Dosage calculation method and system based on radiation environment monitoring data

Abstract

The application relates to the technical field of radiation environment monitoring, and discloses a radiation environment monitoring data-based dose calculation method and system. The method comprises the following steps: acquiring radiation source data, calculating an initial radiation field distribution based on the data, obtaining preliminary dose deposition values of each spatial point in a monitoring area, analyzing particle interaction, and determining an energy redistribution coefficient of each spatial point; adjusting the preliminary dose deposition values point by point by using the coefficient to obtain a corrected three-dimensional dose concentration field data, calculating a gradient vector between spatial points, and generating an optimized gradient distribution map; extracting a boundary point set with a gradient change exceeding a threshold value, constructing an isosurface grid to form a three-dimensional isosurface structure, acquiring an iteration update coefficient of time sequence radiation source data, and obtaining a concentration field time sequence evolution sequence; identifying and marking a high concentration area based on the sequence, fusing the high concentration area and the three-dimensional isosurface structure, and generating a visual distribution model reflecting the dose concentration. The application improves the accuracy of radiation dose calculation.

Description

Technical Field

[0001] This application relates to the field of radiation environment monitoring technology, and in particular to a dose calculation method and system based on radiation environment monitoring data. Background Technology

[0002] Radiation environment monitoring is a core technological support for ensuring the safe operation of nuclear facilities, radiation pollution prevention and control, and emergency response. It is directly related to public health, ecological environment safety, and social stability. Accurate calculation and dynamic tracking of radiation dose are key technological requirements in this field. In practical scenarios such as radiation leak emergency response and medical radiation therapy planning, dose calculation is needed to clarify the spatial distribution, intensity changes, and temporal evolution of radiation particles within the monitoring area, providing a scientific basis for risk assessment, protective measure formulation, and emergency decision-making.

[0003] Current mainstream dose calculation methods primarily rely on Monte Carlo simulations, but this technology has significant drawbacks: First, it can only output statistically significant initial dose deposition values, making it difficult to accurately capture the energy redistribution effects caused by particle collisions and scattering. This results in blurred boundaries in the three-dimensional dose concentration field, insufficient accuracy in identifying high-concentration risk areas, and an inability to meet the needs of refined monitoring. Second, in dynamic scenarios, it lacks a collaborative iterative mechanism between the energy redistribution coefficient and spatial structure. After the time-series radiation source data is updated, the correction process becomes disconnected from boundary extraction, causing discontinuities in the temporal evolution sequence of the concentration field. The radiation diffusion trajectory and intensity changes cannot be accurately reproduced, and in emergency scenarios, data distortion can easily affect emergency response efficiency. Furthermore, existing technologies lack sufficient depth in integrating monitoring data, failing to fully incorporate environmental data such as medium properties and terrain features to optimize the model. The dose distribution results also have low visualization capabilities, making it difficult to intuitively present dynamic changes in high-concentration areas. This prevents monitoring personnel from quickly grasping the risk situation, hindering the practical application of the technology.

[0004] Therefore, there is a pressing need to develop a dose calculation method based on radiation environment monitoring data that combines accuracy, dynamism, and visualization to address the pain points of existing technologies. Summary of the Invention

[0005] To address the aforementioned technical issues, this application provides a dose calculation method and system based on radiation environment monitoring data, which is used to characterize the spatiotemporal evolution of particle interactions and radiation fields, significantly improving the accuracy and dynamic analysis capabilities of radiation dose calculation, while also enabling the visualization of dose distribution.

[0006] In a first aspect, this application provides a dose calculation method based on radiation environment monitoring data, the method comprising:

[0007] Step S1: Obtain radiation source data, calculate the initial radiation field distribution based on the radiation source data, obtain the preliminary dose deposition value at each spatial point in the monitoring area, analyze the inter-particle interaction reflected by the preliminary dose deposition value, and determine the energy redistribution coefficient corresponding to each spatial point;

[0008] Step S2: Using the energy redistribution coefficient, the initial dose deposition value is adjusted point by point to obtain the corrected three-dimensional dose concentration field data. Based on the corrected three-dimensional dose concentration field data, the gradient vector between each spatial point is calculated to obtain the optimized gradient distribution map.

[0009] Step S3: Extract the set of boundary points whose gradient changes exceed a preset threshold from the optimized gradient distribution map, construct an isosurface grid based on the boundary point set to form a three-dimensional isosurface structure, obtain time series radiation source data, iteratively update the energy redistribution coefficient, and obtain the temporal evolution sequence of the concentration field.

[0010] Step S4: Based on the temporal evolution sequence of the concentration field, identify and mark high-concentration regions, fuse the high-concentration regions and the three-dimensional isosurface structure, and generate a visual distribution model reflecting the dose concentration based on the fusion result.

[0011] Secondly, this application provides a dose calculation system based on radiation environment monitoring data, the system comprising:

[0012] The acquisition module is used to acquire radiation source data, calculate the initial radiation field distribution based on the radiation source data, obtain the preliminary dose deposition value at each spatial point in the monitoring area, analyze the inter-particle interaction reflected by the preliminary dose deposition value, and determine the energy redistribution coefficient corresponding to each spatial point.

[0013] The optimization module is used to adjust the initial dose deposition value point by point using the energy redistribution coefficient to obtain the corrected three-dimensional dose concentration field data. Based on the corrected three-dimensional dose concentration field data, the gradient vector between each spatial point is calculated to obtain the optimized gradient distribution map.

[0014] The iterative update module is used to extract the set of boundary points whose gradient changes exceed a preset threshold from the optimized gradient distribution map, construct an isosurface grid based on the boundary point set to form a three-dimensional isosurface structure, acquire time-series radiation source data, iteratively update the energy redistribution coefficient, and obtain the temporal evolution sequence of the concentration field.

[0015] The visualization module is used to identify and mark high-concentration regions based on the temporal evolution sequence of the concentration field, fuse the high-concentration regions and the three-dimensional isosurface structure, and generate a visualization distribution model reflecting the dose concentration based on the fusion result.

[0016] Compared with the prior art, the beneficial effects of the present invention are at least as follows:

[0017] The dose calculation method and system based on radiation environment monitoring data provided in this application significantly improve the accuracy, real-time performance, and visualization of radiation dose calculation through multi-level technical means. This method first calculates the initial radiation field distribution and obtains preliminary dose deposition values ​​based on radiation source data. Then, it determines the energy redistribution coefficient by analyzing inter-particle interactions. Combining finite difference and gradient descent optimization processes, it accurately quantifies the energy transfer behavior within the radiation field, effectively improving the shortcomings of traditional methods in characterizing energy exchange.

[0018] Furthermore, by using the optimized coefficients to adjust the dose values ​​point by point, a three-dimensional dose concentration field is constructed and the gradient distribution is extracted. Then, based on the gradient changes, the boundary point set is extracted and an accurate isosurface structure is constructed, achieving a fine description of the spatial morphology of the radiation field. On this basis, time-series radiation source data is introduced, and by iteratively updating the energy redistribution coefficients and simulating the dynamic evolution of the radiation field, a concentration field evolution sequence reflecting the changes in the time dimension is formed, enhancing the model's ability to describe the radiation diffusion and attenuation process.

[0019] Finally, by integrating high-concentration area information with three-dimensional isosurface structure, a visualized distribution model with temporal evolution characteristics is generated. This model has achieved significant improvements in rendering coverage, precision control, real-time response, and dose calculation accuracy, and can provide more reliable, intuitive, and dynamically analytical dose distribution visualization support for application scenarios such as environmental radiation assessment and medical radiation planning. Attached Figure Description

[0020] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0021] Figure 1 This is a flowchart of the dose calculation method based on radiation environment monitoring data in the embodiments of this application;

[0022] Figure 2 This is an example diagram of the local scope distribution in an embodiment of this application;

[0023] Figure 3 This is a schematic diagram of a triangular mesh surface according to an embodiment of this application;

[0024] Figure 4 This is a comparison chart of the conventional method in this application and the final visualized distribution model in this application in terms of key performance indicators;

[0025] Figure 5 This is a schematic diagram of the dose calculation system based on radiation environment monitoring data according to an embodiment of this application. Detailed Implementation

[0026] This application provides a method and system for dose calculation based on radiation environment monitoring data. The terms "first," "second," "third," "fourth," etc. (if present) in the specification, claims, and accompanying drawings are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments described herein can be implemented in a sequence other than that illustrated or described herein. Furthermore, the terms "comprising" or "having" and any variations thereof are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0027] For ease of understanding, the specific process of the embodiments of this application is described below. Please refer to [link / reference]. Figure 1 One embodiment of the dose calculation method based on radiation environment monitoring data in this application includes:

[0028] Step S1: Obtain radiation source data, calculate the initial radiation field distribution based on the radiation source data, obtain the preliminary dose deposition value at each spatial point in the monitoring area, analyze the inter-particle interaction reflected by the preliminary dose deposition value, and determine the energy redistribution coefficient corresponding to each spatial point.

[0029] In step S1, obtaining the preliminary dose deposition values ​​at each spatial point within the monitoring area includes: extracting radiation source data from a preset database, including the location, energy, and type of the radiation source; processing the radiation source data using the Monte Carlo simulation method to simulate the propagation path and energy deposition process of particles in the monitoring area; calculating the dose deposition value at each spatial point based on the simulation results, generating distribution data of the preliminary dose deposition values, performing gridding processing on the distribution data of the preliminary dose deposition values, constructing a mapping relationship between spatial points and dose values, and outputting the gridded preliminary dose deposition values.

[0030] Specifically, to accurately obtain basic radiation dose distribution data within the monitoring area and provide reliable input for subsequent particle interaction analysis and energy redistribution coefficient calculation, preliminary dose deposition values ​​with both accuracy and adaptability need to be generated through professional simulation and data processing workflows. Specifically, radiation source data is first extracted from a pre-built database, a dedicated database for radiation environment monitoring, storing standardized data on various radiation sources, including radiation source parameters for scenarios such as medical radiation equipment and environmental radiation monitoring stations. The radiation source data specifically includes the location coordinates, energy parameters, and particle types of the radiation source. The location coordinates are the specific spatial points of the radiation source in the three-dimensional coordinate system of the monitoring area. Energy parameters refer to the energy magnitude and energy spectrum distribution of the particles released by the radiation source, including core parameters such as peak energy and average energy. Particle types refer to the types of radiation particles released by the radiation source, commonly including photons, such as gamma-ray photons and X-ray photons; and particle types, such as neutrons, beta particles, and alpha particles. These parameters are directly read by querying the index structure of the pre-built database, ensuring real-time data extraction. Next, the radiation source data is processed using the Monte Carlo simulation method. Monte Carlo simulation is a statistical method based on random sampling, suitable for simulating the random scattering and absorption processes of particles such as gamma rays and neutrons in a medium. First, the particle source is initialized, and the number of simulated particles is set to N, for example, N=1,000,000 in a typical environmental monitoring scenario, to ensure basic statistical accuracy. Then, a random number generator is used to simulate the initial direction and initial energy of the particles starting from the source point. The initial direction is randomly distributed within the range of 0°-360°, and the initial energy matches the distribution of the input energy parameters. Subsequently, the interaction events between the particles and the medium in the monitoring area are calculated. Different particle types correspond to different interaction mechanisms. For example, photons mainly undergo Compton scattering and photoelectric effect, while neutrons mainly undergo elastic scattering and inelastic scattering. The energy deposition location and magnitude after each interaction event are recorded. Through multiple random iterations, the propagation trajectories of all simulated particles are summarized to complete the full simulation of the radiation propagation process. Finally, the complete path data of particle propagation and the energy deposition point data are output.

[0031] Then, based on the simulation results, the dose deposition value at each spatial point is calculated. The total energy deposition of all simulated paths is summarized and divided by the volume of the grid where the corresponding spatial point is located to obtain the average dose deposition value of that spatial point. Preliminary dose deposition value distribution data is generated. For example, the average dose deposition value of a certain grid point in the monitoring area, such as 10.1m, 5.1m, and 3.1m, is 0.1μGy / h. Based on this, a dose distribution matrix covering the entire monitoring area is formed, and this distribution matrix is ​​stored in a standardized file format as the initial radiation field distribution result for easy retrieval and use in subsequent steps. The distribution data of the preliminary dose deposition values ​​are then processed by gridding. Gridding discretizes the continuous monitoring area into a uniform three-dimensional grid. The continuous space is discretized by grid division. In this application, a Cartesian coordinate system is used for division. The standard grid size can be set to 1cm×1cm×1cm. In specific implementation, the resolution parameter can be adjusted according to the scene accuracy requirements. During the gridding process, a linear interpolation algorithm is first used to interpolate and fill the initial dose distribution results to ensure that each grid point has a corresponding dose value. Then, a hash mapping between the spatial points, i.e., grid point coordinates and dose values ​​is established. By verifying the consistency of the mapping, such as checking for no null values ​​and no duplicate mappings, data anomalies are avoided. This processing method can realize the rapid query and location of dose data, and at the same time adapts to the requirements of subsequent finite difference calculations for discretized data, significantly improving the overall calculation efficiency. Finally, the preliminary dose deposition values ​​after gridding are output.

[0032] By clarifying data sources, refining parameter definitions, and standardizing simulation processes and data processing steps, the accuracy, standardization, and adaptability of preliminary dose deposition values ​​were effectively improved, laying a solid data foundation for subsequent analysis of inter-particle interactions and determination of energy redistribution coefficients based on preliminary dose deposition values.

[0033] In step S1, determining the energy redistribution coefficient for each spatial point includes: using the finite difference method to numerically analyze the preliminary dose deposition value, simulating collision events of radiation particles in space, recording the location and intensity of the collision events, calculating the energy transfer between particles, and determining the initial energy redistribution coefficient for each spatial point based on the energy transfer and the total energy of the corresponding spatial point; constructing an energy redistribution model for the initial energy redistribution coefficient, combining the spatial distribution characteristics and intensity distribution patterns of the collision events, adjusting the energy distribution relationship between spatial points within the monitoring area according to the energy redistribution model; using the adjusted energy distribution relationship of the spatial points as a constraint, iteratively optimizing the initial energy redistribution coefficient using the gradient descent method, and outputting the final energy redistribution coefficient.

[0034] Specifically, after obtaining the preliminary dose deposition value after gridding, in order to accurately capture the energy transfer law between radiation particles and make up for the shortcomings of traditional methods in characterizing particle interactions, it is necessary to analyze the interaction between particles and determine the energy redistribution coefficient based on the preliminary dose deposition value, so as to provide a reliable basis for subsequent dose value correction. Specifically, the finite difference method is first used to perform numerical analysis on the preliminary dose deposition value. The finite difference method discretizes the continuous monitoring area space into grid points and uses the difference equation to approximate the particle motion and interaction process. In the process, the preliminary dose deposition value is used as the initial condition. The collision events of radiation particles in space are simulated by calculating the partial derivatives related to the particle trajectory. At the same time, the occurrence position and intensity of each collision event are recorded. The occurrence position is the grid point coordinate, such as (10.2m, 5.3m, 3.2m), and the intensity is the energy release intensity generated by the collision, such as 0.5MeV. Next, based on the recorded location and intensity of the collision event, the energy transfer between particles is calculated using the law of conservation of energy. The energy transfer is equal to the difference between the total energy of the particles before the collision and the sum of the energies of all secondary particles after the collision. A correction factor related to the medium properties at the collision location, distance, and collision intensity is introduced for adjustment. The calculation formula is as follows: ,in, This represents the amount of energy transferred between particles. This represents the total energy of the radiated particles before the collision. The sum of the energies of the n secondary particles after the collision. This represents the energy transfer attenuation coefficient of the medium at the location of the collision. The energy transfer attenuation coefficient is determined by the medium properties at that location. The straight-line distance between the collision location and the radiation source. This is the collision intensity coefficient, which is obtained by mapping the collision intensity value, and its value ranges from [0.5, 1.0]. This term is used to approximately characterize the attenuation of energy during its transmission to the collision point; for example, collision event A within the monitoring area: a Compton scattering event of a γ-ray photon with air molecules occurs at (10.2 m, 5.3 m, 3.2 m), where the medium is air. The distance from the radiation source is d=10m, and the collision intensity is 0.5MeV, corresponding to Total energy of gamma-ray photons before collision If the energy of the scattered photon after the collision is 1.2 MeV and the energy of the recoil electron is 0.3 MeV, then the amount of energy transferred is: The energy transfer between particles is essentially the net energy transferred from a radiating particle to a secondary particle or medium during a collision. It directly quantifies the scale of energy exchange in a single particle collision event and is a core physical quantity describing the energy transfer law of a radiation field. Subsequently, based on the ratio of the energy transfer amount to the total energy of the corresponding spatial point, the initial energy redistribution coefficient for each spatial point is determined. For example, if the total energy of the aforementioned spatial point is 5 MeV, then the initial energy redistribution coefficient is 0.32 ÷ 5 = 0.064. The initial energy redistribution coefficient refers to the proportion of energy transferred at that spatial point due to a particle collision event. The larger the coefficient, the higher the degree of energy redistribution at that spatial point due to interaction. For example, an initial energy redistribution coefficient of 0.064 indicates that 6.4% of the energy at that spatial point has been transferred and redistributed due to particle collisions. The energy distribution relationship between this spatial point and its surrounding spatial points can be adjusted based on this coefficient.

[0035] Subsequently, based on the initial energy redistribution coefficients, and considering the spatial distribution characteristics and intensity distribution patterns of collision events—such as the grid point density distribution of collision events within the monitoring area and the exponential decay of collision intensity with distance from the radiation source—an energy redistribution model is constructed. This model is a matrix model established by statistically analyzing the spatial distribution characteristics and intensity distribution patterns of collision events to characterize the correlation between energy redistribution coefficients at various spatial points. During construction, the weight values ​​of each spatial point are first calculated based on the spatial distribution characteristics and intensity distribution patterns of collision events. To calculate the weight values, the frequency of collision events per unit volume at each spatial point is first statistically analyzed to obtain the spatial distribution density weight. Then, the ratio of the collision intensity at each point to the average intensity of the region is used as the intensity weight. After multiplying the two and normalizing the results, the final weight value for each spatial point is obtained. Finally, a model matrix is ​​generated based on these weight values, with matrix elements reflecting the degree of mutual influence between energy redistribution coefficients at different spatial points. For example, in gamma-ray radiation scenarios, for regions with high concentrations of collision events, a weighted average method is used to assign higher weights to grid points in these regions to enhance their influence on surrounding points. In neutron radiation scenarios, Poisson distribution is incorporated into the model construction to quantify the uncertainty of collision events and improve the model's adaptability to neutron scattering characteristics. Next, the energy distribution relationship between spatial points within the monitoring area is adjusted according to the energy redistribution model. Specifically, matrix multiplication is performed between the constructed model matrix and the vector formed by the initial energy values ​​of each spatial point. The energy of each spatial point is redistributed based on the calculation results, making the energy distribution between spatial points more closely resemble the actual interaction laws of particles. For example, if the initial energy distribution in a certain region is uneven, after model matrix calculation, some energy from high-energy points is transferred to surrounding low-energy points, achieving a smooth transition in energy distribution.

[0036] Subsequently, using the adjusted spatial point energy distribution relationship as a constraint, the initial energy redistribution coefficients are iteratively optimized using the gradient descent method. Gradient descent is an algorithm that optimizes parameters by iteratively finding the minimum value of a function. In this application, the objective function of this algorithm is the sum of squared residuals between the adjusted energy distribution and the actual particle interaction energy distribution. Where m is the total number of spatial points within the monitoring area. The adjusted energy value for the j-th spatial point. The algorithm takes the theoretical energy value based on the laws of particle interaction as input data. This input includes the initial energy redistribution coefficients, the adjusted spatial point energy distribution, and preset convergence conditions. For example, the coefficient difference between two adjacent iterations must be less than a preset threshold (e.g., 0.001), and the residual must be below 0.01. The iterative update formula is as follows: , Let be the energy redistribution coefficient for the (t+1)th iteration. Let be the energy redistribution coefficient for the t-th iteration. This is the learning rate, such as a value of 0.01, used to control the step size of each iteration. Let be the partial derivative of the objective function with respect to the coefficients of the t-th iteration. During training, the initial energy redistribution coefficients are first input into the energy redistribution model. The residual between the energy distribution output by the model and the adjusted spatial point energy distribution is calculated. Then, the energy redistribution coefficients are adjusted in the direction of decreasing residuals. This process is repeated until the convergence condition is met. For example, in high-intensity radiation scenarios, the upper limit of the number of iterations is set to 50. The preliminary dose-related data obtained from the aforementioned Monte Carlo simulation is used as the starting point for iteration to accelerate the convergence speed. In low-energy radiation applications, an adaptive step-size adjustment strategy is adopted, dynamically adjusting the learning rate according to the changes in residuals. The learning rate is increased when the residual is large and decreased when the residual is small, thus reducing computation time. Finally, the final energy redistribution coefficients that meet the convergence condition are output.

[0037] By accurately quantifying the energy transfer in particle collision events and iteratively optimizing the coefficients, the systematic errors introduced by neglecting or simplifying the complex energy exchange between particles in traditional Monte Carlo simulations are fundamentally corrected. This lays the physical foundation for the accurate correction of subsequent dose values ​​and is a key step in improving the overall accuracy of radiation dose calculation.

[0038] The adjustment of the energy distribution relationship between spatial points within the monitoring area includes: setting a local domain for each spatial point; calculating the energy transfer probability between points within the local domain based on the collision event distribution characteristics in the energy redistribution model; constructing a Lagrangian dynamic system using the initial dose deposition value of each spatial point as generalized coordinates and the intensity distribution law in the energy redistribution model as the potential energy field; obtaining the dose distribution evolution trajectory by solving the constrained Euler-Lagrangian equations; extracting the generalized coordinate convergence value when the system reaches steady state from the evolution trajectory as the adjusted dose deposition value for each spatial point; and recalculating the energy gradient distribution between adjacent spatial points to obtain the adjusted energy distribution relationship.

[0039] Specifically, after completing the construction of the energy redistribution model and iterative optimization of the initial coefficients, in order to make the energy distribution between spatial points more closely conform to the actual physical laws of radiation particle interactions, it is necessary to further adjust the energy distribution relationship through dynamic system modeling, providing a precise energy distribution basis for subsequent dose field correction. Specifically, adjusting the energy distribution relationship between spatial points within the monitoring area involves first setting a local domain for each spatial point. The local domain is a spherical region centered on the target spatial point with a preset radius. Its radius is determined based on the medium properties and collision event density of the monitoring area; for example, a local domain with a radius of 0.5m can be set. Figure 2 As shown in the figure, an example of local domain distribution in a two-dimensional view is illustrated. Point P1 is the central spatial point, and the domain boundary drawn around it clearly defines the local interaction range of this point. This schematic diagram intuitively illustrates how each spatial point interacts with neighboring points through its local domain in a discretized grid, providing a spatial structural foundation for subsequent energy transfer probability calculations and the construction of Lagrange dynamics systems. Subsequently, based on the collision event distribution characteristics in the energy redistribution model, the energy transfer probability between points within the local domain is calculated. Specifically, the ratio of the collision event density between two points within the local domain to the total collision event density of the region is multiplied by the energy attenuation factor between the two points to obtain the energy transfer probability between them. The energy attenuation factor is calculated based on the straight-line distance between the two points and the medium properties. The energy transfer probability value reflects the frequency of energy interaction between different spatial points within the local domain. The higher the energy transfer probability, the more frequent the energy exchange between the two points due to particle collisions, and the stronger the correlation of energy distribution. In the subsequent construction of the Lagrange dynamics system, the higher the degree of generalized coordinate coupling between the two points, and the more significant the energy interaction between the two points during the system evolution process, which can more accurately reflect the actual physical laws of energy transfer under the interaction of radiating particles.

[0040] Next, using the initial dose deposition values ​​of each spatial point as generalized coordinates and the intensity distribution law in the energy redistribution model as the potential energy field, a Lagrange dynamic system is constructed. The Lagrange dynamic system is a dynamic model based on the Lagrange function to describe the motion state of the system. Its core is to analyze the evolution law of the system through the difference between the system's kinetic energy and potential energy. In the construction process, the initial dose deposition values ​​of each spatial point are first used as generalized coordinates to characterize the system state. The intensity distribution law in the energy redistribution model is transformed into external potential energy. The magnitude of the external potential energy is positively correlated with the collision intensity of the spatial points. At the same time, interactive potential energy based on the probability of energy transfer between points in the local domain is introduced. Interactive potential energy quantitatively describes the energy coupling effect caused by particle interaction between different spatial points: when the probability of energy transfer between two points is high, the absolute value of its interactive potential energy term increases and the sign is negative, representing an "attractive" interaction, which promotes the dose deposition values ​​to tend to be balanced between the two points, simulating the redistribution process of energy from high concentration points to low concentration points. The system's kinetic energy is defined as a function of the rate of change of the dose deposition value at each spatial point with virtual time. Its physical meaning lies in characterizing the "inertia" or "rate" of system state change, giving the system dynamic characteristics. Virtual time is an artificially introduced continuous parameter used to drive system evolution and does not directly correspond to physical time.

[0041] Combining the aforementioned external potential energy, interactive potential energy, and kinetic energy constitutes the overall Lagrangian function describing the Lagrangian dynamic system. By imposing physical constraints such as energy conservation on the system and solving the corresponding Euler-Lagrangian equations, the trajectory of dose deposition values ​​at each point evolving over virtual time can be obtained, i.e., the dose distribution evolution trajectory. Then, the generalized coordinate convergence value at which the system reaches steady state is extracted from the evolution trajectory. Steady state refers to a state where the difference between the system's kinetic and potential energy tends to stabilize and the rate of change of the generalized coordinates approaches zero. This convergence value is used as the adjusted dose deposition value for each spatial point. Based on the adjusted dose deposition value, the energy gradient distribution between adjacent spatial points is recalculated. The energy gradient distribution is the ratio of the difference in dose deposition values ​​between adjacent spatial points to the distance between the two points. This gradient distribution characterizes the changing trend of energy distribution between spatial points, ultimately yielding the adjusted energy distribution relationship.

[0042] By dividing the local action domain, calculating the energy transfer probability, and modeling the Lagrange dynamics system, the physical laws of energy interaction between spatial points are accurately characterized, providing a solid dynamic theoretical support for energy distribution adjustment. This ensures that the energy distribution state is highly consistent with the actual action law of particles, providing a high-precision energy distribution basis for subsequent correction of the three-dimensional dose concentration field, while improving the physical rationality and data reliability of energy distribution adjustment.

[0043] Step S2: Using the energy redistribution coefficient, the initial dose deposition value is adjusted point by point to obtain the corrected three-dimensional dose concentration field data. Based on the corrected three-dimensional dose concentration field data, the gradient vector between each spatial point is calculated to obtain the optimized gradient distribution map.

[0044] The step S2, which involves obtaining the corrected three-dimensional dose concentration field data, includes: adjusting the initial dose deposition value point by point according to the energy redistribution coefficient; calculating the adjusted dose value for each spatial point; if the adjusted dose value exceeds a preset threshold, marking the corresponding spatial point as a potential high-concentration point; for potential high-concentration points, recording the three-dimensional coordinate position and corresponding dose value information; integrating the adjusted dose values ​​of all spatial points within the monitoring area to generate initial three-dimensional dose concentration field data; constructing a three-dimensional grid model of concentration distribution based on the corrected three-dimensional dose concentration field data; verifying the data continuity of the three-dimensional grid model; smoothing discontinuous areas to generate the corrected three-dimensional dose concentration field data.

[0045] Specifically, after completing the iterative optimization of the energy redistribution coefficient and adjusting the spatial point energy distribution relationship, in order to accurately reflect the true dose distribution state of the radiation field, it is necessary to correct the initial dose deposition value based on the energy redistribution coefficient and construct three-dimensional dose concentration field data to provide a reliable basis for subsequent gradient analysis. Specifically, the initial dose deposition value is first adjusted point by point according to the energy redistribution coefficient. The adjusted dose value of each spatial point is calculated. Specifically, the initial dose deposition value of each spatial point is multiplied by the corresponding energy redistribution coefficient, and the adjustment results are accumulated to complete the point-level dose update. For example, in a fixed radiation source scenario, the initial dose deposition value of a spatial point corresponding to a radiation source with an energy of 10 MeV is 5 μGy / h, and the energy redistribution coefficient is 0.8. Then the adjusted dose value is 5 × 0.8 = 4 μGy / h. Subsequently, the adjusted dose value is compared with a preset threshold. If the adjusted dose value exceeds the preset threshold, it means that the radiation dose level of the spatial point is in a high-risk range, and the corresponding spatial point needs to be marked as a potential high-concentration point. Such points are key targets for radiation protection. For example, in a medical radiation simulation scenario, the preset threshold is set to 8 units. If the adjusted dose value of a spatial point is 9 units, then the point is marked as a potential high-concentration point. This marking operation can highlight high-dose radiation areas and provide clear guidance for subsequent protection planning. Next, for potential high-concentration points, their three-dimensional coordinates and corresponding dose values ​​are recorded. Specifically, the three-dimensional coordinates of potential high-concentration points are extracted, and the coordinate information is associated with the corresponding dose values ​​and stored to form a list of potential high-concentration points for subsequent tracking and analysis. Subsequently, the adjusted dose values ​​of all spatial points within the monitoring area are integrated to generate initial three-dimensional dose-concentration field data. During the integration process, all spatial points need to be traversed to collect adjusted dose values, and interpolation algorithms are applied to fill in missing data areas to ensure the integrity and continuity of the initial three-dimensional dose-concentration field data.

[0046] Based on the initial three-dimensional dose concentration field data, a three-dimensional mesh model of the concentration distribution is constructed. The three-dimensional mesh model is a model that uses the voxel mesh method to divide the three-dimensional space into uniform cubic units and maps the concentration field data onto a regular mesh. In the construction process, the mesh resolution parameter is first defined, for example, 100 units per side. Then, the three-dimensional space of the monitoring area is divided into uniform mesh units according to the set resolution. Then, the statistical mean of the adjusted dose values ​​of all spatial points in each mesh unit is calculated and the mean is assigned to the corresponding mesh unit. For example, if the concentration values ​​of spatial points in a certain mesh unit are 6, 7, and 8 units respectively, and the unit is the radiation dose measurement reference unit microgray per hour (μGy / h), then the concentration value of the mesh unit is 7 units. Finally, all mesh units are assembled to form a complete three-dimensional mesh model. Next, the continuity of the 3D mesh model is verified. During verification, the concentration difference between adjacent mesh cells is calculated and it is determined whether the concentration difference is less than a preset continuity threshold, for example, 0.5 units. If the concentration difference between adjacent mesh cells exceeds the preset continuity threshold, the discontinuous region is smoothed by applying a Gaussian filtering algorithm. Gaussian filtering is a smoothing algorithm based on weighted averaging. It can eliminate data abrupt changes by weighted averaging of the concentration values ​​in discontinuous regions. For example, adjacent cells with concentration values ​​of 7 and 9 are adjusted to 8 after Gaussian filtering smoothing, so that the concentration distribution of the mesh model is more in line with the actual propagation law of the radiation field. Finally, the 3D mesh model data after the data continuity verification and smoothing are determined as the corrected 3D dose concentration field data.

[0047] By adjusting the energy redistribution coefficient point by point, marking and recording potential high-concentration points, integrating data, and constructing and verifying a three-dimensional mesh model, a precise conversion from the initial dose deposition value to the corrected three-dimensional dose concentration field data was achieved. This provides high-precision data support for subsequent gradient vector calculation and gradient distribution map generation, improving the accuracy and reliability of radiation field dose distribution calculation. The generated three-dimensional dose concentration field data can truly reflect the spatial distribution characteristics of the radiation field. In particular, the physical correction of the initial dose deposition value by the energy redistribution coefficient significantly reduces the dose estimation deviation caused by simulation statistical fluctuations and medium inhomogeneity, directly improving the calculation accuracy of radiation dose in spatial distribution.

[0048] In step S2, obtaining the optimized gradient distribution map includes: calculating the concentration difference between adjacent spatial points based on the corrected three-dimensional dose-concentration field data; determining the gradient vector between adjacent spatial points by using the concentration difference as a component and combining it with the spacing between adjacent spatial points; calculating the directional offset angle between each gradient vector and the average gradient vector of the region, and determining whether the directional offset angle exceeds a preset range; if it exceeds the preset range, smoothing the gradient vector; constructing an initial gradient distribution map based on the smoothed gradient vector; verifying the spatial continuity and directional consistency of the gradient vectors in the initial gradient distribution map, and generating the optimized gradient distribution map.

[0049] Specifically, after completing the construction and verification of the corrected three-dimensional dose concentration field data, in order to intuitively present the spatial variation trend of radiation dose, it is necessary to calculate the gradient vector and construct the gradient distribution map based on the concentration field data, so as to provide accurate trend basis for radiation field evolution analysis. Specifically, firstly, based on the corrected three-dimensional dose-concentration field data, the concentration difference between adjacent spatial points is calculated. Using the concentration difference as a component, combined with the distance between adjacent spatial points, the gradient vector between adjacent spatial points is determined. The gradient vector is a vector that characterizes the rate of change and direction of change of physical quantities in space. Its magnitude reflects the intensity of concentration change, and its direction points to the direction of the fastest increase in concentration. The gradient vector is calculated as the ratio of the concentration difference between adjacent spatial points to the distance between the two points. For example, if the coordinates of adjacent spatial point A are (10m, 5m, 3m) and the concentration value is 5μGy / h, and the coordinates of spatial point B are (10.1m, 5m, 3m) and the concentration value is 8μGy / h, and the distance between the two points is 0.1m, then the magnitude of the gradient vector between the two points is (8-5)÷0.1=30μGy / (h・m), and the direction is from A to B. Next, the directional offset angle between each gradient vector and the regional average gradient vector is calculated. The regional average gradient vector is obtained by vector averaging all gradient vectors within the monitoring area, reflecting the overall concentration change trend of the area. The directional offset angle is the angle between a single gradient vector and the regional average gradient vector. Then, it is determined whether the directional offset angle exceeds a preset range, which can be set according to the monitoring scenario requirements, for example, ±30°. If the directional offset angle exceeds the preset range, it indicates that the concentration change trend corresponding to that gradient vector deviates significantly from the overall regional trend, requiring smoothing of the gradient vector. The smoothing process uses a weighted average method, that is, it is weighted and corrected based on the direction and magnitude of the surrounding gradient vectors, making the corrected gradient vector direction more closely match the overall regional change trend. Subsequently, based on the smoothed gradient vectors, an initial gradient distribution map is constructed. The initial gradient distribution map is a visual graphic formed by labeling all gradient vectors according to their spatial positions in a three-dimensional coordinate system, which can intuitively show the direction and intensity of concentration changes in each area. Finally, the initial gradient distribution map is used to verify the spatial continuity and directional consistency of the gradient vectors. Spatial continuity verification checks whether the difference in magnitude between adjacent gradient vectors is less than a preset difference threshold. Directional consistency verification checks whether the angle between the directions of adjacent gradient vectors is less than a preset angle threshold. If the verification conditions are not met, the gradient vectors in the corresponding region are smoothed again. After verification and adjustment, an optimized gradient distribution map is generated.

[0050] By accurately calculating gradient vectors, correcting deviations, and verifying and optimizing distribution maps, a visual presentation of the spatial variation trend of radiation dose is achieved. This provides a high-precision trend basis for the analysis of radiation field evolution, improves the reliability and rationality of gradient distribution maps, and enables radiation monitoring results to more accurately guide the formulation of radiation protection plans, thus ensuring the scientific nature and effectiveness of radiation environment monitoring.

[0051] Step S3: Extract the set of boundary points whose gradient changes exceed a preset threshold from the optimized gradient distribution map, construct an isosurface grid based on the boundary point set to form a three-dimensional isosurface structure, obtain time series radiation source data, iteratively update the energy redistribution coefficient, and obtain the temporal evolution sequence of the concentration field.

[0052] Step S3, forming a three-dimensional isosurface structure, includes: identifying and extracting spatial points whose gradient magnitude exceeds a preset threshold as boundary point sets based on the optimized gradient distribution map; calculating the connection relationships between boundary point sets using an interpolation algorithm; constructing triangular mesh surfaces through the connection relationships to generate an isosurface mesh; obtaining a preliminary three-dimensional isosurface structure based on the isosurface mesh; verifying the topological integrity of the mesh based on the preliminary three-dimensional isosurface structure; and optimizing and adjusting the isosurface mesh to generate the final three-dimensional isosurface structure data.

[0053] Specifically, after constructing and validating the optimized gradient distribution map, to intuitively present the spatial boundary characteristics of radiation dose concentration, it is necessary to extract the boundary point set based on the gradient distribution map and construct a three-dimensional isosurface structure. Simultaneously, time-series data should be combined to achieve temporal evolution analysis of the concentration field. Specifically, firstly, based on the optimized gradient distribution map, spatial points with gradient amplitudes exceeding a preset threshold are identified and extracted as the boundary point set. The gradient amplitude is a scalar value characterizing the drastic change in the gradient vector, obtained by solving the vector difference between adjacent points using the finite difference method. The preset threshold can be set according to the monitoring scenario requirements, for example, 30 μGy / (h・m). All spatial points in the gradient distribution map are traversed, and the gradient amplitude of each point is calculated. Spatial points with amplitudes exceeding the preset threshold are grouped into the boundary point set. These boundary points correspond to key regions of radiation dose concentration change and are the core foundation for constructing the isosurface. Next, for the boundary point set, an interpolation algorithm is used to calculate the connection relationship between the boundary points. The bilinear interpolation algorithm used in this application is a method of inferring unknown point values ​​from four known point values ​​on a two-dimensional plane. In three-dimensional expansion, it can be applied to spatial meshes to calculate the intermediate value between each pair of points to determine the connection strength. During the calculation, a distance matrix between point pairs is first constructed based on the position coordinates in the point set. Then, the bilinear interpolation algorithm is applied to fill in the missing connection values ​​in the distance matrix. For example, in a region with high radiation source energy, when the distance between point pairs is less than 0.5 units, the interpolated connection strength is 1.2 times the original strength. Subsequently, the connection relationship is iteratively updated until the interpolation error of all point pairs is less than a preset threshold, such as 0.01, to ensure the continuity of the connection relationship. Then, triangular mesh surfaces are constructed based on the connection relationship to generate an isosurface mesh. Specifically, point pairs with connection strength greater than the threshold are formed into mesh edges, and then triangular mesh surfaces are constructed and filled using the mesh edges to cover the entire boundary point set region. The triangular mesh surface is the basic unit for constructing the three-dimensional isosurface, which can ensure the stability and smoothness of the isosurface structure. Figure 3As shown in the figure, a two-dimensional projection diagram of a triangular mesh surface is presented. The X and Y axes represent spatial coordinates, respectively. The mesh is composed of multiple adjacent triangular units connected together. The vertices of each triangle are labeled V1, V2, V3, etc., corresponding to spatial points in the extracted boundary point set. This diagram visually demonstrates how to construct a continuous triangular mesh surface based on the connection relationships between boundary points, thus laying the geometric foundation for the subsequent generation of a smooth, closed three-dimensional isosurface structure. Next, based on the isosurface mesh, all triangular mesh surfaces are superimposed to form a closed three-dimensional surface structure, resulting in a preliminary three-dimensional isosurface structure. This structure visually represents the isosurface boundaries of radiation dose concentration, clearly delineating the regions of different concentration levels. Next, for the preliminary 3D isosurface structure, the topological integrity of the mesh is verified. Topological integrity verification is achieved by calculating the Euler characteristic number, which is a parameter characterizing the topological features of the 3D structure. The calculation formula is the number of vertices minus the number of edges plus the number of faces. If the result is 2, it indicates a complete spherical topology. During verification, the connectivity of the mesh faces is checked to ensure that there are no isolated faces. If a break is found, the problem area is marked and the integrity score is recorded. For example, if the score is lower than a preset threshold, such as 0.9, an alarm is triggered. In the static radiation field scene, the constructed 3D isosurface structure has 200 mesh faces, and verifying that the Euler characteristic number is 2 confirms the integrity. Finally, the isosurface mesh was optimized and adjusted. For the problem areas marked in the verification, a smoothing algorithm was applied to adjust the mesh side length by averaging adjacent side lengths, eliminating structural fractures and irregular areas, and generating the final three-dimensional isosurface structure data. At the same time, time-series radiation source data was acquired, and the radiation source parameters at different time points were input into the energy redistribution coefficient calculation process to iteratively update the energy redistribution coefficient. The initial dose deposition value was corrected by combining the coefficients after each update, thereby obtaining the temporal evolution sequence of the concentration field. The temporal evolution sequence of the concentration field represents a continuous data sequence of the spatial distribution and dynamic change process of radiation dose concentration in the monitoring area at different time points. It can intuitively reflect the diffusion, migration, attenuation or enhancement of the radiation concentration field over time, and clearly present the spatiotemporal trajectory of potential high concentration points and the dynamic evolution characteristics of the isosurface structure.

[0054] By accurately extracting boundary point sets, scientifically calculating connectivity relationships, and constructing and verifying three-dimensional isosurface structures, the spatial boundary characteristics of radiation dose concentration are presented intuitively. Combined with time series data, dynamic evolution analysis of the concentration field is realized, providing high-precision structural support and data basis for the study of the spatiotemporal variation law of the radiation field, and improving the visualization and analytical depth of radiation monitoring results.

[0055] In step S3, obtaining the temporal evolution sequence of the concentration field includes: acquiring time-series radiation source data corresponding to different timestamps from a preset database; simulating the dynamic change process of the radiation field over time within the monitoring area based on the time-series radiation source data and a three-dimensional isosurface structure; iteratively updating the energy redistribution coefficient through change simulation; adjusting the concentration field distribution based on the updated energy redistribution coefficient to generate an initial concentration field temporal evolution sequence; verifying the temporal continuity of the data for the initial concentration field temporal evolution sequence, and smoothing the initial concentration field temporal evolution sequence to generate the final concentration field temporal evolution sequence.

[0056] Specifically, after completing the construction of the precise three-dimensional isosurface structure, in order to accurately capture the dynamic changes of the radiation field, it is necessary to iteratively update the energy redistribution coefficient by combining radiation source data in the time dimension, thereby generating a temporal evolution sequence of the concentration field, providing data support for the dynamic monitoring and trend prediction of the radiation field. Specifically, time-series radiation source data corresponding to different timestamps are obtained from a dedicated radiation environment monitoring database. The radiation source data obtained in step S1 is instantaneous static data at a single time node, which can only meet the static calculation requirements of the initial radiation field distribution and cannot reflect the dynamic changes of the radiation source within the monitoring period. To realize the evolution analysis of the radiation field over time, it is necessary to rely on continuous time-series radiation source data with multiple timestamps to fully characterize the fluctuation, migration, or decay process of parameters such as the location, energy, and intensity of the radiation source over time. This is a necessary prerequisite for carrying out dynamic simulation of the radiation field and iteratively updating the energy redistribution coefficient. The database will retain radiation source state data at preset event intervals, such as 10 minutes, starting from the initial moment. The acquisition process involves extracting the above-mentioned full time-series data through the database query interface as the basic input for subsequent dynamic simulation. Next, based on time-series radiation source data and combined with a three-dimensional isosurface structure, the dynamic change of the radiation field within the monitoring area over time is simulated. Here, the Monte Carlo simulation method is used to extend the time dimension. The Monte Carlo simulation method is a probabilistic simulation algorithm based on random sampling. The three-dimensional isosurface structure is used to define the spatial boundary of radiation propagation. During the simulation, the radiation source parameters at each time point are first input into the simulation framework in sequence, and then the propagation trajectory of the radiating particles, their collision interaction with the medium, and the energy decay process are calculated at each time point. For example, in a radiation leakage scenario, the intensity distribution change of the radiation field from the initial moment of leakage to the subsequent diffusion stage can be accurately simulated. Finally, the dynamic distribution data of the radiation field evolution over time is output to support the iterative update of the energy redistribution coefficient. For different radiation types, the simulation parameters can be flexibly adjusted. For example, in the gamma-ray scenario, the focus is on matching the energy decay rate parameter, while in the neutron radiation scenario, the focus is on optimizing the particle capture cross-section parameter to ensure that the simulation results conform to the actual physical laws of different application scenarios.

[0057] Subsequently, the energy redistribution coefficient is iteratively updated through the aforementioned simulation of changes. This iterative update process is based on an extension of the finite difference method. The core logic is to extract the initial dose deposition value at each time point from the dynamic change data of the radiation field obtained from the simulation, and then analyze the interaction law between particles to determine the location and intensity of collision events. Based on this, the initial energy redistribution coefficient corresponding to each time point is calculated. Then, the coefficient is continuously corrected through iterative loops. In each iteration, the radiation field simulation result corresponding to the current coefficient is compared with the result of the previous iteration. If the difference value is less than the preset convergence threshold, such as 0.01, the iteration stops. The number of iterations is usually set to 10-50 times. The coefficient will gradually converge from the initial value, such as 0.5, to a stable value, such as 0.8. In high-energy radiation source scenarios, the iterative process also introduces an energy attenuation factor and corrects the coefficient by combining the fluctuation characteristics of the radiation source intensity in the time series data. In low-intensity radiation diffusion scenarios, the boundary effect is focused on, and the coefficient is adjusted by combining the three-dimensional isosurface structure to ensure that the final energy redistribution coefficient can accurately reflect the energy transfer trend in the time dimension.

[0058] Next, based on the updated energy redistribution coefficients, the concentration field distribution is adjusted to generate an initial concentration field time-series evolution sequence. The adjustment method involves multiplying the initial concentration field data at each time point by the corresponding updated energy redistribution coefficient point by point to obtain the adjusted concentration field distribution at each time point. Then, the adjusted concentration field data at all time points are arranged in timestamp order to form a continuous initial concentration field time-series evolution sequence. Afterwards, the temporal continuity of the initial concentration field time-series evolution sequence is verified. Temporal continuity verification is a check process to ensure a smooth transition of the time-series data. Specifically, the concentration difference between corresponding spatial points at adjacent time points in the sequence is calculated, and it is determined whether this difference is less than a preset continuity threshold, such as 0.5 μGy / h. If the difference exceeds the threshold, it is determined that there is a temporal dimension jump anomaly, and the abnormal region needs to be marked for subsequent processing. Finally, the initial concentration field time-series evolution sequence is smoothed to generate the final concentration field time-series evolution sequence. This is achieved using a Gaussian smoothing filter, a noise reduction algorithm based on weighted averaging. Its core principle is to calculate a weighted average of the concentration value at each time point in the sequence, combined with the concentration values ​​of multiple adjacent time steps, and use this average to replace the original value for smoothing. The smoothing strategy can be adjusted for different radiation change scenarios. For example, in rapidly changing radiation leakage scenarios, a layered smoothing approach is used: first, coarse smoothing removes significant noise, and then fine smoothing preserves the details of concentration changes. In slowly decaying radiation scenarios, the filter radius is adjusted to 10 time steps to highlight the long-term evolution trend. During the smoothing process, gradient vector optimization results are simultaneously incorporated to ensure that potential high-concentration point markers are not lost while eliminating noise, ultimately generating a smooth and accurate final concentration field time-series evolution sequence.

[0059] Through a series of operations, including time-series radiation source data acquisition, dynamic simulation of the radiation field, iterative updating of coefficients, and optimization of the time-series sequence, the accurate generation of the concentration field time-series evolution sequence was achieved, providing high-precision data support for the analysis of the dynamic changes in the radiation field.

[0060] Step S4: Based on the temporal evolution sequence of the concentration field, identify and mark high-concentration regions, fuse the high-concentration regions and the three-dimensional isosurface structure, and generate a visual distribution model reflecting the dose concentration based on the fusion result.

[0061] In step S4, generating a visual distribution model reflecting dose concentration based on the fusion results includes: extracting dynamic change information of high-concentration regions based on the temporal evolution sequence of the concentration field, and marking key high-concentration regions in a three-dimensional isosurface structure by combining high-concentration region markers; fusing the key high-concentration regions and the temporal evolution sequence using a weighted average method to construct a preliminary visual distribution model; generating a three-dimensional dynamic dose distribution map based on the preliminary visual distribution model; calculating the rendering coverage of the three-dimensional dynamic dose distribution map, verifying the visualization integrity of the data, and outputting the final visual distribution model reflecting dose concentration.

[0062] Specifically, after generating the temporal evolution sequence of the concentration field, to intuitively present the spatiotemporal distribution of radiation dose concentration and the dynamics of high-concentration areas, it is necessary to integrate high-concentration area markers with three-dimensional isosurface structures. Through data fusion and visualization modeling, an accurate visualized distribution model is output, providing intuitive support for subsequent radiation field analysis and decision-making. Specifically, firstly, based on the temporal evolution sequence of the concentration field, dynamic change information of high-concentration areas is extracted. High-concentration areas are spatial regions where the concentration value exceeds a preset threshold. Dynamic change information includes the location shift, range scaling, and concentration intensity fluctuations of the area at different timestamps. The extraction process involves selecting data from consecutive time points in the temporal evolution sequence and comparing the three-dimensional coordinate range of the high-concentration area with the corresponding concentration value point by point, thereby identifying the dynamic evolution pattern of the area over time. Next, combined with the high-concentration area markers, key high-concentration areas are marked in the three-dimensional isosurface structure. The high-concentration area markers are the identification information of spatial points with concentrations exceeding the threshold in the previous steps. The marking process involves matching the three-dimensional coordinates of the marker points with the spatial position of the three-dimensional isosurface structure to define the isosurface area corresponding to the high-concentration area as the key area and give it a unique mark, so that the key area can be clearly identified in the three-dimensional isosurface structure, thus clarifying the core objects for subsequent data fusion.

[0063] Subsequently, a weighted average method was used to fuse the high-concentration key areas and the time-series evolution sequence. During the fusion calculation, the coordinates of each point in the key area were first precisely matched with the concentration values ​​in the time-series evolution sequence. Then, the time dimension fusion concentration value of each point was calculated according to the set weights, thereby obtaining the overall fusion data of the key area. The weighting rules are set according to the actual scenario. For example, for a fixed radiation source scenario, weights are assigned according to the time distance, with more recent time points having higher weights to highlight the latest concentration dynamics. For a moving radiation source scenario, the weights are adjusted based on the displacement distance of the center point of the key area, with larger displacement distances resulting in lower weights to weaken the impact of concentration instability caused by positional shifts. Based on this fused data, a preliminary visualization distribution model is constructed. The construction process first organizes the fused data into a data matrix, where the rows of the matrix correspond to spatial points within the key area, the columns correspond to timestamps of the temporal evolution sequence, and the matrix elements are the fused concentration values ​​of each point at the corresponding timestamp. Then, the data matrix is ​​standardized so that the mean of each column is 0 and the variance is 1. Subsequently, principal component analysis is used to extract core features. Principal component analysis calculates the covariance matrix of the data matrix, solves for its eigenvalues ​​and eigenvectors, and selects the top few eigenvectors with a cumulative variance contribution rate exceeding a preset threshold, such as 85%, as the main features. This achieves data dimensionality reduction while retaining core spatiotemporal change information. Finally, a preliminary visualization distribution model is constructed based on the core features. This model can accurately focus on the temporal change trend of high-concentration areas.

[0064] Next, based on the preliminary visualization distribution model, a three-dimensional dynamic dose distribution map is generated. The generation process maps the core feature vectors in the model to a three-dimensional spatial coordinate system, rendering the high-concentration key areas and the overall concentration distribution at each time point in time stamp order, forming a dynamic visualization image that includes a time dimension, intuitively showing the evolution of the radiation concentration field over time. Then, for the three-dimensional dynamic dose distribution map, the rendering coverage is calculated to verify the visualization integrity of the data. Rendering coverage refers to the ratio of the number of successfully drawn spatial points in the three-dimensional dynamic map to the total number of points in the model; it is a core indicator for measuring visualization integrity. During verification, the number of drawn points in the distribution map is first counted against the total number of points to calculate the rendering coverage. If the coverage is lower than a preset threshold, such as 95%, it is considered incomplete and needs to be returned to the data fusion stage for reprocessing. Simultaneously, it is checked whether all points in the high-concentration key areas of the distribution map cover the elements in the data matrix. If any are missing, the missing areas are marked and rendered to ensure that no dynamic changes in high-concentration areas are missed. Finally, by adjusting the visualization parameters, the final visualization distribution model reflecting the dose concentration is output. The visualization parameters include color threshold, transparency, rendering accuracy, etc. When adjusting, they can be optimized according to the type of radiation source. For example, for neutron sources and gamma sources, different color threshold ranges can be set to distinguish energy differences, making the concentration distribution characteristics of different types of radiation clearer. By optimizing the transparency parameter, the key areas of high concentration are highlighted, improving the visual recognition of the model. The final output visualization distribution model can be directly used for subsequent radiation field analysis, protection scheme formulation, and other work.

[0065] By dynamically extracting high-concentration areas, data fusion modeling, and visual verification optimization, a direct presentation of the spatiotemporal distribution of radiation dose concentration is achieved. This precisely focuses on the core information of dynamic changes in high-concentration areas, providing intuitive and accurate visualization support for radiation field analysis. It enhances the readability and application value of radiation monitoring results. The resulting visualized distribution model integrates the accuracy improvements achieved in energy redistribution, spatial gradient optimization, and temporal evolution from the aforementioned steps. By fusing high-precision dose data with three-dimensional geometric structures, the model ensures the numerical accuracy and spatial realism of the presented dose concentration distribution, making radiation safety assessments and protection decisions based on this model more scientifically sound and reliable.

[0066] To verify the technical effects of the present invention, Figure 4The comparison results of traditional methods and the final visualized distribution model of this application in key performance indicators are shown. This application achieves significant improvements in four core performance indicators: rendering coverage, accuracy error, real-time performance, and radiation dose calculation accuracy. Rendering coverage increases from 80.0% of the traditional method to 97.4%, an improvement of 22%; accuracy error decreases from 13.0% to 3.3%, an improvement of 75%; real-time performance increases from 8.0 FPS to 22.0 FPS, an improvement of 175.0%; and radiation dose calculation accuracy increases from 85.0% to 96.2%, an improvement of 13%. These data demonstrate that this application, through a series of techniques such as iterative optimization of energy redistribution coefficients, gradient distribution map construction, three-dimensional isosurface structure, and time-series evolution analysis, significantly improves the accuracy, completeness, and reliability of radiation dose calculation while maintaining high real-time performance, providing a more accurate, complete, and real-time visualization solution for radiation environment monitoring.

[0067] The dose calculation method based on radiation environment monitoring data in the embodiments of this application has been described above. The dose calculation system based on radiation environment monitoring data in the embodiments of this application is described below. Please refer to [link / reference]. Figure 5 One embodiment of the dose calculation system based on radiation environment monitoring data in this application includes:

[0068] The acquisition module is used to acquire radiation source data, calculate the initial radiation field distribution based on the radiation source data, obtain the preliminary dose deposition value at each spatial point in the monitoring area, analyze the inter-particle interaction reflected by the preliminary dose deposition value, and determine the energy redistribution coefficient corresponding to each spatial point.

[0069] The optimization module is used to adjust the initial dose deposition value point by point using the energy redistribution coefficient to obtain the corrected three-dimensional dose concentration field data. Based on the corrected three-dimensional dose concentration field data, the gradient vector between each spatial point is calculated to obtain the optimized gradient distribution map.

[0070] The iterative update module is used to extract the set of boundary points whose gradient changes exceed a preset threshold from the optimized gradient distribution map, construct an isosurface grid based on the boundary point set to form a three-dimensional isosurface structure, acquire time series radiation source data, iteratively update the energy redistribution coefficient, and obtain the temporal evolution sequence of the concentration field.

[0071] The visualization module is used to identify and label high-concentration regions based on the temporal evolution sequence of the concentration field, fuse the high-concentration regions and the three-dimensional isosurface structure, and generate a visualized distribution model reflecting the dose concentration based on the fusion result.

[0072] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the systems and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0073] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0074] The above-described embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application.

Claims

1. A dose calculation method based on radiation environment monitoring data, characterized in that, The method includes: Step S1: Obtain radiation source data, calculate the initial radiation field distribution based on the radiation source data, obtain the preliminary dose deposition value at each spatial point in the monitoring area, analyze the inter-particle interaction reflected by the preliminary dose deposition value, and determine the energy redistribution coefficient corresponding to each spatial point; The step S1, determining the energy redistribution coefficient for each spatial point, includes: using the finite difference method to numerically analyze the preliminary dose deposition value, simulating collision events of radiation particles in space, recording the location and intensity of the collision events, calculating the energy transfer between particles, and determining the initial energy redistribution coefficient for each spatial point based on the energy transfer and the total energy of the corresponding spatial point; constructing an energy redistribution model for the initial energy redistribution coefficient, combining the spatial distribution characteristics and intensity distribution patterns of the collision events, adjusting the energy distribution relationship between spatial points within the monitoring area according to the energy redistribution model; using the adjusted energy distribution relationship of the spatial points as a constraint, iteratively optimizing the initial energy redistribution coefficient using the gradient descent method, and outputting the final energy redistribution coefficient. The adjustment of the energy distribution relationship between spatial points within the monitoring area includes: setting a local domain for each spatial point; calculating the energy transfer probability between points within the local domain based on the collision event distribution characteristics in the energy redistribution model; constructing a Lagrangian dynamic system using the initial dose deposition value of each spatial point as generalized coordinates and the intensity distribution law in the energy redistribution model as the potential energy field; obtaining the dose distribution evolution trajectory by solving the constrained Euler-Lagrangian equations; extracting the generalized coordinate convergence value when the system reaches steady state from the evolution trajectory as the adjusted dose deposition value for each spatial point; and recalculating the energy gradient distribution between adjacent spatial points to obtain the adjusted energy distribution relationship. Step S2: Using the energy redistribution coefficient, the initial dose deposition value is adjusted point by point to obtain the corrected three-dimensional dose concentration field data. Based on the corrected three-dimensional dose concentration field data, the gradient vector between each spatial point is calculated to obtain the optimized gradient distribution map. The step S2, obtaining the corrected three-dimensional dose concentration field data, includes: adjusting the initial dose deposition value point by point according to the energy redistribution coefficient, calculating the adjusted dose value at each spatial point, and marking the corresponding spatial point as a potential high-concentration point if the adjusted dose value exceeds a preset threshold; recording the three-dimensional coordinate position and corresponding dose value information for the potential high-concentration point, and generating initial three-dimensional dose concentration field data by integrating the adjusted dose values ​​of all spatial points within the monitoring area; constructing a three-dimensional grid model of concentration distribution based on the corrected three-dimensional dose concentration field data, verifying the data continuity of the three-dimensional grid model, and smoothing discontinuous areas to generate the corrected three-dimensional dose concentration field data. Step S3: Extract the set of boundary points whose gradient changes exceed a preset threshold from the optimized gradient distribution map, construct an isosurface grid based on the boundary point set to form a three-dimensional isosurface structure, obtain time series radiation source data, iteratively update the energy redistribution coefficient, and obtain the temporal evolution sequence of the concentration field. Step S4: Based on the temporal evolution sequence of the concentration field, identify and mark high-concentration regions, fuse the high-concentration regions and the three-dimensional isosurface structure, and generate a visual distribution model reflecting the dose concentration based on the fusion result.

2. The dose calculation method based on radiation environment monitoring data according to claim 1, characterized in that, Step S1 obtains preliminary dose deposition values ​​at each spatial point within the monitoring area, including: Radiation source data, including the location, energy, and type of the radiation source, is extracted from a preset database. The radiation source data is processed using the Monte Carlo simulation method to simulate the propagation path and energy deposition process of particles in the monitoring area. Based on the simulation results, the dose deposition value at each spatial point is calculated, and the distribution data of the preliminary dose deposition value is generated. The distribution data of the preliminary dose deposition value is then gridded to construct the mapping relationship between spatial points and dose values, and the gridded preliminary dose deposition value is output.

3. The dose calculation method based on radiation environment monitoring data according to claim 1, characterized in that, Step S2 yields the optimized gradient distribution map, including: Based on the corrected three-dimensional dose-concentration field data, the concentration difference between adjacent spatial points is calculated. Using the concentration difference as a component and the spacing between adjacent spatial points, the gradient vector between adjacent spatial points is determined. The directional offset angle between each gradient vector and the regional average gradient vector is calculated, and it is determined whether the directional offset angle exceeds a preset range. If it exceeds the preset range, the gradient vector is smoothed. Based on the smoothed gradient vector, an initial gradient distribution map is constructed. The spatial continuity and directional consistency of the gradient vector are verified on the initial gradient distribution map, and an optimized gradient distribution map is generated.

4. The dose calculation method based on radiation environment monitoring data according to claim 1, characterized in that, Step S3 involves forming a three-dimensional isosurface structure, including: Based on the optimized gradient distribution map, spatial points with gradient amplitudes exceeding a preset threshold are identified and extracted as boundary point sets. For the boundary point sets, an interpolation algorithm is used to calculate the connection relationships between the boundary points. Through the connection relationships, triangular mesh surfaces are constructed to generate isosurface meshes. Based on the isosurface meshes, a preliminary three-dimensional isosurface structure is obtained. For the preliminary three-dimensional isosurface structure, the topological integrity of the mesh is verified. The isosurface meshes are optimized and adjusted to generate the final three-dimensional isosurface structure data.

5. The dose calculation method based on radiation environment monitoring data according to claim 4, characterized in that, Step S3 yields the temporal evolution sequence of the concentration field, including: Time-series radiation source data corresponding to different timestamps are obtained from a preset database. Based on the time-series radiation source data and the three-dimensional isosurface structure, the dynamic change process of the radiation field in the monitoring area over time is simulated. Through change simulation, the energy redistribution coefficient is iteratively updated. Based on the updated energy redistribution coefficient, the concentration field distribution is adjusted to generate an initial concentration field time-series evolution sequence. For the initial concentration field time-series evolution sequence, the temporal continuity of the data is verified, and the initial concentration field time-series evolution sequence is smoothed to generate the final concentration field time-series evolution sequence.

6. The dose calculation method based on radiation environment monitoring data according to claim 1, characterized in that, Step S4 generates a visual distribution model reflecting the dose concentration based on the fusion results, including: Based on the temporal evolution sequence of the concentration field, dynamic change information of high-concentration regions is extracted. Combined with the high-concentration region markers, key high-concentration regions are marked in the three-dimensional isosurface structure. Data fusion of the key high-concentration regions and the temporal evolution sequence is performed using a weighted average method to construct a preliminary visualization distribution model. Based on the preliminary visualization distribution model, a three-dimensional dynamic dose distribution map is generated. For the three-dimensional dynamic dose distribution map, the rendering coverage is calculated to verify the visualization integrity of the data, and the final visualization distribution model reflecting the dose concentration is output.

7. A dose calculation system based on radiation environment monitoring data, used to implement the dose calculation method based on radiation environment monitoring data as described in any one of claims 1-6, characterized in that, The system includes: The acquisition module is used to acquire radiation source data, calculate the initial radiation field distribution based on the radiation source data, obtain the preliminary dose deposition value at each spatial point in the monitoring area, analyze the inter-particle interaction reflected by the preliminary dose deposition value, and determine the energy redistribution coefficient corresponding to each spatial point. The optimization module is used to adjust the initial dose deposition value point by point using the energy redistribution coefficient to obtain the corrected three-dimensional dose concentration field data. Based on the corrected three-dimensional dose concentration field data, the gradient vector between each spatial point is calculated to obtain the optimized gradient distribution map. The iterative update module is used to extract the set of boundary points whose gradient changes exceed a preset threshold from the optimized gradient distribution map, construct an isosurface grid based on the boundary point set to form a three-dimensional isosurface structure, acquire time-series radiation source data, iteratively update the energy redistribution coefficient, and obtain the temporal evolution sequence of the concentration field. The visualization module is used to identify and mark high-concentration regions based on the temporal evolution sequence of the concentration field, fuse the high-concentration regions and the three-dimensional isosurface structure, and generate a visualization distribution model reflecting the dose concentration based on the fusion result.

Images