A calculation method, device and storage medium for brain microwave imaging

Abstract

The present application discloses a calculation method, device and storage medium for brain microwave imaging, which relates to the field of data processing technology, wherein the method comprises: converting the frequency domain scattering field sequence matrix into the time domain waveform amplitude sequence matrix; bringing the time domain waveform amplitude sequence matrix into the recursive confocal imaging algorithm formula in the time domain to obtain the time domain imaging result; obtaining the initial dielectric constant filling image; bringing the time domain imaging result into the loss function as prior information for iteration, obtaining the dielectric constant distribution in the current iteration process, and then using the initial dielectric constant filling image as the initial state of iteration to obtain the final dielectric constant distribution; based on the final dielectric constant distribution and the known dielectric constant intervals corresponding to different lesion types, obtaining the lesion classification and recognition results. The time domain and frequency domain algorithms are applied together to microwave imaging, the time domain algorithm is used to obtain the prior information, and then the prior information is brought into the frequency domain iteration, thereby improving the imaging accuracy.

Description

Technical Field

[0001] The present application relates to the field of data processing technology, and in particular to a calculation method, device and storage medium for brain microwave imaging. Background Art

[0002] In the field of cranial microwave imaging, the general algorithm ideas are divided into two different types of imaging algorithms: time domain imaging algorithms and frequency domain imaging algorithms.

[0003] Among them, the time domain algorithm represented by confocal imaging is characterized by fast speed, but it is only sensitive to areas where the dielectric constant changes. It is more suitable for distinguishing parts where the dielectric constant mutates. For example, it is used to locate the patient's skull during imaging or the contours of abnormal areas such as bleeding. It cannot identify the type of lesion.

[0004] Frequency domain algorithms are mostly iterative algorithms. Iterative algorithms based on Born approximation have higher accuracy and can restore the distribution of dielectric constants in the brain to identify lesion types. However, the algorithm takes longer to iterate than time domain imaging, and is prone to overfitting when the number of antennas is small. Especially when the number of antennas is small, it is easy to have a huge deviation from the actual imaging target due to noise interference and other factors, and the anti-noise ability is slightly low.

[0005] In view of the problems such as noise interference and other factors existing in the existing technology, which lead to huge deviations from the actual imaging target and slightly low anti-noise ability, it is urgent to propose a new solution to solve the shortcomings of the above-mentioned existing technology. Summary of the invention

[0006] In view of the above situation, the embodiments of the present application propose a calculation method, device and storage medium for brain microwave imaging to overcome or at least partially overcome the shortcomings of the prior art.

[0007] The present application discloses a method for calculating brain microwave imaging, the method comprising:

[0008] S100, will use Frequency domain scattered field sequence matrix collected by the antenna array Convert to time domain waveform amplitude sequence matrix ;

[0009] S200, the time domain waveform amplitude sequence matrix Substitute the recursive confocal imaging algorithm formula in the time domain to obtain the time domain imaging results ;

[0010] S300, time domain imaging results Input the diffusion model, segment the time domain imaging results and fill in the initial dielectric constant to obtain the initial dielectric constant filled image ;

[0011] S400, time domain imaging results As a priori information, it is brought into the loss function for iteration to obtain the dielectric constant distribution in the current iteration process, and then the initial dielectric constant is filled in the image As the initial state of the iteration, obtain the final dielectric constant distribution ;

[0012] S500, based on final dielectric constant distribution The known dielectric constant intervals corresponding to different lesion types are used to obtain lesion classification and identification results.

[0013] In some embodiments, the frequency domain scattered field sequence matrix Through the frequency domain total field sequence matrix The background field sequence matrix without the brain Obtain the difference.

[0014] In some embodiments, the step S200 further includes:

[0015] S210, using the current time domain waveform amplitude sequence matrix , calculate the confocal value of each point in the imaging area , get the current recursive time domain result , the recursive confocal imaging algorithm formula is as follows:

[0016] ,in represents the coordinates of any point, represents the transmitting antenna, represents the receiving antenna, is the number of antennas, is the propagation speed of electromagnetic waves, For time, For antenna To point The distance For receiving antenna To point The distance Representative antenna Launch arrival point, then by Point transmission to antenna Compensation for electromagnetic wave amplitude attenuation caused by the transmission distance. is the number of recursions in the current time domain, It is the current time domain waveform amplitude sequence matrix.

[0017] In some embodiments, the step S200 further includes:

[0018] S220, according to the current recursive time domain result of step S210 , use edge detection to extract the current recursive time domain results If a new interface contour appears, the attenuation of the incident wave intensity is calculated using the Fresnel formula. , the intensity of the electromagnetic wave loss caused by interface reflection when the electromagnetic wave enters the brain is corrected, and the corrected time domain waveform amplitude sequence matrix , .

[0019] In some embodiments, S200 further includes: S230, when no new contour appears or the number of recursions is exceeded If it is greater than 5, the calculation ends and the time domain imaging result is obtained. Otherwise, the corrected time domain waveform amplitude sequence matrix Substitute the recursive confocal imaging algorithm formula and repeat the above steps S210-S230.

[0020] In some embodiments, the S300 further includes:

[0021] Diffusion model for time domain imaging results Identify and extract the contour of the time domain imaging result. The contours are the contours of the brain and the lesion area, and the dielectric constants of the brain tissues are filled in the contours of the brain and the lesion area to obtain the initial dielectric constant filling image. .

[0022] In some embodiments, S400 further includes:

[0023] S410, filling the image based on the acquired initial dielectric constant , generate the mask ;

[0024] S420, constructing a loss function Loss;

[0025] S430, the time domain imaging result As a priori information, it is brought into the loss function for iteration to obtain the dielectric constant distribution in the current iteration process. ;

[0026] S440, using mask The dielectric constant distribution in the current iteration process Mask and get a new dielectric constant distribution ;

[0027] S450, and then distribute the new dielectric constant The loss function Loss is introduced and iterated until the loss function value reaches the preset threshold, that is, the final dielectric constant distribution is obtained. .

[0028] In some embodiments, S500 further includes:

[0029] S510, based on the final dielectric constant distribution Divide the dielectric constant interval according to the lesion type and mark the corresponding lesion type;

[0030] S520, using different colors and different brightness to distinguish and display the marked lesion types and output them visually.

[0031] The present application also provides a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the calculation method for brain microwave imaging as described above.

[0032] The present application also provides a computing device for brain microwave imaging, comprising:

[0033] at least one processor; and,

[0034] a memory communicatively connected to the at least one processor; wherein,

[0035] The memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to perform the calculation method of brain microwave imaging as described above.

[0036] At least one of the above technical solutions adopted in the embodiments of the present application can achieve the following beneficial effects:

[0037] The present invention provides a brain image microwave detection algorithm that combines time and frequency domains, so as to obtain prior information using a time domain algorithm, and then bring the prior information into the frequency domain to iteratively improve imaging accuracy and optimize imaging results. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] The drawings described herein are used to provide a further understanding of the present application and constitute a part of the present application. The illustrative embodiments of the present application and their descriptions are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings:

[0039] Figure 1a A logical framework diagram showing a computational method for brain microwave imaging according to an embodiment of the present application.

[0040] Figure 1b A schematic flow chart showing a calculation method for brain microwave imaging according to an embodiment of the present application.

[0041] Figure 2 A schematic diagram of a diffusion model for converting time-domain imaging results according to an embodiment of the present application is shown.

[0042] Figure 3 A schematic diagram of a brain model according to an embodiment of the present application is shown.

[0043] Figure 4a A schematic diagram of a time domain imaging result according to an embodiment of the present application is shown.

[0044] Figure 4b A schematic diagram of a diffusion model conversion result according to an embodiment of the present application is shown.

[0045] Figure 4c A schematic diagram of frequency domain imaging results according to an embodiment of the present application is shown.

[0046] Figure 4d A schematic diagram of result visualization output according to an embodiment of the present application is shown. DETAILED DESCRIPTION

[0047] In order to make the purpose, technical solution and advantages of the present application clearer, the technical solution of the present application will be clearly and completely described below in combination with the specific embodiments of the present application and the corresponding drawings. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present application.

[0048] Based on this, this application can improve the final imaging quality to a certain extent by adding information such as the contrast of the target area tissue in advance. For example, prior information about the spatial distribution and prior information about the dielectric constant value. This application proposes a new calculation method for brain microwave imaging to address the shortcomings of the existing technology.

[0049] This application combines time domain and frequency domain algorithms for imaging, combines time domain and frequency domain algorithms for microwave imaging, brings the imaging results of the time domain algorithm into the frequency domain iteration as prior information, and uses the time domain imaging results to provide prior information for the frequency domain algorithm to optimize the imaging results. This has obvious significance in making up for the shortcomings of the two and achieving more accurate and rapid brain imaging.

[0050] This application mainly uses scattered field data to reconstruct the distribution of electromagnetic parameters (mainly dielectric constant) in the imaging area for imaging. The background field and total field outside the imaging area can be measured, that is, the scattering field generated by the brain can be obtained through measurement and data processing. Then, based on the calculated scattering field data, the position, shape and dielectric constant distribution of the scatterer in the imaging area are obtained through reverse iteration. In a specific embodiment, the imaging area is a target area selected in advance, such as the space contained in the human brain and the surrounding area of ​​the brain, and the scatterer is human brain tissue, lesions, tumors or blood clots, etc. The quantitative index of the so-called imaging quality in this application is defined by the relative error between the dielectric constant distribution of the constructed image and the actual dielectric constant distribution.

[0051] like Figure 1a As shown, Figure 1a The logical framework diagram of the calculation method of brain microwave imaging of one embodiment of the present application. The present application includes a time domain algorithm, a diffusion model, a frequency domain algorithm and recognition visualization. The time domain algorithm has signal compensation, recursive calculation and confocal imaging steps. The diffusion model has image segmentation and numerical filling steps. The frequency domain algorithm has the steps of obtaining the initial state of the dielectric constant distribution, obtaining time domain prior information, masking, iterative calculation, etc. The recognition visualization has the steps of dielectric constant comparison and result output.

[0052] It should be noted that the calculation method of brain microwave imaging of the present application can be applied to a microwave cerebral hemorrhage monitoring and analyzing instrument to realize cerebral hemorrhage detection.

[0053] The technical solutions provided by various embodiments of the present application are described in detail below in conjunction with the accompanying drawings.

[0054] Reference Figure 1b , Figure 1b A flowchart of a calculation method for brain microwave imaging is provided for this application, comprising the following specific steps:

[0055] S100, for use Frequency domain scattered field sequence matrix collected by the antenna array Convert to time domain waveform amplitude sequence matrix .

[0056] This application is for use Frequency domain scattered field sequence matrix collected by the antenna array Perform a frequency domain to time domain transformation and use discrete Fourier transform to convert it into a time domain waveform amplitude sequence matrix .

[0057] The frequency domain scattered field sequence matrix used in the algorithm of this application It can be obtained by: The antenna array is specifically arranged with multiple antennas centered on the brain. Each antenna is located in a space 0 to 10 cm away from the brain. Each antenna can be used for both RF transmission and RF reception. When the device is working, one antenna is in the transmitting state at any time, and the other antennas are in the receiving state to collect scattering parameters. The transmitting antenna is switched until all antenna transmission and reception combinations complete the collection to obtain the frequency domain total field sequence matrix. ,Will The background field sequence matrix without the brain Phase difference to obtain the scattered field sequence matrix .

[0058] S200, the time domain waveform amplitude sequence matrix Substitute the recursive confocal imaging algorithm formula in the time domain to obtain the time domain imaging results .

[0059] like Figure 3 Shown is a schematic diagram of a brain model, in which there is a circular hemorrhage area on the left side of the brain and a circular edema area on the right side.

[0060] The specific process of time domain imaging acquisition is as follows: in order to extract the prior information of the contour position of the multi-layer structure of the brain composed of skin, skull, etc., firstly, the collected radio frequency signal is calculated in the time domain using step S100.

[0061] Then proceed to step S200, which further includes:

[0062] S210, using the current time domain waveform amplitude sequence matrix , is the current time domain recursion times, is the time domain waveform amplitude sequence matrix in the current recursive process, an N*N matrix, In the following formula 1, .

[0063] According to the recursive confocal imaging algorithm formula 1

[0064] (Formula 1)

[0065] Calculate the confocal value for each point in the imaging area Get the current recursive time domain result .in represents the coordinates of any point, represents the transmitting antenna, represents the receiving antenna, is the number of antennas, is the propagation speed of electromagnetic waves, For time, For antenna To point The distance For receiving antenna To point The distance Representative antenna Launch arrival point, then by Point transmission to antenna Compensation for the attenuation of electromagnetic wave amplitude caused by the transmission distance.

[0066] S220, according to the current recursive time domain result of step S210 , use edge detection to extract the current recursive time domain results If a new interface contour appears, the attenuation of the incident wave intensity is calculated using the Fresnel formula. , the signal loss caused by interface reflection when the signal is transmitted into the brain is corrected, and the corrected time domain waveform amplitude sequence matrix , .

[0067] S230, when no new interface contour appears or the number of recursions is If it is greater than 5, the calculation ends and the time domain imaging result is obtained. , refer to Figure 4a , otherwise the corrected time domain waveform amplitude sequence matrix Substitute the recursive confocal imaging algorithm formula and repeat the above steps S210-S230. Repeat the above calculation until no new interface contour appears to obtain the time domain imaging result as shown in Figure 4a shown.

[0068] The following is an introduction to the principles used in this application:

[0069] (1) Calculate the attenuation of the incident wave intensity using the Fresnel formula Here’s how it works:

[0070] In human tissue, the speed of electromagnetic waves It can be expressed by formula 2 as:

[0071] (Formula 2)

[0072] in, is the medium conductivity, is the angular frequency of the electromagnetic wave, is the vacuum permeability, is the relative magnetic permeability, is the dielectric constant of vacuum, is the relative dielectric constant.

[0073] According to Maxwell's equations, we have , replacing the vector Laplacian ,get:

[0074] (Formula 3)

[0075] is the complex propagation constant , further deduced:

[0076] , (Formula 4)

[0077] in is the attenuation constant, is the phase constant, is the imaginary unit, is the real part of the dielectric constant, is the imaginary part of the dielectric constant.

[0078] In the case of weak loss , at this time the distance from the transmitting antenna is The electric field strength at can be expressed as:

[0079] (Formula 5)

[0080] in This is the time domain waveform of the original signal strength under ideal conditions.

[0081] It is the expected signal strength without reflection during signal transmission.

[0082] Since electromagnetic waves will be reflected and refracted when passing through the surface of tissues such as the skin during propagation, the Fresnel formula is used to calculate the intensities of the reflected wave, incident wave, and transmitted wave generated when the signal is transmitted to the brain. In this way, the energy lost when the electromagnetic wave signal is reflected at the interface between different tissues and media during transmission can be calculated to obtain the true intensity of the transmitted wave.

[0083] (2) Use the recursive confocal imaging algorithm formula multiple times to calculate the confocal value of each point in the imaging area Get the current recursive time domain imaging result .

[0084] (3) Extract the current recursive time domain imaging results using edge detection If a new interface contour appears, the Fresnel formula is used to calculate the attenuation of the incident wave intensity. , to compensate for the signal loss caused by interface reflection when the signal is transmitted into the brain Repeat the above calculation until no new interface contour appears to obtain the time domain imaging result. Figure 4a .

[0085] S300, time domain imaging results Input diffusion model, time domain imaging results Perform segmentation and fill initial dielectric constant to obtain initial dielectric constant filling image .

[0086] Specifically: based on the time domain imaging result obtained in step S230 , DiffusionModel for time domain imaging results Identify and extract the contour of the time domain imaging result. The contours are the contours of the brain and the lesion area, and the dielectric constants of the brain tissues are filled in the contours of the brain and the lesion area to obtain the initial dielectric constant filling image. Due to the time domain imaging results It is a line graph. The line graph is filled with colors through the diffusion model. Different colors represent different values ​​of dielectric constants. The initial dielectric constant fills the image. A map of color patches is generated to fill the dielectric constant within the contour.

[0087] This application uses the diffusion model to identify and extract the contours of the brain and lesion area, fill the corresponding dielectric constant of the brain tissue, and improve the resolution for output. After the diffusion model conversion, the initial dielectric constant filled image is obtained , which is the initial iteration state, such as Figure 4b .

[0088] The diffusion model in S300 further includes:

[0089] An input image-conditional generation module (G), where the residual in the residual dense block (RRDB) can capture detailed features without batch normalization.

[0090] The core of the model is a modified U-Net (encoder E and decoder D), which consists of residual blocks enhanced by attention layers of specific resolutions (16*16 and 8*8) and a bottleneck with two residual blocks and one attention layer.

[0091] like Figure 2As shown in the figure, it is a schematic diagram of the diffusion model, which generates results through continuous iteration. In each iteration, the input image condition generation module G is used to extract features from the input time domain imaging results, and the extracted results are added to the current iteration state and input into the network EoD composed of the encoder E and the decoder D to obtain the next state. After multiple iterations, the initial dielectric constant filling image for frequency domain iteration is finally obtained.

[0092] The encoder side of U-Net uses downsampling blocks to gradually reduce the spatial dimensions of the input image and extract high-level features of various scales. In contrast, the decoder side uses upsampling blocks to reconstruct the image to its original size and uses jump connections between the corresponding layers of the encoder and decoder. These jump connections are essential for preserving detailed features and ensuring accurate and high-fidelity image reconstruction. The implementation principle of the diffusion model used in this application can refer to the prior art.

[0093] S400, based on the time domain imaging results As a priori information, the loss function Loss is used for iteration to obtain the dielectric constant distribution in the current iteration process, and then the initial dielectric constant is filled into the image. As the initial state of the iteration, the final dielectric constant distribution is obtained.

[0094] This application uses the time domain imaging result in step S200 The contour features are used as prior information and introduced into the quadratic programming iterative algorithm in the frequency domain. Then, the initial dielectric constant converted from the diffusion model is used to fill the image. As the initial state of the iteration, iterative calculation is performed. Imaging results The contour feature is the following loss function Loss .

[0095] S400 further includes:

[0096] S410, filling the image based on the acquired initial dielectric constant , generate the mask . Mask Used to limit the imaging range in subsequent frequency domain iterations.

[0097] Fill the image with the output of the diffusion model, the initial dielectric constant , as the initial state of the algorithm, according to formula 6:

[0098] (Formula 6)

[0099] Generate Mask .

[0100] S420, construct a loss function Loss, the loss function Loss is as follows:

[0101] , (Formula 7)

[0102] in is the parameter to be optimized, representing the dielectric constant distribution in the current iteration process, is the total number of frequency points, is the frequency point, is the transmitting antenna, For the receiving antenna, is the Green function, is the total electric field, is the coordinate of any point, The position of the current iteration The spatial gradient of the dielectric constant, The time domain imaging result In Location The value at , , , For custom parameters, Used to adjust the overall brightness of the image. Used to adjust the overall sharpness of the image. Used to adjust the tissue contour position gradient, Used to limit the distance range within which the C parameter works.

[0103] S430, the time domain imaging result As a priori information, it is brought into the loss function Loss for iteration to obtain the dielectric constant distribution in the current iteration process .

[0104] S440 using mask The dielectric constant distribution in the current iteration process Mask and get a new dielectric constant distribution .

[0105] Specifically include: , get new .

[0106] S450, and then distribute the new dielectric constant The loss function Loss is introduced and iterated until the loss function value reaches the preset threshold, that is, the final dielectric constant distribution is obtained. , the results are as follows Figure 4c .

[0107] The above steps S410-S450 are the quadratic programming iterative algorithm in the frequency domain adopted by this application.

[0108] S500, based on final dielectric constant distribution The known dielectric constant intervals corresponding to different lesion types are used to obtain lesion classification and identification results.

[0109] S500 also includes, S510, distributing the dielectric constant according to the final dielectric constant The dielectric constant intervals are divided according to the lesion types, and the corresponding lesion types are marked; S520, the marked lesion types are distinguished and displayed using different colors and different brightness and visually output.

[0110] In specific application scenarios, the final dielectric constant distribution Divide and infer the dielectric constant ranges of normal tissue, hemorrhage, infarction, edema and tumor, and mark the corresponding area types; use different colors and brightness to distinguish and display the areas marked as different lesion types, identify the lesion types and output them visually. Visual output to obtain imaging results Figure 4d .

[0111] The present invention realizes microwave imaging by combining time domain and frequency domain algorithms, and uses the time domain imaging result to provide prior information for the frequency domain algorithm to optimize the imaging result.

[0112] The present application also provides a computer-readable storage medium, which stores one or more programs, wherein the one or more programs include instructions, which, when executed by an electronic device including multiple application programs, enable the electronic device to execute Figure 1b The calculation method of brain microwave imaging in the illustrated embodiment is specifically used to execute the aforementioned method.

[0113] An embodiment of the present application also provides a computing device for brain microwave imaging, comprising: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the computing method for brain microwave imaging as described above.

[0114] At the hardware level, the computing device includes a processor, and optionally an internal bus, a network interface, and a memory. The memory may include a memory, such as a high-speed random access memory (RAM), and may also include a non-volatile memory, such as at least one disk storage. Of course, the electronic device may also include hardware required for other services.

[0115] The processor, network interface and memory can be interconnected through an internal bus, which can be an ISA (Industry Standard Architecture) bus, a PCI (Peripheral Component Interconnect) bus or an EISA (Extended Industry Standard Architecture) bus, etc. The bus can be divided into an address bus, a data bus, a control bus, etc.

[0116] The memory is used to store the program. Specifically, the program may include a program code, and the program code includes a computer operation instruction. The memory may include a memory and a non-volatile memory, and provides instructions and data to the processor.

[0117] The processor reads the corresponding computer program from the non-volatile memory into the memory and then runs it, forming a computing device for the calculation method of brain microwave imaging at the logical level. The processor executes the program stored in the memory and is specifically used to execute the above method.

[0118] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems or computer program products. Therefore, the present application may take the form of a complete hardware embodiment, a complete software embodiment or an embodiment combining software and hardware. Moreover, the present application may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program codes.

[0119] The above is only an embodiment of the present application and is not intended to limit the present application. For those skilled in the art, the present application may have various changes and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application should be included in the scope of the claims of the present application.

Claims

1. A method for calculating brain microwave imaging, characterized in that: The method comprises: S100, will use Frequency domain scattered field sequence matrix collected by the antenna array Convert to time domain waveform amplitude sequence matrix ; S200, the time domain waveform amplitude sequence matrix Substitute the recursive confocal imaging algorithm formula in the time domain to obtain the time domain imaging results , the recursive confocal imaging algorithm formula is used for: Using the current time domain waveform amplitude sequence matrix , calculate the confocal value of each point in the imaging area , get the current recursive time domain result ; S300, time domain imaging results Input the diffusion model, segment the time domain imaging results and fill in the initial dielectric constant to obtain the initial dielectric constant filled image ; S400, time domain imaging results As a priori information, it is brought into the loss function for iteration to obtain the dielectric constant distribution in the current iteration process, and then the initial dielectric constant is filled in the image As the initial state of the iteration, obtain the final dielectric constant distribution ; S500, based on final dielectric constant distribution The known dielectric constant intervals corresponding to different lesion types are used to obtain lesion classification and identification results.

2. The method for calculating brain microwave imaging according to claim 1, characterized in that: The frequency domain scattered field sequence matrix Through the frequency domain total field sequence matrix The background field sequence matrix without the brain Obtain the difference.

3. The brain microwave imaging calculation method according to claim 1, characterized in that: The step S200 further comprises: The recursive confocal imaging algorithm formula is as follows: ,in represents the coordinates of any point, represents the transmitting antenna, represents the receiving antenna, is the number of antennas, is the propagation speed of electromagnetic waves, For time, For antenna To point The distance For receiving antenna To point The distance Representative antenna Launch to point, then by Point transmission to antenna Compensation for electromagnetic wave amplitude attenuation caused by the transmission distance. is the number of recursions in the current time domain, It is the current time domain waveform amplitude sequence matrix.

4. The brain microwave imaging calculation method according to claim 3, characterized in that: The step S200 further includes: Step S220, based on the current recursive time domain result of step S210 , use edge detection to extract the current recursive time domain results If a new interface contour appears, the attenuation of the incident wave intensity is calculated using the Fresnel formula. , the intensity of the electromagnetic wave loss caused by interface reflection when the electromagnetic wave enters the brain is corrected, and the corrected time domain waveform amplitude sequence matrix , .

5. The brain microwave imaging calculation method according to claim 4, characterized in that: The S200 further includes: S230, when no new contour appears or the recursive number is reached If it is greater than 5, the calculation ends and the time domain imaging result is obtained. Otherwise, the corrected time domain waveform amplitude sequence matrix Substitute the recursive confocal imaging algorithm formula and repeat the above steps S210-S230.

6. The brain microwave imaging calculation method according to claim 1, characterized in that: The S300 further includes: The diffusion model has an important influence on the time domain imaging results. Identify and extract time domain imaging results The time domain imaging results are The contours are the contours of the brain and the lesion area, and the dielectric constants corresponding to the brain tissues are filled in the contours of the brain and the lesion area to obtain the initial dielectric constant filling image .

7. The brain microwave imaging calculation method according to claim 6, characterized in that: S400 further includes: S410, filling the image based on the acquired initial dielectric constant , generate the mask ; S420, construct loss function Loss ; S430, the time domain imaging result As prior information into the loss function Loss Iterate and obtain the dielectric constant distribution during the current iteration ; S440, using mask The dielectric constant distribution in the current iteration process Mask and get a new dielectric constant distribution ; S450, and then distribute the new dielectric constant Substitute into the loss function Loss Iterate until the loss function value reaches the preset threshold, and the final dielectric constant distribution is obtained. .

8. The brain microwave imaging calculation method according to claim 7, characterized in that: S500 further includes: S510, based on the final dielectric constant distribution Divide the dielectric constant interval according to the lesion type and mark the corresponding lesion type; S520, using different colors and different brightness to distinguish and display the marked lesion types and output them visually.

9. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the computing method for brain microwave imaging as claimed in any one of claims 1 to 8 is implemented.

10. A computing device for brain microwave imaging, characterized in that: include: at least one processor; as well as, a memory communicatively connected to the at least one processor; wherein, The memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the computational method for brain microwave imaging as described in any one of claims 1-8.

Images