A method, system and device for calculating the lead equivalent of a radiation protection device

Abstract

The application discloses a kind of lead equivalent calculation methods, systems and equipment of radiation protection equipment, it is related to ionizing radiation protection technical field.The method comprises: obtaining the historical data of lead equivalent measurement of radiation protection equipment sample;Based on LSTM neural network, a deep learning model is constructed;Based on the historical data of lead equivalent greater than 0.5 and the historical data of lead equivalent less than or equal to 0.5, the deep learning model is trained respectively, to obtain two lead equivalent calculation models;Standard radiation field and monitoring equipment are used to measure the radiation transmittance of the sample to be measured;According to the radiation transmittance of the sample to be measured, one of the two lead equivalent calculation models is selected as the current calculation model to calculate the lead equivalent of the sample to be measured.The lead equivalent calculation method, system and equipment of radiation protection equipment disclosed in the application can effectively improve the calculation efficiency, precision and stability of batch sample lead equivalent, save labor cost and time cost.

Description

Technical Field

[0001] This invention relates to the field of ionizing radiation protection technology, and in particular to a method, system and equipment for calculating the lead equivalent of radiation protection equipment. Background Technology

[0002] In practice, especially during medical irradiation, radiation protection equipment is typically required for shielding. To ensure the quality of radiation protection equipment and safeguard the health and safety of radiation workers and the public, the lead equivalent of a sample of the equipment must be determined before use.

[0003] In a standard radiation field, there is a correlation between the radiation transmittance of a sample and its lead equivalent. Theoretically, the radiation transmittance of each radiation protection device sample to a specific X-ray material can be measured and calculated in a standard radiation field. By using different combinations of standard lead sheets to find the same radiation transmittance, the lead equivalent of that combination of standard lead sheets is the lead equivalent of the sample. However, this method is too cumbersome, with high labor and time costs, making it difficult to widely apply in practice. Furthermore, because radiation transmittance is affected by various factors such as sample properties, radiation field characteristics, and environmental conditions, it is difficult to establish a unified and accurate calculation formula between radiation transmittance and lead equivalent through physical definitions. Therefore, there is an urgent need to achieve automated and intelligent calculation of lead equivalent in practical applications. Summary of the Invention

[0004] The purpose of this invention is to provide a method, system, and equipment for calculating the lead equivalent of radiation protection equipment, so as to improve the calculation efficiency, accuracy, and stability of lead equivalent of batch samples, and save labor and time costs.

[0005] To achieve the above objectives, the present invention provides the following solution:

[0006] A method for calculating the lead equivalent of radiation protection equipment, comprising:

[0007] Obtain historical data on lead equivalent measurements of radiation protection equipment samples; the historical data includes the sample's radiation transmittance and the corresponding lead equivalent.

[0008] Building deep learning models based on LSTM neural networks;

[0009] Two lead equivalent calculation models were obtained by training deep learning models based on historical data with lead equivalent greater than 0.5 and historical data with lead equivalent less than or equal to 0.5 respectively. The input of the lead equivalent calculation model is the sample radiation transmittance, and the output is the lead equivalent.

[0010] The radiation transmittance of the sample under test was measured using a standard radiation field and monitoring equipment.

[0011] Based on the radiation transmittance of the sample to be tested, one of two lead equivalent calculation models is selected as the current calculation model to calculate the lead equivalent of the sample to be tested.

[0012] Optionally, the construction of the deep learning model based on the LSTM neural network specifically includes:

[0013] A deep learning model is constructed, comprising an input layer, a hidden layer, and an output layer; the input layer performs logarithmic processing on the input data; the hidden layer adopts a network structure that alternates between LSTM layers and fully connected layers.

[0014] Optionally, the LSTM layer includes multiple LSTM units; the LSTM unit includes a forget gate, an input gate, and an output gate.

[0015] Optionally, the step of training deep learning models based on historical data with lead equivalent greater than 0.5 and historical data with lead equivalent less than or equal to 0.5 respectively to obtain two lead equivalent calculation models specifically includes:

[0016] Deep learning models are trained based on historical data with lead equivalent greater than 0.5 and historical data with lead equivalent less than or equal to 0.5. During the training process, after the output layer outputs the lead equivalent, the difference between the expected output lead equivalent and the actual output lead equivalent is calculated using the loss function. The gradient of the loss function with respect to each model parameter is calculated using backpropagation and the derivative chain rule. The parameters of the deep learning model are iteratively updated according to the gradient, so that the loss function is continuously reduced until a deep learning model with accuracy that meets the test requirements is obtained as the lead equivalent calculation model.

[0017] Optionally, the measurement of the radiation transmittance of the sample under test using a standard radiation field and monitoring equipment specifically includes:

[0018] Record the standard capacitance value C0 and the corresponding monitoring equipment reading V0 when no sample is placed in the standard radiation field;

[0019] Record the adaptation capacitance value C and the corresponding monitoring equipment reading V after placing the sample to be tested in the standard radiation field;

[0020] Using formula Calculate the radiation transmittance x of the sample to be tested.

[0021] Optionally, the step of selecting one of two lead equivalent calculation models as the current calculation model based on the radiation transmittance of the sample to be tested, and calculating the lead equivalent of the sample to be tested, specifically includes:

[0022] When the radiation transmittance of the sample to be tested is greater than or equal to 3.91554 × 10 -5When the lead equivalent is less than 0.0375, the lead equivalent calculation model trained based on historical data with a lead equivalent greater than 0.5 is selected as the current calculation model to calculate the lead equivalent of the sample to be tested.

[0023] When the radiation transmittance of the sample to be tested is greater than or equal to 0.0375 and less than 1, the lead equivalent calculation model trained based on historical data with lead equivalent less than or equal to 0.5 is selected as the current calculation model to calculate the lead equivalent of the sample to be tested.

[0024] A lead equivalent calculation system for radiation protection equipment includes:

[0025] The historical data acquisition module is used to acquire historical data on the lead equivalent measurement of radiation protection equipment samples; the historical data includes the sample's radiation transmittance and the corresponding lead equivalent.

[0026] The deep learning model building module is used to build deep learning models based on LSTM neural networks.

[0027] The model training module is used to train deep learning models based on historical data with lead equivalent greater than 0.5 and historical data with lead equivalent less than or equal to 0.5, respectively, to obtain two lead equivalent calculation models; the input of the lead equivalent calculation model is the sample radiation transmittance, and the output is the lead equivalent.

[0028] The radiation transmittance acquisition module is used to measure the radiation transmittance of the sample under test using a standard radiation field and monitoring equipment.

[0029] The lead equivalent calculation module is used to select one of two lead equivalent calculation models as the current calculation model based on the radiation transmittance of the sample to be tested, in order to calculate the lead equivalent of the sample.

[0030] An electronic device includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement a method for calculating the lead equivalent of the radiation protection equipment.

[0031] Optionally, the memory is a non-transitory computer-readable storage medium.

[0032] According to specific embodiments provided by the present invention, the present invention discloses the following technical effects:

[0033] This invention provides a method, system, and device for calculating the lead equivalent of radiation protection equipment. By constructing and utilizing a deep learning model, training the model with relatively accurate historical data, and optimizing and correcting the training results until the model stabilizes, a lead equivalent calculation model is obtained. Compared to the traditional method of manually reading simulation curves, this lead equivalent calculation model significantly improves computational efficiency, accuracy, and the correctness and stability of batch data, effectively saving labor and time costs. It achieves rapid, accurate, automated, and intelligent calculation of the lead equivalent of radiation protection equipment. This invention can significantly improve the calculation efficiency of lead equivalent of radiation protection equipment, while also enhancing the accuracy and stability of lead equivalent calculation results for batch samples. Attached Figure Description

[0034] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0035] Figure 1 A flowchart illustrating the method for calculating the lead equivalent of radiation protection equipment provided by the present invention;

[0036] Figure 2 A schematic diagram of the structure of the deep learning model provided by this invention;

[0037] Figure 3 A schematic diagram of the LSTM unit provided by this invention;

[0038] Figure 4 A schematic diagram illustrating the lead equivalent calculation process provided by this invention. Detailed Implementation

[0039] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0040] The purpose of this invention is to provide a method, system, and equipment for calculating the lead equivalent of radiation protection equipment, so as to improve the calculation efficiency, accuracy, and stability of lead equivalent of batch samples, and save labor and time costs.

[0041] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0042] Figure 1 A flowchart illustrating the method for calculating the lead equivalent of radiation protection equipment provided by this invention. (For example...) Figure 1 As shown, the present invention discloses a method for calculating the lead equivalent of radiation protection equipment, comprising:

[0043] Step 1: Obtain historical data on lead equivalent measurements of radiation protection equipment samples.

[0044] The historical data obtained by this invention includes the sample radiation transmittance and the corresponding lead equivalent. By sorting through the historical data, erroneous historical data that deviates significantly from the normal values ​​are deleted.

[0045] Step 2: Build a deep learning model based on the LSTM neural network.

[0046] This invention constructs a deep learning model comprising an input layer, a hidden layer, and an output layer, the specific structure of which is as follows: Figure 2 As shown, the sample radiation transmittance x is used as the input to the constructed deep learning model, and the input layer performs logarithmic processing on the input data; the hidden layer adopts a network structure of alternating LSTM layers and fully connected layers, the LSTM layer includes multiple LSTM units, each LSTM unit includes a forget gate, an input gate and an output gate; the lead equivalent data y corresponding to the input data is used as the output of the deep learning model.

[0047] As a specific implementation, the hidden layer adopts a network structure with alternating superposition of two LSTM layers and three fully connected layers. It has good prediction performance and fast speed, and can effectively reduce time costs compared with other network structures.

[0048] Step 3: Train deep learning models based on historical data with lead equivalent greater than 0.5 and historical data with lead equivalent less than or equal to 0.5 respectively to obtain two lead equivalent calculation models.

[0049] Two separate models were trained for data with lead equivalent greater than 0.5 and data with lead equivalent less than 0.5, because experiments have shown that this segmented prediction yields the highest accuracy. The radiation transmittance-lead equivalent data were fed into the deep learning model for training. To ensure the deep learning model converges quickly, the input data x was logarithmically processed in the input layer of the deep learning model, as shown in formula (1) below.

[0050] x'=lgx (1)

[0051] In the formula, x' is the output of the input layer of the deep learning model.

[0052] The output x' of the input layer of the deep learning model enters the fully connected layer. Each node of the fully connected layer uses the sigmoid function, and the specific calculation formula (2) is shown below.

[0053] x d1 =sig(x')=1 / (1+exp(-x')) (2)

[0054] In the formula, x d1 This is the output of the first fully connected layer.

[0055] Figure 3 A schematic diagram of the LSTM unit provided by this invention is shown below. Figure 3 As shown, x d1 After processing by the first fully connected layer, the data enters the first LSTM layer. The LSTM layer consists of multiple LSTM units, each containing a forget gate, an input gate, and an output gate. The calculation process for the forget gate is shown in formula (3), based on the output value h of the previous LSTM unit. n-1 and the current input value x n The output f of the forget gate is obtained using the sigmoid activation function. n .

[0056] f n =σ(W f ·[h n-1 ,x n ]+b f (3)

[0057] The calculation process of the input gate is shown in formulas (4) to (6), and is also based on h. n-1 and x n Using the sigmoid activation function to obtain i n i n Let h be a weight vector representing the proportion of data updates. n-1 and x n Using the tanh activation function to obtain candidate cell states Make f n The cell state value C of the previous LSTM unit n-1 Multiply, plus i n and Multiplication yields a new cell state value C. n .

[0058] i n =σ(W i ·[h n-1 ,x n ]+b i (4)

[0059]

[0060]

[0061] The calculation process of the output gate is shown in formulas (7) and (8), and is also based on h. n-1 and x n Obtaining o using the sigmoid activation function n Then, the tanh activation function is used on C. n Process and with o n Multiply to obtain the output h of the LSTM unit. n .

[0062] o n =σ(W o ·[h n-1 ,x n ]+b o (7)

[0063] h n =o n *tanh(C n (8)

[0064] In formulas (1) to (7), W f W i W C and W o The output f of the forget gate is respectively n The weight vector i of the input gate n Cell state candidate values and output gate o n The corresponding coefficient matrix, b f b i b C and b o f n i n , and o n The corresponding bias vector. Therefore, the output of the first LSTM layer is x. L1 .

[0065] x L1 After passing through the first LSTM layer, the signal enters the second fully connected layer. The output of the second fully connected layer is x. d2 ;x d2 After passing through the second fully connected layer, the output of the second LSTM layer is x. L2 ;x L2 After passing through the second LSTM layer, the signal enters the third fully connected layer. The output of the third fully connected layer is x. d3 x d3After passing through the output layer, the output of the deep learning model is obtained, which is the lead equivalent y.

[0066] After obtaining the output lead equivalent y, the difference between the expected output and the actual output is calculated using the loss function. Backpropagation is used to calculate the gradient of the loss function with respect to each model parameter through the chain rule of derivatives. Based on the gradient, the parameters of the deep learning model are iteratively updated. By continuously updating the parameters of the deep learning model, the loss function is continuously reduced, and finally a lead equivalent calculation model with accuracy that meets the test requirements is obtained.

[0067] To improve the prediction accuracy of the lead equivalent calculation model, this invention trains two lead equivalent calculation models separately using historical data with lead equivalent greater than 0.5 and historical data with lead equivalent less than 0.5. The lead equivalent calculation model trained using historical data with lead equivalent greater than 0.5 is designated as Model 1, and the lead equivalent calculation model trained using historical data with lead equivalent less than or equal to 0.5 is designated as Model 2.

[0068] Step 4: Measure the radiation transmittance of the sample under test using a standard radiation field and monitoring equipment.

[0069] When calculating the lead equivalent in actual measurements, it is necessary to first obtain the radiation transmittance value corresponding to the sample to be tested. To obtain the radiation transmittance, it is necessary to record the standard capacitance value C0 and the corresponding monitoring equipment reading V0 in the standard radiation field when the sample to be tested is not placed; after placing the sample, record the adaptation capacitance value C and the corresponding monitoring equipment reading V. Calculate the radiation transmittance value x that needs to be input into the model according to formula (9).

[0070]

[0071] Step 5: Select one of the two lead equivalent calculation models as the current calculation model based on the radiation transmittance of the sample to be tested, and calculate the lead equivalent of the sample to be tested.

[0072] Figure 4 A schematic diagram illustrating the lead equivalent calculation process provided by this invention. (See diagram below.) Figure 4 As shown, when the radiation transmittance of the sample to be tested is less than 3.91554 × 10⁻⁶, -5 When the lead equivalent is greater than 3.5, it is directly determined that the lead equivalent is greater than 3.5; when the radiation transmittance of the sample is greater than or equal to 3.91554 × 10⁻⁶. -5When the lead equivalent is less than 0.0375, the lead equivalent calculation model (Model 1) trained based on historical data with a lead equivalent greater than 0.5 is selected as the current calculation model, and Model 1 is used to calculate the lead equivalent of the sample to be tested; when the radiation transmittance of the sample to be tested is greater than or equal to 0.0375 and less than 1, the lead equivalent calculation model (Model 2) trained based on historical data with a lead equivalent less than or equal to 0.5 is selected as the current calculation model, and Model 2 is used to calculate the lead equivalent of the sample to be tested; when the radiation transmittance of the sample to be tested is 1, the lead equivalent is determined to be 0.

[0073] The deep learning model constructed in the lead equivalent calculation method for radiation protection equipment provided by this invention has a simple structure, which can significantly improve the calculation efficiency, accuracy and stability of lead equivalent of batch samples, save labor and time costs, and realize the automated and intelligent calculation of lead equivalent.

[0074] Based on the lead equivalent calculation method for radiation protection equipment provided by this invention, a lead equivalent calculation system for radiation protection equipment is also provided, comprising:

[0075] The historical data acquisition module is used to acquire historical data on the lead equivalent measurement of radiation protection equipment samples; the historical data includes the sample's radiation transmittance and the corresponding lead equivalent.

[0076] The deep learning model building module is used to build deep learning models based on LSTM neural networks.

[0077] The model training module is used to train deep learning models based on historical data with lead equivalent greater than 0.5 and historical data with lead equivalent less than or equal to 0.5, respectively, to obtain two lead equivalent calculation models; the input of the lead equivalent calculation model is the sample radiation transmittance, and the output is the lead equivalent.

[0078] The radiation transmittance acquisition module is used to measure the radiation transmittance of the sample under test using a standard radiation field and monitoring equipment.

[0079] The lead equivalent calculation module is used to select one of two lead equivalent calculation models as the current calculation model based on the radiation transmittance of the sample to be tested, in order to calculate the lead equivalent of the sample.

[0080] Furthermore, the present invention also provides an electronic device, which may include: a processor, a communication interface, a memory, and a communication bus. The processor, communication interface, and memory communicate with each other via the communication bus. The processor can call a computer program stored in the memory to execute the lead equivalent calculation method for the radiation protection equipment.

[0081] Furthermore, when the computer program in the aforementioned memory is implemented as a software functional unit and sold or used as an independent product, it can be stored in a non-transitory computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or a 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 the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory, random access memory, magnetic disks, or optical disks.

[0082] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on its differences from other embodiments. Similar or identical parts between embodiments can be referred to interchangeably. For the systems disclosed in the embodiments, since they correspond to the methods disclosed in the embodiments, the descriptions are relatively simple; relevant parts can be referred to the method section.

[0083] This document uses specific examples to illustrate the principles and implementation methods of the present invention. The descriptions of the above embodiments are only for the purpose of helping to understand the method and core ideas of the present invention. Furthermore, those skilled in the art will recognize that, based on the ideas of the present invention, there will be changes in the specific implementation methods and application scope. Therefore, the content of this specification should not be construed as a limitation of the present invention.

Claims

1. A method for calculating the lead equivalent of radiation protection equipment, characterized in that, include: Obtain historical data on lead equivalent measurements of radiation protection equipment samples; the historical data includes the sample's radiation transmittance and the corresponding lead equivalent. Building deep learning models based on LSTM neural networks; Two lead equivalent calculation models were obtained by training deep learning models based on historical data with lead equivalent greater than 0.5 and historical data with lead equivalent less than or equal to 0.5 respectively. The input of the lead equivalent calculation model is the sample radiation transmittance, and the output is the lead equivalent. The radiation transmittance of the sample under test was measured using a standard radiation field and monitoring equipment. Based on the radiation transmittance of the sample to be tested, one of two lead equivalent calculation models is selected as the current calculation model to calculate the lead equivalent of the sample to be tested. When the radiation transmittance of the sample to be measured is greater than or equal to 3.91554x10 -5 and less than 0.0375, a lead equivalent calculation model trained based on historical data of a lead equivalent greater than 0.5 is selected as a current calculation model to calculate the lead equivalent of the sample to be measured. When the radiation transmittance of the sample to be tested is greater than or equal to 0.0375 and less than 1, the lead equivalent calculation model trained based on historical data with lead equivalent less than or equal to 0.5 is selected as the current calculation model to calculate the lead equivalent of the sample to be tested.

2. The method for calculating the lead equivalent of radiation protection equipment according to claim 1, characterized in that, The construction of a deep learning model based on an LSTM neural network specifically includes: A deep learning model is constructed, comprising an input layer, a hidden layer, and an output layer; the input layer performs logarithmic processing on the input data; the hidden layer adopts a network structure that alternates between LSTM layers and fully connected layers.

3. The method for calculating the lead equivalent of radiation protection equipment according to claim 2, characterized in that, The LSTM layer includes multiple LSTM units; each LSTM unit includes a forget gate, an input gate, and an output gate.

4. The method for calculating the lead equivalent of radiation protection equipment according to claim 2, characterized in that, The process involves training deep learning models based on historical data with lead equivalents greater than 0.5 and historical data with lead equivalents less than or equal to 0.5, respectively, to obtain two lead equivalent calculation models. Specifically, these models include: Deep learning models are trained based on historical data with lead equivalent greater than 0.5 and historical data with lead equivalent less than or equal to 0.

5. During the training process, after the output layer outputs the lead equivalent, the difference between the expected output lead equivalent and the actual output lead equivalent is calculated using the loss function. The gradient of the loss function with respect to each model parameter is calculated using backpropagation and the derivative chain rule. The parameters of the deep learning model are iteratively updated according to the gradient, so that the loss function is continuously reduced until a deep learning model with accuracy that meets the test requirements is obtained as the lead equivalent calculation model.

5. The method for calculating the lead equivalent of radiation protection equipment according to claim 1, characterized in that, The measurement of the radiation transmittance of the sample under test using a standard radiation field and monitoring equipment specifically includes: Record the standard capacitance value C0 and the corresponding monitoring equipment reading V0 when no sample is placed in the standard radiation field; Record the adaptation capacitance value C and the corresponding monitoring equipment reading V after placing the sample to be tested in the standard radiation field; Using formula Calculate the radiation transmittance of the sample to be tested. .

6. A lead equivalent calculation system for radiation protection equipment, characterized in that, include: The historical data acquisition module is used to acquire historical data on the lead equivalent measurement of radiation protection equipment samples; the historical data includes the sample's radiation transmittance and the corresponding lead equivalent. The deep learning model building module is used to build deep learning models based on LSTM neural networks. The model training module is used to train deep learning models based on historical data with lead equivalent greater than 0.5 and historical data with lead equivalent less than or equal to 0.5, respectively, to obtain two lead equivalent calculation models; the input of the lead equivalent calculation model is the sample radiation transmittance, and the output is the lead equivalent. The radiation transmittance acquisition module is used to measure the radiation transmittance of the sample under test using a standard radiation field and monitoring equipment. The lead equivalent calculation module is used to select one of two lead equivalent calculation models as the current calculation model based on the radiation transmittance of the sample to be tested, in order to calculate the lead equivalent of the sample to be tested. When the radiation transmittance of the sample to be tested is greater than or equal to 3.91554 × 10 -5 When the lead equivalent is less than 0.0375, the lead equivalent calculation model trained based on historical data with a lead equivalent greater than 0.5 is selected as the current calculation model to calculate the lead equivalent of the sample to be tested. When the radiation transmittance of the sample to be tested is greater than or equal to 0.0375 and less than 1, the lead equivalent calculation model trained based on historical data with lead equivalent less than or equal to 0.5 is selected as the current calculation model to calculate the lead equivalent of the sample to be tested.

7. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the method for calculating the lead equivalent of radiation protection equipment as described in any one of claims 1 to 5.

8. The electronic device according to claim 7, characterized in that, The memory is a non-transitory computer-readable storage medium.

Images