Melanoma-based boron concentration distribution model acquisition method, terminal and medium

Abstract

The application provides a melanoma-based boron concentration distribution model acquisition method, a terminal and a medium. The target area of melanoma, the microneedle administration area and the administration concentration are obtained, PET image data of the target area at each sampling time after microneedle administration is collected, microneedle boron concentration data of the administration area at each sampling time is collected, a sample data set is constructed based on the PET image data and the corresponding microneedle boron concentration data, iterative training of the boron concentration distribution model is performed based on the sample data set, and a trained boron concentration distribution model is obtained. The method provided by the application can not only quickly and conveniently obtain the simulation result of the boron concentration distribution corresponding to each time, but also effectively improve the data resolution of the simulation result of the boron concentration distribution and improve the accuracy of the boron concentration distribution data, thereby providing a data basis for subsequent BNCT analysis and treatment scheme formulation.

Description

Technical Field

[0001] This application belongs to the field of boron concentration acquisition technology, and relates to a method, terminal and medium for acquiring a boron concentration distribution model based on melanoma, specifically involving a method for acquiring a boron concentration distribution model based on melanoma, a method for acquiring boron concentration distribution, a terminal and a medium. Background Technology

[0002] In the field of melanoma treatment, neutron capture therapy (BNCT) has unique advantages over traditional treatment methods (such as photons or carbon ions), including: (1) melanoma is resistant to traditional photon irradiation but not to BNCT, and BNCT is effective for both aerobic and hypoxic tumor cells; (2) boron drugs selectively accumulate in melanin and non-melanin tumor cells, which can precisely destroy cancer cells with complex shapes and reduce damage to normal tissues; (3) BNCT can deliver a high radiation dose to cancer cells in a single irradiation, while traditional photon irradiation requires 6 to 7 weeks to complete the same dose; based on this, BNCT has great potential in the treatment of malignant skin melanoma.

[0003] In existing technologies, there are relatively few studies on boron concentration distribution in melanoma. Although neutron excitation spectroscopy, mass spectrometry, and positron emission tomography-computed tomography (PET-CT) can be used to measure boron concentration in tissues for other types of tumors, the dispersed distribution of melanomas and the often small size of individual melanomas result in problems such as low resolution and long scan times for boron concentration data obtained by PET-CT. This makes it impossible to accurately and quickly obtain high-resolution boron concentration distribution information in melanoma tissue at various time points. Furthermore, boron concentration data obtained by PET-CT can only reflect the boron concentration distribution at a single time point and cannot characterize the spatiotemporal variation of boron concentration, thus reducing the effectiveness of boron non-invasive co-transfer therapy (BNCT) for melanoma.

[0004] Therefore, how to accurately and quickly obtain the boron concentration distribution within melanoma regions has become a technical problem that needs to be solved in this field. Summary of the Invention

[0005] In view of the shortcomings of the prior art described above, the purpose of this application is to provide a method for obtaining a boron concentration distribution model based on melanoma, a method for obtaining boron concentration distribution, a terminal and a medium, to solve the problem that the existing methods cannot quickly and accurately obtain the boron concentration distribution in the melanoma region.

[0006] In a first aspect, this application provides a method for obtaining a boron concentration distribution model based on melanoma, comprising:

[0007] The process involves acquiring the target region of the melanoma, the microneedle delivery region, and the drug concentration. The target region is the PET image acquisition area for boron concentration information, which covers an area larger than the melanoma. The delivery region is the area where boron diffuses after microneedle administration. PET image data of the target region at each sampling time after microneedle administration and microneedle boron concentration data of the delivery region at each sampling time are acquired. A sample dataset is constructed based on the PET image data and the corresponding microneedle boron concentration data. Iterative training is performed on a preset boron concentration distribution model based on the sample dataset to obtain a trained boron concentration distribution model. The boron concentration distribution model is a model that simulates the spatial distribution of boron concentration within the target region at subsequent times based on the PET image data at the initial time.

[0008] In some embodiments, the iterative training, performed in a single execution, includes:

[0009] Image features of the microneedle boron concentration data in the current sample data pair are extracted; based on the image features and the image feature prediction results corresponding to the previous sample data pair, the image feature prediction results corresponding to the current sample data pair are obtained through a time-series prediction method; image reconstruction is performed on the current image feature prediction results to obtain a simulated image of boron concentration distribution corresponding to the current sample data pair; image comparison is performed between the current simulated image of boron concentration distribution and the PET image data in the current sample data pair to obtain the image differences between the two, and the model parameters are optimized based on the image differences.

[0010] In some embodiments, during the first execution of the iterative training, the method for obtaining the image feature prediction result corresponding to the previous sample data includes:

[0011] Image features corresponding to the PET image data in the first sample data pair and image features corresponding to the microneedle boron concentration data in the first sample data pair are extracted respectively; feature fusion is performed on the image features of the two; and the image feature prediction result corresponding to the fused feature is obtained using the time-series prediction method.

[0012] In some embodiments, the method for obtaining the image features includes:

[0013] Using a multi-scale convolution module, feature extraction is performed on each of the microneedle boron concentration data to obtain the image features corresponding to each of the microneedle boron concentration data, as follows:

[0014]

[0015] in, Image features of boron concentration data for microneedles It is a multi-scale convolution module; The data includes microneedle boron concentration data; and, using the residual module, the spatial structure features of the PET image data in the first sample data pair are extracted to obtain the global spatial features of the PET image data, which are:

[0016]

[0017] in, Image features corresponding to PET image data. For residual modules; The data consists of PET image data from the first sample data pair.

[0018] In some embodiments, the method of obtaining the image features further includes:

[0019] Adaptive feature alignment is performed on the image features of each modality, as follows:

[0020]

[0021] in, The image features of the boron concentration data of the microneedles; The image features corresponding to the PET image data; This is an adaptive feature alignment module.

[0022] In some embodiments, the method for obtaining the boron concentration data of the microneedles includes:

[0023] Within the drug delivery area, sampling points at different sampling distances are determined, and the boron concentration at each sampling point at different sampling times is obtained. Based on the sampling distance, sampling time, and corresponding boron concentration, sampling data pairs after microneedle drug delivery are constructed. Based on each of the sampling data pairs, the pre-constructed boron drug diffusion function is fitted to obtain a fitted microneedle boron concentration diffusion model. Based on the sampling time of the PET image data, the fitted microneedle boron concentration diffusion model is used to obtain the microneedle boron concentration data corresponding to that sampling time.

[0024] In some embodiments, for each sampling point within the microneedle coverage area, the method for acquiring the microneedle boron concentration data at the initial time includes:

[0025] Based on the drug volume and boron concentration administered via microneedles, and combined with the tissue mass at the sampling point, the boron concentration corresponding to the sampling point is obtained as follows:

[0026]

[0027] Where V is the volume of drug solution administered via microneedle; C is the boron concentration administered via microneedle; and M is the tissue mass at the sampling point.

[0028] In a second aspect, this application provides a method for obtaining boron concentration distribution based on melanoma, comprising:

[0029] The drug concentration for microneedle administration is determined, and based on this drug concentration, a boron concentration distribution model corresponding to and trained with this drug concentration is determined; PET image data of the target region at the initial time is acquired; based on the PET image data, the boron concentration distribution model is used to acquire a simulated image of the boron concentration distribution at the target time; wherein, the boron concentration distribution model is acquired using any of the above-described methods for obtaining boron concentration distribution models based on melanoma.

[0030] In a third aspect, this application provides a terminal, including: a processor and a memory, wherein the memory and the processor are communicatively connected; the memory is used to store a computer program, and the processor is used to execute the computer program stored in the memory, so that the terminal executes the boron concentration distribution model acquisition method based on melanoma as described above, or executes the boron concentration distribution acquisition method based on melanoma as described above.

[0031] In a fourth aspect, this application provides a computer storage medium storing a computer program, which, when executed by a processor, implements the method for obtaining a boron concentration distribution model based on melanoma as described above, or executes the method for obtaining a boron concentration distribution based on melanoma as described above.

[0032] As described above, the boron concentration distribution model acquisition method, boron concentration distribution acquisition method, terminal, and medium based on melanoma provided in this application construct sample data pairs consisting of PET image data and microneedle boron concentration data at corresponding time points. The boron concentration distribution model is iteratively trained using each sample data pair to obtain a trained boron concentration distribution model. Based on this trained model and combined with the PET image data at the initial time point, simulated images of boron concentration distribution at subsequent time points can be obtained. Compared to existing technologies, the method in this application fully utilizes the measured boron concentration data after microneedle administration and the microneedle boron concentration data fitted based on the measured boron concentration data. That is, it fully utilizes the variation law of microneedle boron concentration and considers the spatial distribution characteristics of boron drug concentration in PET image data. This not only allows for the rapid and convenient acquisition of simulated boron concentration distribution results at each time point but also improves upon the low resolution of PET data, effectively increasing the data resolution of the boron concentration distribution simulation results and enhancing the accuracy of the boron concentration distribution data. This provides a better data foundation for subsequent BNCT analysis and calculation. Attached Figure Description

[0033] Figure 1The diagram shown is a flowchart illustrating the method for obtaining a boron concentration distribution model based on melanoma, as provided in the embodiments of this application.

[0034] Figure 2 The diagram shows a flowchart illustrating the method for acquiring PET image data as described in this application embodiment;

[0035] Figure 3 The diagram shows a flowchart illustrating the method for obtaining boron concentration data from microneedles as described in this application embodiment.

[0036] Figure 4 The diagram shows a flowchart of the iterative training process described in this embodiment during execution.

[0037] Figure 5 The diagram shows a flowchart of the method for obtaining boron concentration distribution based on melanoma as described in the embodiments of this application.

[0038] Figure 6 The diagram shown is a structural schematic of the terminal described in this application embodiment. Detailed Implementation

[0039] The following specific examples illustrate the implementation of this application. Those skilled in the art can easily understand other advantages and effects of this application from the content disclosed in this specification. This application can also be implemented or applied through other different specific embodiments, and various details in this specification can also be modified or changed based on different viewpoints and applications without departing from the spirit of this application. It should be noted that, unless otherwise specified, the following embodiments and features in the embodiments can be combined with each other.

[0040] It should be noted that in the following description, reference is made to the accompanying drawings, which illustrate several embodiments of this application. It should be understood that other embodiments may also be used, and changes in mechanical composition, structure, electrical system, and operation may be made without departing from the spirit and scope of this application. The following detailed description should not be considered limiting, and the scope of the embodiments of this application is defined only by the claims of the published patent. The terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the application. Spatial terms such as “upper,” “lower,” “left,” “right,” “below,” “below,” “lower part,” “above,” “upper part,” etc., may be used herein to illustrate the relationship between one element or feature shown in the figures and another element or feature.

[0041] Furthermore, as used herein, the singular forms “a,” “an,” and “the” are intended to include the plural forms as well, unless the context indicates otherwise. It should be further understood that the terms “comprising,” “including,” indicate the presence of the stated feature, operation, element, component, item, kind, and / or group, but do not preclude the presence, occurrence, or addition of one or more other features, operations, elements, components, items, kinds, and / or groups. The terms “or” and “and / or” as used herein are to be interpreted inclusively, or mean any one or any combination thereof.

[0042] To make the objectives, technical solutions, and advantages of the present invention clearer, the technical solutions in the embodiments of the present invention will be further described in detail below with reference to the accompanying drawings. It should be understood that the specific embodiments described herein are only for explaining the present invention and are not intended to limit the invention.

[0043] Before providing a further detailed description of the present invention, the nouns and terms used in the embodiments of the present invention are explained. The nouns and terms used in the embodiments of the present invention are subject to the following interpretations:

[0044] (1) PET-CT: Positron emission tomography / X-ray computed tomography is a combination of PET and CT. It integrates PET and CT into one unit and is controlled by a workstation. During scanning, PET and CT imaging are performed simultaneously as needed, and the workstation fuses the two images together to achieve better identification and localization.

[0045] (2) Microneedle drug delivery: the micron-sized needle tip pierces the stratum corneum of the skin to form a temporary mechanical channel, and the drug is released directly into the epidermis or dermis so that the drug enters the systemic circulation through the capillaries.

[0046] To address the technical problems in the prior art, this application provides a method for obtaining a boron concentration distribution model based on melanoma in the first aspect, which is used to obtain a boron concentration distribution model corresponding to the microneedle administration concentration under microneedle drug delivery method.

[0047] The boron concentration distribution model is used to simulate the spatial distribution of boron concentration in the melanoma region at subsequent times based on PET-CT image data (hereinafter referred to as PET image data) at an initial time; the initial time is a time when or after microneedle administration; the subsequent time is a time after the initial time.

[0048] Please see Figure 1 The diagram shows a flowchart illustrating the method for obtaining a boron concentration distribution model based on melanoma, as provided in an embodiment of this application. Figure 1 As shown, the method includes the following steps:

[0049] S10, Obtain the target area of ​​melanoma, and determine the microneedle delivery area and drug concentration for the same melanoma;

[0050] The target area is the PET acquisition area for boron concentration information, and the range of this acquisition area is larger than the range of the melanoma.

[0051] The drug delivery area is the diffusion area of ​​boron drug after microneedle administration, and the extent of this diffusion area is larger than the area of ​​the melanoma.

[0052] Specifically, after determining the coverage area of ​​the melanoma, the coverage area is expanded according to a preset expansion ratio, and the expanded area is used as the target area to ensure that the target area can cover the entire area of ​​the melanoma.

[0053] Furthermore, based on the microneedle administration location as the center and the boron drug diffusion radius of the microneedle administration, an administration area for microneedle administration is constructed; wherein, the boron drug diffusion radius is a preset value, which can be obtained based on experiments or experience.

[0054] It should be noted that the diffusion radius of the boron drug is positively correlated with the dosage concentration of the microneedle drug; that is, the higher the dosage concentration, the larger the diffusion radius of the boron drug, and vice versa.

[0055] S20, collect PET image data of the target area at different sampling times, and collect microneedle boron concentration data of the drug delivery area corresponding to the sampling time;

[0056] The PET image data is acquired via PET and represents the spatial distribution information of boron concentration within the target area.

[0057] The microneedle boron concentration data is data characterizing the spatial distribution of boron concentration within the administration area after microneedle administration.

[0058] Specifically, based on the area of ​​the target region, PET images of the region are acquired at several sampling times after microneedle drug administration to obtain PET image data of the target region.

[0059] Furthermore, based on the regional range of the drug delivery area, microneedle boron concentration data corresponding to each sampling time is obtained for that regional range.

[0060] In this application, the microneedle boron concentration data corresponds to the PET image data, and the sampling times of the two are the same or similar, that is, the sampling time difference between the two is less than a time difference threshold.

[0061] To ensure that PET image data accurately and comprehensively reflects the dynamic distribution of boron concentration over time after microneedle administration, in a specific embodiment, the method for acquiring the PET image data is as follows: Figure 2 As shown, it includes:

[0062] S21, determine each sampling time and acquire the PET image corresponding to each sampling time;

[0063] The PET images are the raw image data acquired during the PET scanning process.

[0064] Specifically, based on preset time intervals, a sampling time series for PET is constructed as follows:

[0065]

[0066] in, It is a sampled time series. This is the initial sampling time; Let n be the sampling time (n=1,2,...,n).

[0067] After microneedle drug delivery, a PET scan is initiated. At the sampling time, PET images are acquired at the corresponding time points to obtain the PET images for each time point, as follows:

[0068]

[0069] Where x, y, z are the coordinates of the PET image in three-dimensional space; H, W, D are the spatial resolutions of the PET image along the z-axis, x-axis, and y-axis, respectively; t i For sampling time series At any time in the process.

[0070] S22, perform image preprocessing on each of the PET image data to obtain preprocessed PET image data;

[0071] In this embodiment, the image preprocessing includes at least coordinate registration; that is, since there are some differences in the acquisition position, angle and other acquisition states at different times, the acquired PET images will have certain imaging deviations. In order to eliminate the imaging deviations in the image acquisition process, coordinate registration is performed on the PET images acquired at each time.

[0072] Specifically, based on the coordinates of the PET image at the initial time, coordinate registration is performed on PET images acquired at other times besides the initial time to ensure that PET images at different times are located in the same coordinate system, that is:

[0073]

[0074] in, For PET images before registration, The registered PET image; R is the PET image at the initial time, and R is the coordinate registration function used to align the PET images at each time (except the initial time) to the coordinate system corresponding to the initial time.

[0075] It should be noted that, in other embodiments, the image preprocessing may also include other existing preprocessing methods such as image enhancement and image filtering; for example, the image preprocessing process further includes:

[0076] Standardization processing was performed on the PET images acquired at each time point to eliminate the acquisition system bias corresponding to different PET images acquired at different times.

[0077] S23, extract the boron concentration information from the PET images to obtain PET image data corresponding to each PET image.

[0078] Specifically, for a single PET image, the boron concentration information corresponding to each pixel in the PET image is extracted to obtain the spatial distribution information of boron concentration in the PET image; the obtained spatial distribution information of boron concentration is used as the PET image data corresponding to the PET image.

[0079] To quickly, accurately, and conveniently obtain microneedle boron concentration data corresponding to the sampling time of PET image data, in a specific embodiment, the method for obtaining the microneedle boron concentration data is as follows: Figure 3 As shown, it includes:

[0080] S201, within the drug administration area, determine sampling points at different sampling distances;

[0081] The sampling point is a collection point used to collect boron concentration information within the tissue; the sampling distance is the distance between the sampling point and the area covered by the microneedle.

[0082] In one specific embodiment, microneedle drug delivery is performed using a microneedle array, with the area covered by the microneedle array serving as the drug delivery source area. Based on this drug delivery source area, different sampling areas for microneedle drug delivery are sequentially constructed according to a preset sampling interval 'a'. Specifically, the area located outside the drug delivery source area and at a distance 'a' from it is designated as the first sampling area; the area located outside the drug delivery source area and at a distance '2a' from it is designated as the second sampling area, and so on, to obtain each sampling area for the current microneedle drug delivery. Within each sampling area, several sampling points are randomly selected as the corresponding sampling points within the sampling area.

[0083] S202, Obtain the boron concentration at each sampling point at different sampling times, and construct sampling data pairs after microneedle drug delivery based on sampling distance, sampling time and corresponding boron concentration;

[0084] Each of the sampled data pairs includes the sampling interval, sampling time, and corresponding boron concentration value corresponding to the sampling point;

[0085] Specifically, based on a preset time interval, several sampling times are determined after the microneedle drug delivery ends; through tissue sampling and boron concentration measurement, the boron concentration values ​​corresponding to each sampling point at different sampling times are obtained to construct each sampling data pair.

[0086] It should be noted that the tissue sampling can be carried out using existing biological tissue sampling methods, and the boron concentration determination can be carried out using existing in vitro biological determination methods, such as mass spectrometry or chromatography, which can accurately quantify the boron content, i.e., boron concentration.

[0087] S203, Based on each of the sampling data pairs, the pre-constructed boron drug diffusion function is fitted to obtain the fitted microneedle boron concentration diffusion model;

[0088] The boron drug diffusion function is used to characterize the spatiotemporal diffusion pattern of boron drug within the drug delivery area after microneedle administration.

[0089] In this embodiment, to facilitate the acquisition of boron concentration data, the diffusion process of the boron drug within the drug delivery area is considered as a uniform diffusion process. Therefore, the constructed boron drug diffusion function is:

[0090] B=C(r,t,D)

[0091] Where C is the boron diffusion function, B is the measured boron concentration, r is the sampling distance, t is the sampling time, and D is a preset diffusion coefficient used to reflect the diffusion ability of the drug.

[0092] Each of the sampled data pairs is input into the boron drug diffusion function, and the boron drug diffusion function is fitted using each sampled data pair to obtain the fitted boron drug diffusion function.

[0093] For example, the boron diffusion function is a polynomial linear function.

[0094] It should be noted that, for each sampling point within the microneedle coverage area, in order to obtain the boron concentration at the initial time (microneedle administration time) of the sampling point more quickly and conveniently, the boron concentration corresponding to the sampling point is obtained based on the drug volume and boron concentration administered by the microneedle, combined with the tissue mass of the sampling point, as follows:

[0095]

[0096] Where V is the volume of drug solution administered via microneedle; C is the boron concentration administered via microneedle; and M is the tissue mass at the sampling point.

[0097] S204, based on the sampling time of PET image data, the boron concentration data of microneedles corresponding to the sampling time is obtained by using the fitted microneedle boron concentration diffusion model.

[0098] Specifically, the sampling time corresponding to the PET image data is obtained, and this sampling time is input into the fitted microneedle boron concentration diffusion model to obtain the change function between boron concentration and sampling distance, that is, the functional relationship of boron concentration changing with sampling distance; based on the functional relationship, the spatial distribution of microneedle boron concentration in the drug delivery area is obtained, and this spatial distribution is used as the microneedle boron concentration data corresponding to the sampling time.

[0099] S30, Based on the PET image data and the corresponding microneedle boron concentration data, construct a sample dataset;

[0100] The sample dataset includes multiple sample data pairs;

[0101] The single sample data pair includes the PET image data and the microneedle boron concentration data at the sampling time corresponding to the PET image data.

[0102] The sampling time of the PET image data is obtained; based on the sampling time, the boron concentration data of the microneedles with the same sampling time and the PET image data are combined into a sample data pair;

[0103] The microneedle boron concentration data serves as the input data for the sample pair, and the PET image data serves as the label data for the sample pair.

[0104] This step is performed on each PET image data to obtain sample data pairs;

[0105] The sample data pairs are arranged in chronological order according to their corresponding sampling times to obtain the sample dataset.

[0106] S40, Based on the sample dataset, the pre-built boron concentration distribution model is trained to obtain the trained boron concentration distribution model.

[0107] The boron concentration distribution model includes a feature extraction module, a time-series prediction module, an image reconstruction module, and an image comparison module.

[0108] The feature extraction module is used to extract image features from boron concentration data to obtain high-dimensional feature information.

[0109] The time series prediction module uses a Long Short-Term Memory (LSTM) network to obtain the dependence of boron concentration distribution in the time dimension. Through the feature sequences of continuous time points, it learns the feature transmission and change patterns between different time points.

[0110] The image reconstruction module is used to convert boron concentration distribution features into a simulated image of boron concentration distribution;

[0111] The image comparison module is used to compare the differences between simulated boron concentration distribution images and PET images.

[0112] Specifically, the boron concentration distribution model is iteratively trained using each sample data pair in the sample dataset to obtain the trained boron concentration distribution model.

[0113] To facilitate the description of this process, the following will take a single sample data pair as an example to describe the single iteration training process in detail;

[0114] like Figure 4 As shown, for a single sample data pair (the i-th sample data pair) as an example, the single iteration training process includes the following during execution:

[0115] Using a feature extraction module, image features of the boron concentration data of the microneedles in the current sample data pair are obtained; and image feature prediction results corresponding to the previous sample data are obtained; wherein, the image feature prediction results are feature processing results obtained by using an LSTM network to process the previous sample data pair;

[0116] The image features, along with the image feature prediction results corresponding to the previous sample data, are input into the LSTM network to obtain the image feature prediction results corresponding to the current sample data.

[0117] The image reconstruction module is used to perform image reconstruction on the current image feature prediction results to obtain a simulated image of the boron concentration distribution corresponding to the current sample data.

[0118] Furthermore, based on the image comparison module, an image comparison is performed between the current simulated boron concentration distribution image and the PET image in the current sample data pair to obtain the image differences between the two, so as to optimize the parameters of the model based on the image differences.

[0119] Based on the above steps, the boron concentration distribution model is trained using each sample data pair to obtain the trained boron concentration distribution model.

[0120] It should be noted that, during the first model iteration training process, the image feature prediction result corresponding to the previous sample data is the image feature prediction result obtained based on the first sample data;

[0121] Specifically, the method for obtaining the image feature prediction results corresponding to the above sample data includes:

[0122] Using the feature extraction module, image features corresponding to PET image data and microneedle boron concentration data in the first sample data pair are extracted respectively; that is, the PET image data corresponding to the initial time t0 is input into the feature extraction module to obtain the image features corresponding to the PET image data; and the microneedle boron concentration data corresponding to the initial time t0 is input into the feature extraction module to obtain the image features corresponding to the microneedle boron concentration data.

[0123] The image features corresponding to the PET image data and the image features of the microneedle boron concentration data are fused to obtain the fused features corresponding to the initial time t0; the fused features are then input into an LSTM network for processing to obtain the image feature prediction results corresponding to the first sample data.

[0124] When performing iterative training of the model corresponding to the second sample data pair, the image feature prediction results corresponding to the first sample data pair, combined with the image features of the boron concentration data of the microneedles in the second sample data pair, are input into the LSTM network for processing to obtain the image feature prediction results corresponding to the second sample data pair. The image reconstruction module is then used to reconstruct the image based on the image feature prediction results to obtain the simulated image of the boron concentration distribution corresponding to the second sample data pair. The image comparison module is then used to compare the simulated image of the boron concentration distribution corresponding to the second sample data pair with the PET concentration data in the second sample data pair to obtain the corresponding comparison results. Based on the comparison results, the current boron concentration distribution model is optimized to obtain a new boron concentration distribution model. The new boron concentration distribution model is then iteratively trained using subsequent sample data pairs to obtain the trained model.

[0125] In a more specific embodiment, the LSTM network is:

[0126]

[0127] in, This is the output of the LSTM network, i.e., the image feature prediction result; This is the image feature prediction result corresponding to the first sample data pair, that is, the image feature prediction result corresponding to the initial time t0; This is the image feature prediction result corresponding to the second sample data pair, that is, the image feature prediction result corresponding to the first sampling time t1; The prediction result of the image features corresponding to the last sample data, that is, the last sampling time t. n The corresponding image feature prediction results.

[0128] Based on the learned time-series features, a fully connected layer or convolutional network is used to map the time-series features to the image feature prediction results corresponding to each sampling time, thus obtaining the image feature prediction results for each sampling time, as follows:

[0129]

[0130] in, For the i-th sampling time t i Image feature prediction results; This is a temporal mapping function (usually a multilayer perceptron or convolutional network); For the i-th sampling time t i The corresponding network parameters.

[0131] In one specific embodiment, the image reconstruction module employs a convolutional neural network including deconvolutional layers, as follows:

[0132]

[0133] in, For the i-th sampling time t i Image feature prediction results; Simulated image of the reconstructed boron concentration distribution.

[0134] To ensure the network generates accurate and stable results during training, specifically high-precision simulated images of boron concentration distribution, a total loss function is constructed in a particular embodiment based on mean squared error (MSE) loss, attention constraint loss, feature alignment constraint loss, microneedle concentration constraint loss, and biological sample constraint loss. The function is as follows:

[0135]

[0136] in, Image resolution loss, used to measure the global error of super-resolution;

[0137] To minimize data loss during microneedle measurements and ensure the accuracy of local concentrations;

[0138] To mitigate data loss and ensure the accuracy of concentration across time and space;

[0139] To mitigate attention-constrained loss, optimize the network's focus on key regions;

[0140] The feature alignment loss reduces feature differences between modalities;

[0141] The weights of each constraint term.

[0142] In one specific embodiment, the feature extraction module includes a multi-scale convolution module. The method for extracting image features of the microneedle boron concentration data using the feature extraction module includes:

[0143] Using a multi-scale convolution module, feature extraction is performed on the boron concentration data of each microneedle to obtain the diffusion characteristics of boron drugs in local tissues, reflecting the multi-scale spatial characteristics of boron concentration in the drug delivery area, as follows:

[0144]

[0145] in, Image features of boron concentration data for microneedles It is a multi-scale convolution module; This is data on boron concentration in microneedles.

[0146] Furthermore, in order to further improve the attention paid to key regions in the image by the time-series prediction module, the feature extraction module also includes a global attention mechanism (GA) module and a channel attention mechanism (CA) module to dynamically identify key feature regions in the image and assign them higher weights, thereby improving the accuracy of boron concentration reconstruction.

[0147] The Global Attention (GA) module is used to capture key global features and optimize the reconstruction of the overall boron concentration distribution; the Channel Attention (CA) module is used to weight feature channels, highlight effective features, and reduce invalid information.

[0148] The two attention mechanisms are weighted and fused together as follows:

[0149] .

[0150] Furthermore, the feature extraction module also includes a residual module. Therefore, the implementation method of using the feature extraction module to extract the image features corresponding to the PET image data in the first sample data pair includes:

[0151] Using the Residual module, feature extraction is performed on the PET image data to obtain the global spatial features of the PET image data, which are:

[0152]

[0153] in, Image features corresponding to PET image data. For residual modules; The data consists of PET image data from the first sample data pair.

[0154] In one specific embodiment, the method for performing feature fusion of the image features corresponding to the PET image data and the image features of the microneedle boron concentration data is as follows:

[0155]

[0156] in, As a feature of fusion, This is the feature fusion module.

[0157] Furthermore, due to the differences in the distribution of different modal data in the feature space, feature mismatch may occur, thereby affecting the model prediction accuracy. To improve the matching degree between features of different modalities, the process of performing feature fusion on the image features corresponding to the PET image data and the image features of the microneedle boron concentration data further includes:

[0158] Adaptive feature alignment is performed on image features of each modality. This involves adjusting the feature contribution of different modalities through dynamic weighting to learn the mapping relationship between features. This allows features of different modalities to be aligned in the same feature space, thereby reducing feature deviation between modalities and achieving effective fusion of information from each modality in the feature space.

[0159] More specifically, the adaptive feature alignment is as follows:

[0160]

[0161] in, The image features of the boron concentration data of the microneedles; The image features corresponding to the PET image data; This is an adaptive feature alignment module.

[0162] Based on the same inventive concept, this application also provides a method for obtaining boron concentration distribution based on melanoma, which is used to obtain a simulated image of boron concentration distribution at a certain target time after microneedle drug administration;

[0163] The target time is a pre-set time, which can be any time after the microneedle is administered;

[0164] The simulated boron concentration distribution image is a set of simulated data, which is used to simulate the spatial distribution of boron concentration at a certain moment after microneedle administration.

[0165] In this embodiment, the method for obtaining boron concentration distribution based on melanoma is as follows: Figure 5 As shown, it includes:

[0166] S1, determine the drug concentration for microneedle delivery; based on the drug concentration, determine the boron concentration distribution model corresponding to the drug concentration;

[0167] The boron concentration distribution model is used to simulate the spatial distribution of boron concentration in the melanoma region at subsequent times based on PET image data at an initial time; the initial time is a time at or after microneedle administration; and the subsequent times are times located after the initial time.

[0168] In this embodiment, the boron concentration distribution model is a pre-trained model; specifically, the boron concentration distribution model is trained using the boron concentration distribution model acquisition method provided in the above embodiment, which will not be repeated here.

[0169] S2, acquire PET image data of the target area at the initial time; based on the PET image data, use the boron concentration distribution model to acquire the simulated boron concentration distribution image at the target time.

[0170] Specifically, a certain moment during or after microneedle administration is taken as the initial moment; PET image data corresponding to the initial moment is sampled; the PET image data and the target moment are input into the boron concentration distribution model to output a simulated boron concentration distribution image corresponding to the target moment.

[0171] Based on the same technical concept, the method for obtaining a boron concentration distribution model based on melanoma or the method for obtaining boron concentration distribution based on melanoma provided in the embodiments of the present invention can be implemented on the terminal side or the server side.

[0172] Please see Figure 6 This is a schematic diagram of an optional hardware structure of a terminal provided in an embodiment of the present invention. The terminal can be a mobile phone, computer device, tablet device, personal digital processing device, factory back-end processing device, etc. The terminal includes: at least one processor 61, a memory 62, at least one network interface 64, and a user interface 63. The various components in the device are coupled together through a bus system 65. It is understood that the bus system 65 is used to realize the connection and communication between these components. In addition to a data bus, the bus system also includes a power bus, a control bus, and a status signal bus.

[0173] The user interface 63 may include a monitor, keyboard, mouse, trackball, clicker, button, touchpad, or touch screen.

[0174] It is understood that memory 62 can be volatile memory or non-volatile memory, or both. Non-volatile memory can be read-only memory (ROM) or programmable read-only memory (PROM), used as an external cache. By way of example, but not limitation, many forms of RAM are available, such as static random access memory (SRAM) and synchronous static random access memory (SSRAM). The memory characterized in the embodiments of the present invention is intended to include, but is not limited to, these and any other suitable categories of memory.

[0175] In this embodiment of the invention, the memory 62 is used to store various types of data to support the operation of the terminal. Examples of this data include: any executable program for operation on the terminal 60, such as the operating system 621 and application programs 622; the operating system 621 contains various system programs, such as the framework layer, core library layer, driver layer, etc., for implementing various basic services and handling hardware-based tasks. Application programs 622 may contain various applications, such as media players, browsers, etc., for implementing various application services. The method for obtaining the boron concentration distribution model based on melanoma provided in this embodiment of the invention, or the method for obtaining the boron concentration distribution based on melanoma, can be included in application program 622.

[0176] The methods disclosed in the above embodiments of the present invention can be applied to processor 61, or implemented by processor 61. Processor 61 may be an integrated circuit chip with signal processing capabilities. In the implementation process, each step of the above method can be completed by the integrated logic circuit of the hardware in processor 61 or by instructions in the form of software. The processor mentioned above may be a general-purpose processor, a digital signal processor (DSP), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. Processor 61 can implement or execute the methods, steps and logic block diagrams disclosed in the embodiments of the present invention. General-purpose processor 61 may be a microprocessor or any conventional processor, etc. The steps of the accessory optimization method provided in the embodiments of the present invention can be directly reflected as being executed by a hardware decoding processor, or being executed by a combination of hardware and software modules in the decoding processor. The software module may be located in a storage medium, which is located in memory. The processor reads the information in the memory and combines it with its hardware to complete the steps of the aforementioned method.

[0177] In an exemplary embodiment, terminal 60 may be used by one or more application-specific integrated circuits (ASICs), DSPs, programmable logic devices (PLDs), or complex programmable logic devices (CPLDs) to execute the aforementioned method.

[0178] This invention also provides a computer-readable storage medium storing a computer program thereon, which, when called by a processor, implements the method for obtaining a boron concentration distribution model based on melanoma or the method for obtaining boron concentration distribution based on melanoma provided by this invention.

[0179] Computer-readable storage media can be tangible devices capable of holding and storing instructions used by an instruction execution device. Computer-readable storage media can be, for example, (but not limited to) electrical storage devices, magnetic storage devices, optical storage devices, electromagnetic storage devices, semiconductor storage devices, or any suitable combination of the foregoing. More specific examples (a non-exhaustive list) of computer-readable storage media include: portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), static random access memory (SRAM), portable compact disc read-only memory (CD-ROM), digital multifunction disc (DVD), memory sticks, floppy disks, and mechanical encoding devices.

[0180] The computer-readable program represented herein can be downloaded from a computer-readable storage medium to various computing / processing devices, or downloaded via a network, such as the Internet, a local area network, a wide area network, and / or a wireless network, to an external computer or external storage device. A network adapter card or network interface in each computing / processing device receives computer-readable program instructions from the network and forwards these instructions to the computer-readable storage medium in the respective computing / processing device.

[0181] In summary, the boron concentration distribution model acquisition method, boron concentration distribution acquisition method, terminal, and medium provided in this application, by jointly constructing sample data pairs from different modal information of low-resolution PET images and microneedle drug delivery diffusion data, fully expresses and utilizes the boron concentration distribution features contained in different data sources. Furthermore, by iteratively training the boron concentration distribution model using each sample data pair, a trained boron concentration distribution model is obtained, which accurately reflects the spatiotemporal patterns of boron drug diffusion within the melanoma region. Compared to existing technologies, the method described in this application fully utilizes the boron concentration variation patterns of microneedles and considers the spatial distribution characteristics of boron drug concentration in PET image data. Therefore, it can not only quickly and conveniently obtain simulation results of boron concentration distribution at various times, but also... This invention also addresses the low resolution of PET data, effectively improving the data resolution of boron concentration distribution simulation results and enhancing the accuracy of boron concentration distribution data. Furthermore, by performing feature alignment on the data sources of each modality, this application effectively avoids information loss and incompleteness issues in a single modality, effectively reducing feature mismatch between different modalities, improving overall feature representation capabilities, and making the reconstructed boron concentration distribution more accurate. Additionally, the introduction of an attention mechanism allows the network to dynamically focus on key regions of the boron concentration distribution, reducing resource waste in non-critical areas, thereby improving the accuracy of the reconstruction results in key regions and enhancing the efficiency and real-time performance of model reconstruction. Moreover, the design of the joint supervised loss function ensures the network's stability and efficiency during training, enabling it to achieve a balanced optimization objective under different constraints and effectively preventing overfitting. The network architecture of this invention exhibits good generalization ability under different scenarios and data sources, adapting to various boron concentration distribution reconstruction tasks and providing reliable technical support for practical medical applications.

[0182] This invention achieves high-precision and high-stability reconstruction of boron concentration distribution during BNCT treatment of malignant melanoma through the design of multimodal information fusion, adaptive feature alignment, attention mechanism, and joint supervision loss function. It significantly improves the predictability and controllability of subsequent BNCT treatment effects and has important theoretical value and application prospects.

[0183] The above embodiments are merely illustrative of the principles and effects of this application and are not intended to limit this application. Any person skilled in the art can modify or alter the above embodiments without departing from the spirit and scope of this application. Therefore, all equivalent modifications or alterations made by those skilled in the art without departing from the spirit and technical concept disclosed in this application should still be covered by the claims of this application.

Claims

1. A melanoma-based boron concentration distribution model acquisition method, comprising: acquiring a target region of a melanoma, a microneedle administration region, and an administration concentration; the target region is a PET acquisition region of boron concentration information, which covers and is larger in range than the melanoma; the administration region is a boron drug diffusion region after microneedle administration; based on PET image data and microneedle boron concentration data at the corresponding sampling time, a sample data set is constructed; based on the sample data set, the preset boron concentration distribution model is iteratively trained to obtain a trained boron concentration distribution model; wherein the boron concentration distribution model is a model for simulating the spatial distribution of boron concentration in the target region at a subsequent time based on PET image data at an initial time; when the iterative training is performed once, it includes: extracting image features of the microneedle boron concentration data in the current sample data pair; based on the image features and the image feature prediction result corresponding to the previous sample data pair, the image feature prediction result corresponding to the current sample data pair is obtained by a time series prediction method; image reconstruction is performed on the current image feature prediction result to obtain a boron concentration distribution simulation image corresponding to the current sample data pair; image comparison is performed on the current boron concentration distribution simulation image and the PET image data in the current sample data pair to obtain the image difference therebetween, and the model parameters are optimized based on the image difference.

2. The method of claim 1, wherein, When the iterative training is performed for the first time, the manner of obtaining the image feature prediction result corresponding to the previous sample data pair comprises: extracting image features corresponding to the PET image data in the first sample data pair and extracting image features corresponding to the microneedle boron concentration data in the first sample data pair, respectively; performing feature fusion on the image features of the two; using the time series prediction method to obtain the image feature prediction result corresponding to the fused features.

3. The method of claim 2, wherein the boron concentration profile model is obtained by: The manner of obtaining the image features comprises: using a multi-scale convolution module to perform feature extraction on each of the microneedle boron concentration data to obtain image features corresponding to each of the microneedle boron concentration data, which are: ； wherein, is an image feature of the microneedle boron concentration data, is a multi-scale convolution module; is microneedle boron concentration data; and, using a residual module to extract spatial structure features in the PET image data in the first sample data pair to obtain global spatial features of the PET image data, which are: ； wherein, image features corresponding to the PET image data, is a residual module; is the PET image data of the first sample data pair.

4. The method of claim 3, wherein the boron concentration profile model is obtained by: The manner of obtaining the image features further comprises: performing adaptive feature alignment on the image features of each modality, which are: ； wherein, is an image feature of the micro-needle boron concentration data; is an image feature corresponding to the PET image data; is an adaptive feature alignment module.

5. The method of claim 1, wherein the method further comprises: The manner of obtaining the microneedle boron concentration data comprises: determining each sampling point at different sampling distances within the administration region and obtaining the boron concentration of each sampling point at different sampling times to construct a sampling data pair after microneedle administration based on the sampling distance, the sampling time, and the corresponding boron concentration; based on each of the sampling data pairs, a boron drug diffusion function constructed in advance is fitted to obtain a fitted microneedle boron concentration diffusion model; based on the sampling time of the PET image data, the fitted microneedle boron concentration diffusion model is used to obtain the microneedle boron concentration data corresponding to the sampling time.

6. The method of claim 5, wherein the boron concentration profile model is obtained by: For each sampling point in the microneedle coverage region, at the initial time, the manner of obtaining the microneedle boron concentration data comprises: Based on the micro-needle drug liquid volume and boron concentration, combined with the tissue mass of the sampling point, the boron concentration corresponding to the sampling point is obtained, which is: ； Wherein, V is the micro-needle drug liquid volume; C is the micro-needle drug boron concentration; M is the tissue mass of the sampling point.

7. A melanoma-based boron concentration distribution acquisition method, comprising: determining the administration concentration of the micro-needle administration, based on the administration concentration, determining the trained boron concentration distribution model corresponding to the administration concentration; acquiring PET image data of the target region at the initial time; based on the PET image data, using the boron concentration distribution model to obtain the corresponding boron concentration distribution simulation image at the target time; Wherein, the boron concentration distribution model is obtained by the melanoma-based boron concentration distribution model acquisition method of any one of claims 1 to 6.

8. A terminal, characterized by comprising: comprising: a processor and a memory, the memory being in communication with the processor; the memory is used to store computer programs, and the processor is used to execute the computer programs stored in the memory, so that the terminal executes the melanoma-based boron concentration distribution model acquisition method of any one of claims 1 to 6, or executes the melanoma-based boron concentration distribution acquisition method of claim 7.

9. A computer storage medium storing a computer program, the computer program comprising instructions, which, when executed by a computer, cause the computer to perform the method according to any one of claims 1 to 8. The computer program is executed by the processor to realize the melanoma-based boron concentration distribution model acquisition method of any one of claims 1 to 6, or realize the melanoma-based boron concentration distribution acquisition method of claim 7.

Images