Method for converting radioactivity of each nuclide using artificial neural network

Abstract

This method for converting the radioactivity of each nuclide using an artificial neural network comprises the steps of: analyzing a spectrum of a radionuclide by using an artificial neural network and outputting an output layer of the artificial neural network as the radioactivity ratio of each nuclide; and converting the radioactivity ratio of each nuclide into the radioactivity of each nuclide.

Description

A method for converting the radioactivity of individual nuclides using artificial neural networks

[0001] This paper relates to a method for converting the radioactivity of individual nuclides using an artificial neural network.

[0002] An artificial neural network is one of the artificial intelligence technologies modeled after the structure of human neurons and is composed of an input layer, a hidden layer, and an output layer. When the hidden layer consists of multiple layers, it is called deep learning.

[0003] An artificial neural network is a network structure composed of interconnected neurons or nodes. Training data entered through the input layer is processed through the hidden layer, and the final result is output through the output layer. The output layer uses various output functions, such as sigmoid and softmax.

[0004] The sigmoid function is used for binary classification, where one of two options is selected, and the softmax function is used for multi-class classification, where one of three or more options is selected.

[0005] The softmax function normalizes the input values ​​to values ​​between 0 and 1, and has the characteristic that the sum of the outputs is always 1.

[0006] When using an artificial neural network for radioactive nuclide analysis, the softmax function is used in the output layer because it is created as an algorithm that predicts the abundance ratio of various nuclides.

[0007] However, since the output layer data obtained by the softmax function is the radioactivity ratio of individual nuclides included in the radioactive nuclide, even if the amount of radioactivity in the sample to be analyzed is large, the amount of radioactivity of individual nuclides cannot be known.

[0008] For example, when analyzing one sample containing a mixture of three individual nuclides, A (20 Bq), B (30 Bq), and C (50 Bq), and another sample containing a mixture of three individual nuclides, D (40 Bq), E (60 Bq), and F (100 Bq), using an artificial neural network, if the output layer is a softmax function, the radioactivity ratios of each individual nuclide, 0.2, 0.3, and 0.5, are output equally for each individual nuclide, A, B, and C, or D, E, and F, so the amount of radioactivity of each nuclide cannot be known.

[0009] One embodiment is to provide a method for converting the radioactivity of individual nuclides using an artificial neural network, which converts the radioactivity ratio of individual nuclides into the radioactivity amount of individual nuclides when analyzing radioactive nuclides using an artificial neural network.

[0010] One aspect provides a method for converting the radioactivity of individual nuclides using an artificial neural network, including the steps of analyzing a spectrum of a radioactive nuclide using an artificial neural network, outputting an output layer of the artificial neural network as a radioactivity ratio of individual nuclides, and converting the radioactivity ratio of individual nuclides into the radioactivity of individual nuclides.

[0011] The step of converting the radioactivity ratio of the above individual nuclides into the radioactivity of the individual nuclides can be calculated using the following mathematical formula 1.

[0012] [Mathematical Formula 1]

[0013]

[0014] Here, A j is the radioactivity of the jth nuclide, n is the total net count rate, and r i is the radioactivity ratio of the i-th nuclide, and r j is the radioactivity ratio of the jth nuclide, and ε i is the efficiency of the ith nuclide, and y j is the gamma-ray emission rate of the jth nuclide.

[0015] The step of outputting the output layer of the above artificial neural network as the radioactivity ratio of each nuclide can use a softmax function in the output layer.

[0016] In the above mathematical expression 1, the above ε i can be calculated using the mathematical formula 2 below.

[0017] [Mathematical Formula 2]

[0018]

[0019] Here, A c is the radioactivity of the certified material, and n i is the net counting rate of the ith nuclide, and y i is the gamma-ray emission rate of the ith nuclide.

[0020] According to one embodiment, a method for converting the radioactivity of individual nuclides using an artificial neural network is provided, which converts the radioactivity ratio of individual nuclides into the radioactivity amount of individual nuclides during radionuclide analysis using an artificial neural network.

[0021] Figure 1 is a flowchart illustrating a method for converting the radioactivity of individual nuclides using an artificial neural network according to one embodiment.

[0022] Hereinafter, with reference to the attached drawings, embodiments of the present invention will be described in detail so that those skilled in the art can easily implement the present invention. The present invention may be implemented in various different forms and is not limited to the embodiments described herein.

[0023] Additionally, throughout the specification, whenever a part is said to "include" a component, this does not mean that it excludes other components, but rather that it may include other components, unless otherwise specifically stated.

[0024] Hereinafter, a method for converting the radioactivity of individual nuclides using an artificial neural network according to one embodiment will be described with reference to FIG. 1.

[0025] One embodiment relates to a method for converting the radioactivity ratio of individual nuclides into the radioactivity of individual nuclides when analyzing the spectrum of radioactive nuclides using an artificial neural network and configuring the output layer with a softmax function. The radioactivity of individual nuclides can be obtained by using the efficiency of individual nuclides obtained in advance, the gamma-ray emission rate of individual nuclides, and the radioactivity ratio of individual nuclides, which is the output value of the softmax function. The artificial neural network can be implemented using various known computing devices, but is not limited thereto.

[0026] Figure 1 is a flowchart illustrating a method for converting the radioactivity of individual nuclides using an artificial neural network according to one embodiment.

[0027] Referring to Figure 1, first, the spectrum of a radioactive nuclide is analyzed using an artificial neural network, and the output layer of the artificial neural network outputs the radioactivity ratio of each nuclide (S100).

[0028] For example, an artificial neural network can be used to analyze the spectrum of radionuclides for radionuclide analysis. The softmax function can be used in the output layer, allowing the artificial neural network to output the radioactivity ratios of individual nuclides. The softmax function can be expressed using Equation 1 below, but is not limited thereto. Various known mathematical expressions for softmax functions can also be used.

[0029] [Mathematical Formula 1]

[0030]

[0031] Next, the radioactivity ratio of individual nuclides is converted into the radioactivity of individual nuclides (S200).

[0032] For example, the total net coefficient (n T ) is the value obtained by subtracting the background count rate from the count rate measured by the radioactive nuclide, and when multiple nuclides are mixed, it can be expressed by the following mathematical formula 2.

[0033] [Mathematical Formula 2]

[0034]

[0035] Here, n T is the total net coefficient, and A i is the radioactivity of the ith nuclide, and ε i is the efficiency of the ith nuclide, and y i is the gamma-ray emission rate of the ith nuclide.

[0036] Efficiency of the ith nuclide (ε i ) can be calculated using the mathematical formula 3 below.

[0037] [Mathematical Formula 3]

[0038]

[0039] Here, A c is the radioactivity of the certified material, and n i is the net counting rate of the ith nuclide, and y i is the gamma-ray emission rate of the ith nuclide.

[0040] The above-described mathematical formula 3 may correspond to the mathematical formula 2 described in the claims, but is not limited thereto.

[0041] The softmax function value of the output layer of the artificial neural network is r, which is the radioactivity ratio of the ith nuclide, which is the relative contribution of the radioactivity and emission rate of each nuclide to the entire spectrum. i , and can be expressed by the mathematical formula 4 below.

[0042] [Equation 4]

[0043]

[0044] r i is the radioactivity ratio of the i-th nuclide, and A i is the radioactivity of the ith nuclide, and y i is the gamma-ray emission rate of the ith nuclide.

[0045] The relative contribution ratio of the ith nuclide and the jth nuclide can be expressed by the mathematical equation 5 below.

[0046] [Equation 5]

[0047]

[0048] Therefore, even if the spectra of multiple nuclides are measured, the radioactivity of each nuclide can be expressed as a relative ratio between the individual nuclides. For example, the softmax output results of measuring N nuclides and analyzing them with an artificial neural network are r1, r2, …, r N And A2, …, A N It can be expressed as mathematical formula 6 below with A1 as the following.

[0049] [Equation 6]

[0050]

[0051] A2, …, A in mathematical expression 6 N After substituting into the mathematical equation 2 described above and organizing for A1, it can be expressed as the mathematical equation 7 below.

[0052] [Equation 7]

[0053]

[0054] Likewise, A1, A3, …, A N After substituting into Equation 2, the radioactivity of the jth nuclide, A j If we organize it, the following mathematical formula 8 is derived.

[0055] [Equation 8]

[0056]

[0057] Here, A j is the radioactivity of the jth nuclide, n is the total net count rate, and r i is the radioactivity ratio of the i-th nuclide, and r j is the radioactivity ratio of the jth nuclide, and ε i is the efficiency of the ith nuclide, and y j is the gamma-ray emission rate of the jth nuclide.

[0058] The above-described mathematical formula 8 may correspond to the mathematical formula 1 described in the claims, but is not limited thereto.

[0059] As above, using mathematical expression 8, the output values ​​of the softmax function are the radioactivity ratios of individual nuclides, r1, r2, …, r N can be converted into the radioactivity value of individual nuclides. The conversion of the radioactivity ratio of individual nuclides into the radioactivity of individual nuclides can be calculated using mathematical formula 8.

[0060] For example, when analyzing the spectrum of a radioactive nuclide using an artificial neural network, if the output layer is configured with a softmax function, the radioactivity ratio for each nuclide is output, and according to one embodiment, this value can be converted into the radioactivity of each nuclide.

[0061] A method for converting the radioactivity of individual nuclides using an artificial neural network is provided, which converts the radioactivity ratio of individual nuclides into the radioactivity amount of individual nuclides when analyzing radioactive nuclides using an artificial neural network.

[0062] Although the embodiments of the present invention have been described in detail above, the scope of the present invention is not limited thereto, and various modifications and improvements made by those skilled in the art using the basic concept of the present invention defined in the following claims also fall within the scope of the present invention.

Claims

1. A step of analyzing the spectrum of a radioactive nuclide using an artificial neural network and outputting the output layer of the artificial neural network as the radioactivity ratio of each nuclide; and A step of converting the radioactivity ratio of the above individual nuclides into the radioactivity of the individual nuclides. A method for converting the radioactivity of individual nuclides using an artificial neural network including .

2. In paragraph 1, The step of converting the radioactivity ratio of the above individual nuclides into the radioactivity of the individual nuclides is a method of converting the radioactivity of the individual nuclides using an artificial neural network, which is calculated using the following mathematical formula 1. [Mathematical Formula 1] Here, A j is the radioactivity of the jth nuclide, n is the total net count rate, and r i is the radioactivity ratio of the i-th nuclide, and r j is the radioactivity ratio of the jth nuclide, and ε i is the efficiency of the ith nuclide, and y j is the gamma-ray emission rate of the jth nuclide.

3. In paragraph 2, The step of outputting the output layer of the artificial neural network as the radioactivity ratio of each nuclide is a method for converting the radioactivity of each nuclide using an artificial neural network that uses a softmax function in the output layer.

4. In paragraph 2, In the above mathematical expression 1, the above ε i A method for converting the radioactivity of individual nuclides using an artificial neural network calculated using the mathematical formula 2 below. [Mathematical Formula 2] Here, A c is the radioactivity of the certified material, and n i is the net counting rate of the ith nuclide, and y i is the gamma-ray emission rate of the ith nuclide.

Images