the performance of the SM denoising training model meets the preset performance requirements, and use the model that meets the preset performance requirements as the SM denoising model.

[0070] Specifically, the preset performance requirement may be a preset image resolution of an image obtained by reconstructing an image using the SM denoising training model, or whether the reconstructed image resolution is improved. This is only an example, but not limited to this. The image resolution can be obtained by visualizing the half-value weight width curve. The SM denoised sample data with noise added in the test set is used as the input of the SM denoising model, and the SM sample data without noise added is used as the label to test the SM denoising model to obtain the denoised SM component corresponding to the input. The input and output results are shown in FIG. Figure 5 As shown, Figure 5 (a) is the SM noise sample data with noise added, Figure 5 (b) is the denoised SM component obtained by the SM denoising model after the SM noisy sample data. By comparison Figure 5 (a) and Figure 5 (b) It can be seen that Figure 5 (b) The noise is reduced.

[0071] The SM components before and after denoising are used for imaging, and the half-height weight width is visualized for analysis, as shown below: Figure 6 The comparison results shown are Figure 6 (a): The rightmost picture shows the image reconstruction result using the SM component before denoising, where the blue solid line represents the pixel value of each point on the 162nd row, corresponding to the blue broken line of the half-height weighted width visualization graph on the far left; the middle picture shows the image reconstruction result using the SM component after denoising, where the red solid line represents the pixel value of each point on the 162nd row, corresponding to the red broken line of the half-height weighted width visualization graph on the far left. Figure 6 (b): The rightmost picture shows the image reconstruction result of imaging using the SM component before denoising, where the blue solid line represents the pixel value of each point on the 248th row, corresponding to the blue broken line of the half-height weighted width visualization graph on the far left; the middle picture shows the image reconstruction result of imaging using the SM component after denoising, where the red solid line represents the pixel value of each point on the 248th row, corresponding to the red broken line of the half-height weighted width visualization graph on the far left.

[0072] according to Figure 6 From the half-height weight width visualization in (a), we can see that at the arrow point, the highest point of the red line is higher than that of the blue line. Figure 6 From the half-height weighted width visualization diagram in (b), we can see that at the arrow point, the lowest point of the red line is lower than that of the blue line. This means that the two branches of the imaging target are more finely divided, with higher pixel values ​​being higher and lower values ​​being lower, so the two branches can be better separated. If the line in the half-height weighted width visualization diagram becomes a straight line, it means that nothing can be separated.

[0073] The method for improving the spatial resolution of magnetic particle imaging provided in this embodiment, the SM denoising model is a deep learning model, which can automatically extract the global and local information of the SM component, and master the ability of noise removal from the training data pair, and directly use it for the already collected SM components to improve the spatial resolution of MPI.

[0074] Step S204: denoise the actual SM data using the SM denoising model to obtain denoised actual SM data, and reconstruct the magnetic particle image using the denoised actual SM data. Figure 1 Step S104 of the illustrated embodiment will not be described in detail here.

[0075] The method for improving the spatial resolution of magnetic particle imaging provided in this embodiment realizes denoising of the SM component by training the SM denoising model, thereby reducing the noise information of the reconstructed magnetic particle image, thereby improving the spatial resolution of MPI, obtaining high-resolution imaging results, and better assisting doctors in clinical image interpretation and diagnosis.

[0076] In a specific embodiment, for example, SM 6 in the OpenMPI data set is used for an experiment, and the method for improving the spatial resolution of magnetic particle imaging provided by the present invention is used to add noise to the SM with high SNR to construct a sample pair training model, and then the SM with low SNR is denoised. After the image reconstruction is compared using the SM with low SNR before and after denoising, it can be found that after denoising the SM with low SNR, the resolution of image reconstruction is improved. The SM signal-to-noise ratio is improved without modifying the system hardware settings, thereby improving the imaging resolution.

[0077] In this embodiment, a device for improving the spatial resolution of magnetic particle imaging is also provided, and the device is used to implement the above-mentioned embodiments and preferred embodiments, and the descriptions that have been made will not be repeated. As used below, the term "module" can implement a combination of software and / or hardware of a predetermined function. Although the devices described in the following embodiments are preferably implemented in software, the implementation of hardware, or a combination of software and hardware, is also possible and conceived.

[0078] This embodiment provides a device for improving the spatial resolution of magnetic particle imaging, such as Figure 7 As shown, including: The sample acquisition module 701 is used to acquire SM component data and convert the SM component data from the complex domain to the RGB domain to obtain SM sample data.

[0079] The data noise adding module 702 is used to add white noise to the SM sample data to obtain SM noise adding sample data, and the SM sample data and the corresponding SM noise adding sample data constitute a matching data pair.

[0080] The model training module 703 is used to train the preset SM denoising model using the matching data pairs to obtain the SM denoising model.

[0081] The SM denoising module 704 is used to denoise the actual SM data using the SM denoising model to obtain denoised actual SM data, and reconstruct the magnetic particle image using the denoised actual SM data.

[0082] In some optional implementations, the sample acquisition module 701 includes: The threshold setting unit is used to set the lower threshold of the SM component according to the validity of the SM component.

[0083] The effective data screening unit is used to compare the SM component calibrated by magnetic particle imaging with the lower threshold, and take the SM component not less than the lower threshold as the SM component data.

[0084] The HSV vector conversion unit is used to convert the SM component data into an HSV vector using the HSV color model.

[0085] The RGB vector conversion unit is used to convert the HSV vector into an RGB vector by using a preset HSV-RGB conversion standard.

[0086] The RGB domain data determination unit is used to obtain SM sample data in the RGB domain by using amplitude linear scaling conversion.

[0087] In some optional implementations, the model training module 703 includes: The sample data division unit is used to divide the matching data pairs into a training set, a test set, and a validation set according to a preset ratio.

[0088] The model training and evaluation unit is used to train the preset SM denoising model using the training set and the validation set, and to evaluate the trained model using the test set to obtain the SM denoising model.

[0089] In some optional embodiments, the model training evaluation unit includes: The parameter initialization subunit is used to randomly initialize the parameters of the preset SM denoising model.

[0090] The model training subunit is used to use the SM noisy sample data in the training set and the validation set as model input, the corresponding SM sample data as labels, and jointly input the preset SM denoising model, update the parameters of the preset SM denoising model, and obtain the SM denoising training model.

[0091] The model evaluation subunit is used to evaluate whether the performance of the SM denoising training model meets the preset performance requirements using the test set, and the model that meets the preset performance requirements is used as the SM denoising model.

[0092] The further functional description of each of the above modules and units is the same as that of the above corresponding embodiments and will not be repeated here.

[0093] In this embodiment, the device for improving the spatial resolution of magnetic particle imaging is presented in the form of a functional unit, where the unit refers to an ASIC (Application Specific Integrated Circuit) circuit, a processor and memory that executes one or more software or fixed programs, and / or other devices that can provide the above functions.

[0094] The embodiment of the present invention also provides a computer device having the above Figure 7 The device shown here is for improving the spatial resolution of magnetic particle imaging.

[0095] See also Figure 8 , Figure 8 is a schematic diagram of the structure of a computer device provided by an optional embodiment of the present invention, such as Figure 8 As shown, the computer device includes: one or more processors 10, a memory 20, and interfaces for connecting various components, including high-speed interfaces and low-speed interfaces. Various components are connected to each other using different buses for communication, and can be installed on a common mainboard or installed in other ways as needed. The processor can process the instructions executed in the computer device, including instructions stored in or on the memory to display the graphical information of the GUI on an external input / output device (such as, a display device coupled to the interface). In some optional embodiments, if necessary, multiple processors and / or multiple buses can be used together with multiple memories and multiple memories. Similarly, multiple computer devices can be connected, and each device provides some necessary operations (for example, as a server array, a group of blade servers, or a multi-processor system). Figure 8 A processor 10 is taken as an example.

[0096] The processor 10 may be a central processing unit, a network processor or a combination thereof. The processor 10 may further include a hardware chip. The hardware chip may be a dedicated integrated circuit, a programmable logic device or a combination thereof. The programmable logic device may be a complex programmable logic device, a field programmable gate array, a general purpose array logic or any combination thereof.

[0097] The memory 20 stores instructions executable by at least one processor 10, so that the at least one processor 10 executes the method shown in the above embodiment.

[0098] The memory 20 may include a program storage area and a data storage area, wherein the program storage area may store an operating system, an application required for at least one function; the data storage area may store data created according to the use of the computer device, etc. In addition, the memory 20 may include a high-speed random access memory, and may also include a non-transient memory, such as at least one disk storage device, a flash memory device, or other non-transient solid-state storage device. In some optional embodiments, the memory 20 may optionally include a memory remotely arranged relative to the processor 10, and these remote memories may be connected to the computer device via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0099] The memory 20 may include a volatile memory, such as a random access memory; the memory may also include a non-volatile memory, such as a flash memory, a hard disk or a solid state drive; the memory 20 may also include a combination of the above types of memory.

[0100] The computer device further comprises a communication interface 30 for the computer device to communicate with other devices or a communication network.

[0101] The embodiment of the present invention also provides a computer-readable storage medium. The method according to the embodiment of the present invention can be implemented in hardware, firmware, or can be implemented as a computer code that can be recorded in a storage medium, or can be implemented as a computer code that is originally stored in a remote storage medium or a non-temporary machine-readable storage medium and will be stored in a local storage medium through a network download, so that the method described herein can be stored in such software processing on a storage medium using a general-purpose computer, a dedicated processor, or programmable or dedicated hardware. Among them, the storage medium can be a magnetic disk, an optical disk, a read-only storage memory, a random access memory, a flash memory, a hard disk or a solid-state hard disk, etc.; further, the storage medium can also include a combination of the above types of memories. It can be understood that a computer, a processor, a microprocessor controller, or programmable hardware includes a storage component that can store or receive software or computer code. When the software or computer code is accessed and executed by a computer, a processor, or hardware, the method shown in the above embodiment is implemented.

[0102] A part of the present invention may be applied as a computer program product, such as a computer program instruction, which, when executed by a computer, can call or provide the method and / or technical solution according to the present invention through the operation of the computer. Those skilled in the art should understand that the existence of computer program instructions in computer-readable media includes, but is not limited to, source files, executable files, installation package files, etc., and accordingly, the way in which computer program instructions are executed by a computer includes, but is not limited to: the computer directly executes the instruction, or the computer compiles the instruction and then executes the corresponding compiled program, or the computer reads and executes the instruction, or the computer reads and installs the instruction and then executes the corresponding installed program. Here, the computer-readable medium may be any available computer-readable storage medium or communication medium accessible to the computer.

[0103] Although the embodiments of the present invention have been described in conjunction with the accompanying drawings, those skilled in the art may make various modifications and variations without departing from the spirit and scope of the present invention, and such modifications and variations are all within the scope defined by the appended claims.

Claims

1. A method for improving the spatial resolution of magnetic particle imaging, characterized in that: The method comprises: Acquire SM component data, and convert the SM component data from a complex domain to an RGB domain to obtain SM sample data; Adding white noise to the SM sample data to obtain SM noisy sample data, wherein the SM sample data and the corresponding SM noisy sample data constitute a matching data pair; Using the matching data to train a preset SM denoising model to obtain a SM denoising model; The actual SM data is denoised using the SM denoising model to obtain denoised actual SM data, and the denoised actual SM data is used to reconstruct a magnetic particle image.

2. The method according to claim 1, characterized in that Get SM component data, including: Setting the lower threshold of the SM component according to the effectiveness of the SM component; The SM component calibrated by magnetic particle imaging is compared with the lower threshold, and the SM component not less than the lower threshold is used as SM component data.

3. The method according to claim 1 or 2, characterized in that: The SM component data is converted from the complex domain to the RGB domain to obtain SM sample data, including: Using the HSV color model, converting the SM component data into an HSV vector; Using a preset HSV-RGB conversion standard, convert the HSV vector into an RGB vector; The SM sample data in the RGB domain is obtained by linear amplitude scaling.

4. The method according to claim 1, characterized in that: The preset SM denoising model includes: a feature extraction unit, a multi-scale information mining unit, and a multi-scale information fusion unit, wherein: A feature extraction unit, used to extract feature information from noisy SM component data; A multi-scale information mining unit, used to extract useful feature information of each preset scale from the feature information according to multiple preset scales; The multi-scale information fusion unit is used to fuse the useful feature information of each preset scale to obtain the useful information in the noisy SM component.

5. The method according to claim 1, characterized in that The method of training a preset SM denoising model using the matching data to obtain the SM denoising model includes: Dividing the matching data pairs into a training set, a test set, and a validation set according to a preset ratio; The preset SM denoising model is trained using the training set and the validation set, and the trained model is evaluated using the test set to obtain the SM denoising model.

6. The method according to claim 5, characterized in that The preset SM denoising model is trained using the training set and the validation set, and the trained model is validated using the test set to obtain the SM denoising model, including: Randomly initializing the parameters of the preset SM denoising model; The SM denoising sample data in the training set and the validation set are used as model inputs, and the corresponding SM sample data are used as labels, and are input into the preset SM denoising model together, and the parameters of the preset SM denoising model are updated to obtain the SM denoising training model; The test set is used to evaluate whether the performance of the SM denoising training model meets the preset performance requirements, and the model that meets the preset performance requirements is used as the SM denoising model.

7. A device for improving the spatial resolution of magnetic particle imaging, characterized in that: The device comprises: A sample acquisition module, used to acquire SM component data, and convert the SM component data from a complex domain to an RGB domain to obtain SM sample data; A data noise adding module, used for adding white noise to the SM sample data to obtain SM noise adding sample data, wherein the SM sample data and the corresponding SM noise adding sample data constitute a matching data pair; A model training module, used to train a preset SM denoising model using the matching data to obtain a SM denoising model; The SM denoising module is used to denoise the actual SM data using the SM denoising model to obtain the denoised actual SM data, and reconstruct the magnetic particle image using the denoised actual SM data.

8. A computer device, characterized in that: include: A memory and a processor, wherein the memory and the processor are communicatively connected to each other, the memory stores computer instructions, and the processor executes the method according to any one of claims 1 to 6 by executing the computer instructions.

9. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a computer to execute the method according to any one of claims 1 to 6.

10. A computer program product, characterized in that The method comprises computer instructions for causing a computer to execute the method according to any one of claims 1 to 6.

Images