Early warning method and system based on 18F-FDG efficiency calculation model

Abstract

The invention relates to the technical field of computer processing, in particular to an early warning method and system based on an 18F-FDG efficiency calculation model. The method comprises the following steps: obtaining a corresponding fluorine ion yield influence factor and a 18F-FDG synthesis influence factor; performing primary and secondary influence division and criterion weight determination on the fluorine ion yield influence factor and the 18F-FDG synthesis influence factor to obtain a relative importance weight corresponding to each influence factor; efficiency calculation model iterative training and efficiency prediction are carried out based on the relative importance weight corresponding to each influence factor so as to generate the current corresponding 18F-FDG production efficiency; and performing 18F-FDG production early warning based on a preset first-level early warning threshold value and a second-level early warning threshold value so as to generate a corresponding 18F-FDG production early warning adjustment measure. According to the method, each influence factor in the 18F-FDG production process can be comprehensively considered to accurately predict the corresponding 18F-FDG efficiency.

Description

Technical Field

[0001] The present invention relates to the field of computer processing technology, and in particular to a method based on 18 Early warning method and system for F-FDG efficiency calculation model. Background Art

[0002] Fluorodeoxyglucose injection ( 18 F-FDG) is a radioactive tracer commonly used in positron emission tomography (PET) imaging examinations and is widely used in the medical field. 18 The production process of F-FDG usually requires efficient synthesis and quality control to ensure its clinical application effect. Therefore, how to monitor and predict the real-time 18 F-FDG production efficiency and early warning of potential problems have become the key to improving 18 The key to the quality and efficiency of F-FDG production. 18 Real-time monitoring of key parameters in the F-FDG production process (such as reaction temperature, reaction time, raw material concentration, equipment performance, etc.) 18 The computational model of F-FDG production efficiency can evaluate the efficiency fluctuation during the production process. 18 Indicators such as process review and data tracing of the F-FDG efficiency calculation method are difficult to capture the factors affecting production efficiency in real time during the production process. At the same time, there is a lack of dynamic response analysis between various influencing factors, and it is impossible to achieve early warning effects. Summary of the Invention

[0003] Based on this, the present invention is necessary to provide a 18 An early warning method and system for an F-FDG efficiency calculation model are provided to solve at least one of the above technical problems.

[0004] To achieve the above purpose, a 18 The early warning method of the F-FDG efficiency calculation model includes the following steps:

[0005] Step S1: By 18 The historical production process of F-FDG was used to obtain the factors affecting the yield of fluoride ions produced by cyclotron nuclear reactions and 18 The factors affecting F-FDG synthesis include the abundance of oxygen-18 water, the amount of oxygen-18 water used, the amount of oxygen-18 water consumed during transmission in a closed target, the proton beam target position, the proton beam energy, the proton beam current, and the irradiation time. 18 The factors affecting F-FDG synthesis include the amount of precursors prepared for production, reagent content, fluoride ion pipeline transmission consumption, QMA full load, precursor purity, catalyst activity, reaction temperature, reaction time, 18F-FDG synthesis channel number, automated synthesis efficiency and 18 F-FDG production;

[0006] Step S2: Factors affecting fluoride ion yield and 18 The F-FDG synthesis influencing factors were divided into primary and secondary influences to obtain 18 The main and secondary influencing factors corresponding to F-FDG efficiency; 18 The main influencing factors and secondary influencing factors corresponding to F-FDG efficiency are determined by criterion weights to obtain the corresponding 18 F-FDG relative importance weight;

[0007] Step S3: Based on the corresponding 18 F-FDG relative importance weight 18 The main influencing factors and secondary influencing factors corresponding to F-FDG efficiency are iteratively trained to generate the efficiency calculation model. 18 F-FDG efficiency calculation model;

[0008] Step S4: Get the current 18 The main and secondary influencing factors corresponding to the F-FDG production process are input into the trained 18 F-FDG efficiency calculation model 18 F-FDG efficiency prediction to generate the current corresponding 18 F-FDG efficiency; based on the preset first-level warning threshold and second-level warning threshold, the current corresponding 18 F-FDG efficiency 18 F-FDG production warning to generate corresponding 18 F-FDG production early warning adjustment measures.

[0009] Furthermore, step S1 includes the following steps:

[0010] Step S11: Using a mass spectrometer to obtain the corresponding abundance of oxygen-18 water produced by the cyclotron nuclear reaction, and using a flow sensor and a liquid level sensor to monitor the corresponding flow and storage amount of oxygen-18 water in the target practice process in real time, so as to calculate the corresponding oxygen-18 water consumption based on the corresponding flow and storage amount of oxygen-18 water; and by monitoring each target transfer stop point or target transfer position on the target transfer track during the target practice process, obtain the corresponding oxygen-18 water transmission consumption in the closed target;

[0011] Step S12: using a cyclotron to monitor and obtain the number of single or double target positions corresponding to the proton beam in the target practice phase to obtain the proton beam target position; using a radiation monitoring sensor to accurately measure the energy and flow rate corresponding to the proton beam in the target practice phase and the irradiation duration to obtain the proton beam energy, proton beam current, and irradiation time; using the corresponding oxygen-18 water abundance, oxygen-18 water usage, oxygen-18 water transmission consumption in the closed target, proton beam target position, proton beam energy, proton beam current, and irradiation time as factors affecting fluoride ion yield;

[0012] Step S13: By 18 During the historical production of F-FDG, capacitive level sensors and weighing sensors were used to accurately measure the volume and mass of materials in the reaction tubes to obtain the amount of material required for production. Spectral analysis sensors were used to quickly and accurately detect the concentration and composition of production reagents to determine the reagent content, including trifluoromannose, anhydrous acetonitrile, K2.2.2 / K2CO3 eluent, and NaOH.

[0013] Step S14: Obtain the corresponding fluoride ion dosage after the nuclear reaction and the fluoride ion dosage after passing through the QMA column through the corresponding fluoride ion pipeline transmission process in the synthesis link to obtain the fluoride ion pipeline transmission consumption and the QMA full load; obtain the precursor purity, catalyst activity, reaction temperature, reaction time, 18 The number of F-FDG synthesis channels, including FDG1 and FDG2, is used to obtain the corresponding automated synthesis efficiency through two water removal and one nucleophilic reaction processes in the synthesis process. 18 F-FDG production; the corresponding production preparation precursor amount, reagent content, fluoride ion pipeline transmission consumption, QMA full load, precursor purity, catalyst activity, reaction temperature, reaction time, 18 F-FDG synthesis channel number, automated synthesis efficiency and 18 F-FDG production as 18 Factors affecting F-FDG synthesis.

[0014] Furthermore, the reaction tube described in step S13 includes a QMA column, a long tC18 column and a combination column, wherein the combination column includes an IC-H column, an AL2O3 column and a tC18 column.

[0015] Furthermore, the step S14 of obtaining the corresponding automated synthesis efficiency by performing two water removals and one nucleophilic reaction process in the synthesis step includes:

[0016] The corresponding water removal rate is obtained by performing two water removal processes on the reaction reagents in the synthesis process, and the water content is obtained by using a desiccant to absorb water and distill to remove water;

[0017] The moisture content is statistically analyzed based on the water removal rate to obtain the moisture residual 18 Residual water content before F-FDG synthesis reaction;

[0018] Based on the residual water content, the corresponding nucleophilic reaction process in the synthesis link was analyzed. 18 F-FDG synthesis loss assessment to obtain the effect of reaction reagent water content on 18 F-FDG synthesis reaction affects the loss;

[0019] The corresponding nucleophilic reaction temperature and nucleophilic reaction time are obtained during the nucleophilic reaction process in the synthesis link, and the nucleophilic reaction rate analysis is performed on the corresponding nucleophilic reaction process in the synthesis link based on the nucleophilic reaction temperature and nucleophilic reaction time to obtain the corresponding nucleophilic reaction rate at the corresponding reaction temperature and time. 18 F-FDG synthesis reaction rate;

[0020] Obtain the corresponding raw material input amount and 18 F-FDG synthesis yield, and based on 18 The influence of F-FDG synthesis reaction loss and the corresponding reaction temperature and time 18 The F-FDG synthesis reaction rate is affected by the amount of raw materials input and 18 F-FDG synthesis yield 18 The F-FDG synthesis rate was quantified to obtain 18 The automated synthesis efficiency corresponding to F-FDG.

[0021] Furthermore, the step of obtaining the corresponding moisture content of the reaction reagent by performing two corresponding water removal processes in the synthesis step includes the following steps:

[0022] The corresponding drying water absorption reaction time and distillation water reaction time are obtained through the two corresponding water removal processes in the synthesis process;

[0023] The drying water absorption capacity of the corresponding desiccant in the synthesis process is analyzed to obtain the drying water absorption capacity of the reaction reagent;

[0024] The corresponding distillation water migration amount of the reaction reagent is obtained by the corresponding distillation water separation process in the synthesis link, and the reaction reagent drying water absorption amount and the reaction reagent distillation water migration amount corresponding to the two dehydration processes are synthesized based on the drying water absorption reaction time and the distillation water reaction time to obtain the corresponding reaction reagent moisture content.

[0025] Furthermore, step S2 includes the following steps:

[0026] Step S21: Factors affecting fluoride ion yield and18 Each factor in the F-FDG synthetic impact factor is combined with other factors to perform correlation measurement calculations to quantify the corresponding correlation coefficients between the two combinations of impact factors and form the corresponding impact factor correlation matrix to generate 18 F-FDG impact factor correlation matrix;

[0027] Step S22: Based on 18 The correlation matrix of F-FDG influencing factors on the influencing factors of fluoride ion yield and 18 The influence factors of F-FDG synthesis were constructed by the influence factor association grid to integrate the influence factors of fluoride ion yield and 18 Each impact factor in the F-FDG synthetic impact factor is used as a node, and according to 18 The correlation coefficient between the two influencing factor combinations in the F-FDG influencing factor correlation matrix determines the edges connecting them. If the correlation coefficient is greater than or equal to 0.85, the edges between the nodes are connected; if the correlation coefficient is less than 0.85, no processing is done and the corresponding 18 F-FDG impact factor association network;

[0028] Step S23: 18 The F-FDG impact factor association network is analyzed for node degree and betweenness centrality, in order to analyze the number of edges connected to the corresponding node as the corresponding degree of the node, and to analyze the betweenness centrality of the corresponding node to obtain the 18 The node degree and betweenness centrality corresponding to the F-FDG impact factor;

[0029] Step S24: Based on each 18 The node degree and betweenness centrality corresponding to the F-FDG impact factor have an impact on the fluoride ion yield factor and 18 Each factor in the F-FDG synthesis impact factor is divided into primary and secondary impacts. 18 If the node degree of the F-FDG impact factor exceeds 5 or the betweenness centrality exceeds 8, it is determined as the main impact factor, otherwise it is determined as the secondary impact factor to obtain 18 The main and secondary influencing factors corresponding to F-FDG efficiency;

[0030] Step S25: 18 The main influencing factors and secondary influencing factors corresponding to F-FDG efficiency are determined by criterion weights to obtain the corresponding 18 F-FDG relative importance weight.

[0031] Furthermore, step S25 includes the following steps:

[0032] Step S251: By18 Obtain the corresponding F-FDG historical production process 18 F-FDG historical production efficiency and use it as the target layer;

[0033] Step S252: 18 The main influencing factors corresponding to F-FDG efficiency are used as the criterion layer, and 18 The secondary influencing factors corresponding to F-FDG efficiency are used as the indicator layer; the target layer, criterion layer and indicator layer are constructed into the corresponding expert evaluation system architecture according to the hierarchical structure;

[0034] Step S253: Expert scoring is performed based on the correlation coefficients between each influencing factor corresponding to the criterion layer and the indicator layer within the expert evaluation system architecture, and a corresponding scoring judgment matrix is constructed based on the scoring results of each influencing factor, where the elements in the scoring judgment matrix represent the score ratio relationship between each influencing factor;

[0035] Step S254: Obtain the corresponding maximum eigenvalue and its corresponding eigenvector through the scoring judgment matrix, and perform consistency check calculation based on the maximum eigenvalue and its corresponding eigenvector to calculate the corresponding consistency index and random consistency index, and calculate the ratio of the consistency index and the random consistency index to obtain the corresponding consistency ratio;

[0036] Step S255: If the consistency ratio is greater than or equal to 0.1, the corresponding scoring result is determined as the corresponding relative importance weight to obtain the corresponding 18 F-FDG relative importance weight; if the consistency ratio is less than 0.1, it will return to the expert scoring until the scoring judgment matrix passes the consistency test.

[0037] Furthermore, step S3 includes the following steps:

[0038] Step S31: Based on the corresponding 18 The relative importance weight of F-FDG was combined with multiple linear regression to analyze the 18 The main influencing factors and secondary influencing factors corresponding to F-FDG efficiency are used to train the efficiency calculation model to 18 The relative importance weight of F-FDG is calculated with the main impact factor and the secondary impact factor as input. 18 The F-FDG efficiency is used as output, generating 18 Calculate the initial model for F-FDG efficiency and output the corresponding prediction 18 F-FDG efficiency;

[0039] Step S32: Predict 18 F-FDG efficiency and historical average 18Comparative analysis of F-FDG efficiency was performed to obtain the deviation between the predicted efficiency and the historical average efficiency;

[0040] Step S33: Compare and judge the deviation between the predicted efficiency and the historical average efficiency according to the preset deviation threshold of 1%. If the deviation between the predicted efficiency and the historical average efficiency is less than or equal to the deviation threshold of 1%, the model will not be optimized; if the deviation between the predicted efficiency and the historical average efficiency is greater than the deviation threshold of 1%, the model will be re-optimized. 18 The corresponding initial model for F-FDG efficiency calculation 18 The relative importance weight of F-FDG is determined by the criterion weight, and the re-determined 18 F-FDG relative importance weight iterative optimization corresponding to 18 F-FDG efficiency calculation initial model to generate 18 F-FDG efficiency calculation model.

[0041] Furthermore, step S4 includes the following steps:

[0042] Step S41: Get the current 18 The main and secondary influencing factors corresponding to the F-FDG production process;

[0043] Step S42: The current 18 The main and secondary influencing factors corresponding to the F-FDG production process are input into the trained 18 F-FDG efficiency calculation model 18 F-FDG efficiency prediction is to compare the corresponding main influencing factors and secondary influencing factors with 18 The weights corresponding to each influencing factor in the F-FDG efficiency calculation model are weighted to generate the current corresponding 18 F-FDG efficiency;

[0044] Step S43: by presetting the corresponding first-level warning threshold of 85% and the second-level warning threshold of 70%;

[0045] Step S44: Based on the preset first-level warning threshold of 85% and second-level warning threshold of 70%, the current corresponding 18 F-FDG efficiency 18 F-FDG production warning, if the current corresponding 18 When the F-FDG efficiency is lower than the first-level warning threshold of 85%, a corresponding minor warning signal will be issued to indicate that there is an impact. 18 Potential factors for F-FDG efficiency and alert production personnel to conduct further inspection and analysis in time; if the current corresponding 18 When the F-FDG efficiency is lower than the secondary warning threshold of 70%, a corresponding serious warning signal will be issued to indicate the occurrence of18 The system will warn the production personnel to stop production immediately to avoid producing defective products, check whether the pipeline transmission system is leaking or blocked, and activate the preset emergency plan for radioactive material transmission pipeline leakage when necessary. 18 The corresponding spare targets and spare synthesis modules in the F-FDG production process are supplied first, and when the radiation dose in the target room drops to an acceptable range for personnel, the accelerator is inspected, maintained and troubleshooted to generate the corresponding 18 F-FDG production early warning adjustment measures.

[0046] Furthermore, the present invention also provides a method based on 18 The F-FDG efficiency calculation model is used to perform the above-mentioned early warning system based on 18 The early warning method of F-FDG efficiency calculation model is based on 18 The early warning system of the F-FDG efficiency calculation model includes:

[0047] 18 F-FDG impact factor analysis module is used to analyze the 18 The historical production process of F-FDG was used to obtain the factors affecting the yield of fluoride ions produced by cyclotron nuclear reactions and 18 The factors affecting F-FDG synthesis include the abundance of oxygen-18 water, the amount of oxygen-18 water used, the amount of oxygen-18 water consumed during transmission in a closed target, the proton beam target position, the proton beam energy, the proton beam current, and the irradiation time. 18 The factors affecting F-FDG synthesis include the amount of precursors prepared for production, reagent content, fluoride ion pipeline transmission consumption, QMA full load, precursor purity, catalyst activity, reaction temperature, reaction time, 18 F-FDG synthesis channel number, automated synthesis efficiency and 18 F-FDG production;

[0048] Impact factor weight determination module is used to determine the impact factor weight based on 18 The factors affecting the historical production efficiency of F-FDG on the yield of fluoride ions and 18 The F-FDG synthesis influencing factors were divided into primary and secondary influences to obtain 18 The main and secondary influencing factors corresponding to F-FDG efficiency; 18 The main influencing factors and secondary influencing factors corresponding to F-FDG efficiency are determined by criterion weights to obtain the corresponding 18 F-FDG relative importance weight;

[0049] Efficiency calculation model training module is used to calculate the efficiency of the model based on the corresponding 18 F-FDG relative importance weight18 The main influencing factors and secondary influencing factors corresponding to F-FDG efficiency are iteratively trained to generate the efficiency calculation model. 18 F-FDG efficiency calculation model;

[0050] 18 F-FDG production warning module is used to obtain current 18 The main and secondary influencing factors corresponding to the F-FDG production process are input into the trained 18 F-FDG efficiency calculation model 18 F-FDG efficiency prediction to generate the current corresponding 18 F-FDG efficiency; based on the preset first-level warning threshold and second-level warning threshold, the current corresponding 18 F-FDG efficiency 18 F-FDG production warning to generate corresponding 18 F-FDG production early warning adjustment measures.

[0051] Beneficial effects of the present invention:

[0052] 1. The present invention is based on 18 Compared with the prior art, the early warning method of the F-FDG efficiency calculation model has the following advantages: 18 During the F-FDG production process, by analyzing the factors affecting the fluoride ion yield, such as the abundance of target material (oxygen-18 water), the amount of target material (oxygen-18 water), the transmission consumption of target material (oxygen-18 water) in the closed target, the proton beam target position, the proton beam energy, the proton beam current and the irradiation time, we can understand the influence of fluoride ion yield on the oxygen-18 water nuclear reaction. 18 The interaction between O and proton beam and the key factors affecting fluoride ion yield are studied to precisely control the 18 The reaction conditions during the production of F-FDG provide data support to help optimize the efficiency of the fluorination reaction. 18 Factors affecting F-FDG synthesis include the amount of precursors prepared for production, reagent content, fluoride ion pipeline transmission consumption, QMA full load, precursor purity, catalyst activity, reaction temperature, reaction time, 18 F-FDG synthesis channel number, automated synthesis efficiency and 18 F-FDG production is affected 18 The final yield and purity of F-FDG are important factors. Accurately understanding these factors can help optimize 18 The amount of raw materials and reagents used in the F-FDG synthesis process, the channel selection of the synthesis reaction, and the automation level of the production process can be used to capture the factors affecting production efficiency in real time during the production process.18 The primary and secondary impact factors of F-FDG synthesis are divided into two groups, which helps to systematically identify the 18 The factors that have the greatest impact on F-FDG efficiency can provide a key focus for further optimizing the production process and effectively screen out the most important factors from a large number of influencing factors. 18 The most influential factors and secondary factors of F-FDG efficiency. In addition, the determination of criterion weights is a key step in quantifying the relative importance of each influencing factor. By establishing a reasonable weight distribution model, the influence of each influencing factor on the efficiency of F-FDG can be accurately calculated. 18 The contribution of F-FDG efficiency can be evaluated and a theoretical basis can be provided for subsequent model optimization. In the process of weight determination, the correlation between the influencing factors can be evaluated through hierarchical analysis and expert scoring, so as to accurately determine the weight of each factor, which can better realize the dynamic response analysis between the influencing factors. Then, based on the 18 The relative importance weight of F-FDG is used to iteratively train the efficiency calculation model, which can build an accurate 18 The F-FDG efficiency prediction model can combine theoretical research with practical production through repeated optimization of various influencing factors and model training, thereby improving the accuracy and practicality of the prediction. This step helps to use machine learning, data mining and other technical methods to conduct a comprehensive analysis of the influencing factors using historical production data to explore the influencing factors. 18 The potential law of F-FDG efficiency is continuously optimized and adjusted to gradually improve the stability and reliability of the efficiency calculation model. The efficiency calculation model obtained by training is not only 18 The optimization of F-FDG production process provides theoretical support and can effectively guide production decisions in actual production, thereby improving 18 F-FDG production efficiency. Finally, by 18 The F-FDG efficiency calculation model is applied in the actual production process, which can monitor and predict the current 18 The efficiency of F-FDG production is to reduce the current 18 When the main and secondary influencing factors in the F-FDG production process are input into the trained model, the model can give the corresponding 18 The predicted value of F-FDG efficiency provides a quantitative basis for adjusting the production process. Furthermore, based on the set first and second level warning thresholds, the model can 18When the F-FDG production efficiency deviates from the normal range, a graded warning signal will be automatically issued. The first-level warning can remind the potential risks in the production process, while the second-level warning can help production personnel take more specific response measures, such as adjusting reaction conditions or replacing raw materials. This mechanism can effectively reduce risks in production and avoid waste of resources and time delays caused by low production efficiency. At the same time, through real-time 18 F-FDG production early warning adjustment measures can respond in the first time and adjust the production strategy, thereby improving 18 Corresponding early warning effect during F-FDG production process.

[0053] 2. The present invention proposes 18 The early warning system of the F-FDG efficiency calculation model is composed of 18 F-FDG impact factor analysis module, impact factor weight determination module, efficiency calculation model training module and 18 F-FDG production early warning module is composed of 18 The early warning method of the F-FDG efficiency calculation model is used to realize the operation between the computer programs running on each module based on 18 The early warning method of the F-FDG efficiency calculation model and the mutual cooperation of the internal structure of the system can greatly reduce duplication of work and manpower input, and can quickly and effectively provide more accurate and efficient 18 The early warning process of the F-FDG efficiency calculation model is simplified 18 Operational procedures of the early warning system for the F-FDG efficiency calculation model. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] Other features, objects and advantages of the present invention will become more apparent upon reading the detailed description of non-limiting embodiments thereof made with reference to the following drawings:

[0055] Figure 1 The present invention is based on 18 Schematic diagram of the steps of the early warning method of the F-FDG efficiency calculation model;

[0056] Figure 2 for Figure 1 Detailed step flow diagram of step S1;

[0057] Figure 3 for Figure 1 Detailed step flow chart of step S2 in FIG. DETAILED DESCRIPTION

[0058] The following is a clear and complete description of the technical method of the present invention in conjunction with the accompanying drawings. It is obvious that the embodiments described are part of the embodiments of the present invention, but not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without making any creative efforts are within the scope of protection of the present invention.

[0059] In addition, the accompanying drawings are merely schematic illustrations of the present invention and are not necessarily drawn to scale. Identical reference numerals in the figures denote identical or similar parts, and thus repetitive descriptions thereof will be omitted. Some of the block diagrams shown in the accompanying drawings are functional entities that do not necessarily correspond to physically or logically separate entities. These functional entities may be implemented in software, in one or more hardware modules or integrated circuits, or in different network and / or processor and / or microcontroller approaches.

[0060] It should be understood that although the terms "first," "second," and the like may be used herein to describe various elements, these elements should not be limited by these terms. These terms are used solely to distinguish one element from another. For example, a first element may be referred to as a second element, and similarly, a second element may be referred to as a first element, without departing from the scope of the exemplary embodiments. The term "and / or" as used herein includes any and all combinations of one or more of the listed associated items.

[0061] To achieve this, please refer to Figures 1 to 3 The present invention provides a method based on 18 An early warning method for a F-FDG efficiency calculation model, comprising the following steps:

[0062] Step S1: By 18 The historical production process of F-FDG was used to obtain the factors affecting the yield of fluoride ions produced by cyclotron nuclear reactions and 18 The factors affecting F-FDG synthesis include the abundance of oxygen-18 water, the amount of oxygen-18 water used, the amount of oxygen-18 water consumed during transmission in a closed target, the proton beam target position, the proton beam energy, the proton beam current, and the irradiation time. 18 The factors affecting F-FDG synthesis include the amount of precursors prepared for production, reagent content, fluoride ion pipeline transmission consumption, QMA full load, precursor purity, catalyst activity, reaction temperature, reaction time, 18 F-FDG synthesis channel number, automated synthesis efficiency and 18 F-FDG production;

[0063] Step S2: Factors affecting fluoride ion yield and 18The influencing factors of F-FDG synthesis were divided into primary and secondary factors to obtain the primary and secondary influencing factors corresponding to 18F-FDG efficiency; 18 The main influencing factors and secondary influencing factors corresponding to F-FDG efficiency are determined by criterion weights to obtain the corresponding 18 F-FDG relative importance weight;

[0064] Step S3: Based on the corresponding 18 F-FDG relative importance weight 18 The main influencing factors and secondary influencing factors corresponding to F-FDG efficiency are iteratively trained to generate the efficiency calculation model. 18 F-FDG efficiency calculation model;

[0065] Step S4: Get the current 18 The main and secondary influencing factors corresponding to the F-FDG production process are input into the trained 18 F-FDG efficiency calculation model 18 F-FDG efficiency prediction to generate the current corresponding 18 F-FDG efficiency; based on the preset first-level warning threshold and second-level warning threshold, the current corresponding 18 F-FDG efficiency 18 F-FDG production warning to generate corresponding 18 F-FDG production early warning adjustment measures.

[0066] In the embodiment of the present invention, please refer to Figure 1 As shown, the present invention is based on 18 Schematic diagram of the steps of the early warning method of the F-FDG efficiency calculation model. In this example, the 18 The early warning method of the F-FDG efficiency calculation model includes the following steps:

[0067] Step S1: By 18 The historical production process of F-FDG was used to obtain the factors affecting the yield of fluoride ions produced by cyclotron nuclear reactions and 18 The factors affecting F-FDG synthesis include the abundance of oxygen-18 water, the amount of oxygen-18 water used, the amount of oxygen-18 water consumed during transmission in a closed target, the proton beam target position, the proton beam energy, the proton beam current, and the irradiation time. 18 The factors affecting F-FDG synthesis include the amount of precursors prepared for production, reagent content, fluoride ion pipeline transmission consumption, QMA full load, precursor purity, catalyst activity, reaction temperature, reaction time, 18 F-FDG synthesis channel number, automated synthesis efficiency and 18 F-FDG production;

[0068] In the embodiment of the present invention, by 18 Relevant influencing factors are obtained during the historical production process of F-FDG. For the influencing factors of fluoride ion yield, the abundance of oxygen-18 water is determined by mass spectrometry. The mass spectrometer parameters are set to a specific range to accurately identify the isotope peak of oxygen-18 water. For example, the mass number scanning range is set to 18-20amu and the resolution is 1000. The abundance value of oxygen-18 water is obtained. The amount of oxygen-18 water is calculated by the flow sensor and the liquid level sensor. The flow sensor is installed on the conveying pipeline and uses the electromagnetic induction principle to measure the flow rate in real time. The liquid level sensor is installed in the storage tank and measures the liquid level based on the static pressure principle. The total flow rate is obtained by integrating the speed data, and the consumption is calculated in combination with the liquid level change. The transmission consumption in the closed target is measured by installing high-precision mass flow meters at key positions on the target transmission track, with an accuracy of up to 0.1 grams. The proton beam target position is determined by obtaining the target position switching signal from the cyclotron control system. It is clear at a glance whether it is a single target position or a double target position. The proton beam energy is measured using an energy detector such as a magnetic spectrometer; the proton beam current is measured using an ionization chamber type detector based on the characteristic that radiation ionizes the gas to generate current; the irradiation time is measured by a high-precision timer, starting from the time the proton beam is turned on and stopping at the end. 18 The factors affecting F-FDG synthesis are as follows: the amount of precursor prepared for production is measured by measuring the volume of the material in the reaction tube with a capacitive liquid level sensor and the mass with a weighing sensor; the concentration and composition of production reagents (such as trifluoromannose and anhydrous acetonitrile) are measured using a spectral analysis sensor, such as a UV-visible spectrophotometer, according to the Lambert-Beer law; the fluoride ion pipeline transmission consumption and the QMA full load are measured by installing a radioactivity activity meter at the starting end of the pipeline and after passing through the QMA column, respectively; the difference between the two is the transmission consumption; the activity when the QMA column is saturated is the full load; the purity of the precursor is determined by high-performance liquid chromatography (HPLC) and compared with the chromatogram of the standard; the catalyst activity is determined by measuring the catalytic rate under specified conditions through a specific catalytic reaction experiment; the reaction temperature is monitored in real time by inserting a thermocouple temperature sensor into the reaction vessel; the reaction time is measured from the start to the end of the reaction using a high-precision timer. 18 The number of F-FDG synthesis channels is determined by checking the operating status of the synthesis equipment, and the automated synthesis efficiency is calculated according to the corresponding method. 18 The F-FDG yield is directly measured by radioactivity meter. Through these specific tools and methods, the factors affecting the fluoride ion yield and the 18 Factors affecting F-FDG synthesis.

[0069] Step S2: Factors affecting fluoride ion yield and 18 The F-FDG synthesis influencing factors were divided into primary and secondary influences to obtain 18The main and secondary influencing factors corresponding to F-FDG efficiency; 18 The main influencing factors and secondary influencing factors corresponding to F-FDG efficiency are determined by criterion weights to obtain the corresponding 18 F-FDG relative importance weight;

[0070] In the embodiment of the present invention, the factors affecting the fluoride ion yield and the 18 The influencing factors of F-FDG synthesis are divided into primary and secondary influences. First, the Pearson correlation coefficient between each pair of influencing factors is calculated by the DataFrame.corr() method. For example, the correlation coefficient between the number of target shots and the corresponding production preparation column quantity is 0.7, and the correlation coefficient between the automated synthesis rate and the target water consumption is 0.85. The correlation coefficient threshold is set, such as 0.85. The influencing factors with a correlation coefficient greater than or equal to 0.85 are determined as the primary influencing factors, and those with a correlation coefficient less than 0.85 are determined as secondary influencing factors. After determining the primary and secondary influencing factors, the hierarchical analysis method (AHP) is used to determine the criterion weights. By using the yaahp software to construct a hierarchical structure model, the primary influencing factors are used as the criterion layer, and the secondary influencing factors are used as the indicator layer. The target is 18 F-FDG efficiency, organize relevant field experts (such as chemical process experts, production engineers) to score the relative importance of each influencing factor, build a judgment matrix, and the software automatically calculates the maximum eigenvalue and its corresponding eigenvector, and performs consistency test. If the consistency ratio is less than 0.1, the test is passed, and the eigenvector is normalized to obtain the corresponding value of each influencing factor. 18 F-FDG relative importance weight.

[0071] Step S3: Based on the corresponding 18 F-FDG relative importance weight 18 The main influencing factors and secondary influencing factors corresponding to F-FDG efficiency are iteratively trained to generate the efficiency calculation model. 18 F-FDG efficiency calculation model;

[0072] In the embodiment of the present invention, the scikit-learn library of Python is used to calculate the influence factors corresponding to the 18 F-FDG relative importance weight 18 The main influencing factors and secondary influencing factors corresponding to F-FDG efficiency are iteratively trained for the efficiency calculation model. The main influencing factors and secondary influencing factors in the historical production data are sorted into a feature matrix. Each row represents the influencing factor data of a production process. 18F-FDG efficiency is used as a label, and the LinearRegression() function is used to create a multivariate linear regression model object. The weight vector and the feature matrix are weighted and calculated to start training the model. For example, after the first round of training, the prediction error of the model is large. By adjusting the weight (such as re-using AHP to determine the weight based on the training results), training is performed again. After multiple iterative trainings, until the prediction error of the model meets the requirements (such as the mean square error (ie, deviation) is less than the preset value), the model is generated. 18 F-FDG efficiency calculation model is stored as a Python pickle file for subsequent use.

[0073] Step S4: Get the current 18 The main and secondary influencing factors corresponding to the F-FDG production process are input into the trained 18 F-FDG efficiency calculation model 18 F-FDG efficiency prediction to generate the current corresponding 18 F-FDG efficiency; based on the preset first-level warning threshold and second-level warning threshold, the current corresponding 18 F-FDG efficiency 18 F-FDG production warning to generate corresponding 18 F-FDG production early warning adjustment measures.

[0074] In the embodiment of the present invention, by 18 During the F-FDG production process, the corresponding main and secondary influencing factors are obtained through the real-time data acquisition system. For example, the number of target shots is obtained by real-time monitoring of the target shooting process by sensors, and the automatic synthesis rate is obtained from the control system of the automatic synthesis equipment. These real-time data are organized into feature vectors in the same format as those used in the training model, and the trained model is loaded by using the pickle.load() function. 18 F-FDG efficiency calculation model, input the feature vector into the model for prediction, and generate the current corresponding 18 F-FDG efficiency, assuming the preset first-level warning threshold is 85%, and the second-level warning threshold is 70% (determined based on historical production data and process requirements), the current 18 F-FDG efficiency and historical standards 18 F-FDG efficiency (obtained from historical production data) compared to the calculated percentage, if the current 18 If the F-FDG efficiency percentage is lower than 85% but higher than 70%, a text message will be sent to the production staff through the SMS platform of the production early warning system to alert them of any impact. 18Potential factors affecting F-FDG efficiency; if it is lower than 70%, not only will a text message be sent, but also a warning message will be displayed on the display screen in the production workshop, and the automatic control system will be activated to suspend production equipment, and at the same time generate maintenance, adjust process parameters, replace unqualified reagents, etc. 18 F-FDG production early warning adjustment measures to ensure production stability and product quality.

[0075] Furthermore, step S1 includes the following steps:

[0076] Step S11: Using a mass spectrometer to obtain the corresponding abundance of oxygen-18 water produced by the cyclotron nuclear reaction, and using a flow sensor and a liquid level sensor to monitor the corresponding flow and storage amount of oxygen-18 water in the target practice process in real time, so as to calculate the corresponding oxygen-18 water consumption based on the corresponding flow and storage amount of oxygen-18 water; and by monitoring each target transfer stop point or target transfer position on the target transfer track during the target practice process, obtain the corresponding oxygen-18 water transmission consumption in the closed target;

[0077] Step S12: using a cyclotron to monitor and obtain the number of single or double target positions corresponding to the proton beam in the target practice phase to obtain the proton beam target position; using a radiation monitoring sensor to accurately measure the energy and flow rate corresponding to the proton beam in the target practice phase and the irradiation duration to obtain the proton beam energy, proton beam current, and irradiation time; using the corresponding oxygen-18 water abundance, oxygen-18 water usage, oxygen-18 water transmission consumption in the closed target, proton beam target position, proton beam energy, proton beam current, and irradiation time as factors affecting fluoride ion yield;

[0078] Step S13: By 18 During the historical production of F-FDG, capacitive level sensors and weighing sensors were used to accurately measure the volume and mass of materials in the reaction tubes to obtain the amount of material required for production. Spectral analysis sensors were used to quickly and accurately detect the concentration and composition of production reagents to determine the reagent content, including trifluoromannose, anhydrous acetonitrile, K2.2.2 / K2CO3 eluent, and NaOH.

[0079] Step S14: Obtain the corresponding fluoride ion dosage after the nuclear reaction and the fluoride ion dosage after passing through the QMA column through the corresponding fluoride ion pipeline transmission process in the synthesis link to obtain the fluoride ion pipeline transmission consumption and the QMA full load; obtain the precursor purity, catalyst activity, reaction temperature, reaction time, 18 The number of F-FDG synthesis channels, including FDG1 and FDG2, is used to obtain the corresponding automated synthesis efficiency through two water removal and one nucleophilic reaction processes in the synthesis link, and the corresponding 18F-FDG production; the corresponding production preparation precursor amount, reagent content, fluoride ion pipeline transmission consumption, QMA full load, precursor purity, catalyst activity, reaction temperature, reaction time, 18 F-FDG synthesis channel number, automated synthesis efficiency and 18 F-FDG production as 18 Factors affecting F-FDG synthesis.

[0080] As an embodiment of the present invention, refer to Figure 2 As shown, Figure 1 Detailed step flow diagram of step S1 in FIG. 1 , in this embodiment, step S1 includes the following steps:

[0081] Step S11: Using a mass spectrometer to obtain the corresponding abundance of oxygen-18 water produced by the cyclotron nuclear reaction, and using a flow sensor and a liquid level sensor to monitor the corresponding flow and storage amount of oxygen-18 water in the target practice process in real time, so as to calculate the corresponding oxygen-18 water consumption based on the corresponding flow and storage amount of oxygen-18 water; and by monitoring each target transfer stop point or target transfer position on the target transfer track during the target practice process, obtain the corresponding oxygen-18 water transmission consumption in the closed target;

[0082] In an embodiment of the present invention, when acquiring data related to oxygen-18 water produced by a cyclotron nuclear reaction, a mass spectrometer is connected to a nuclear reaction product collection device, and the scanning parameters of the mass spectrometer are set so that the isotopic abundance of oxygen-18 water can be accurately detected. For example, the scanning mass range is set to 18-20 amu, and the resolution is adjusted to 1000 to clearly distinguish the mass spectrum peaks of oxygen-18 water from other impurities, thereby accurately obtaining the abundance of oxygen-18 water. A flow sensor is installed on the oxygen-18 water delivery pipeline in the target shooting link. Its working principle is based on electromagnetic induction or ultrasonic measurement, and it measures the water flow rate in real time. The liquid level sensor is installed in the oxygen-18 water storage tank. Inside, a static pressure level gauge is used to convert the liquid level height by measuring the liquid pressure, and then obtain the storage capacity, and the flow rate data of the flow sensor is integrated with time. For example, within 10 minutes, the flow rate data is integrated to obtain a total flow of 5 ml. On the target transfer track, a high-precision mass flow meter with an accuracy of up to 0.1 gram is installed at each target transfer stop or target transfer position. When oxygen-18 water enters the closed target through the target transfer track, the mass flow meter monitors its mass change in real time, thereby obtaining the transmission consumption in the closed target. For example, in a certain target transfer process, the mass before entering the closed target is 10 grams, and after the target transfer, it becomes 9.8 grams, and the transmission consumption is 0.2 grams.

[0083] Step S12: using a cyclotron to monitor and obtain the number of single or double target positions corresponding to the proton beam in the target practice phase to obtain the proton beam target position; using a radiation monitoring sensor to accurately measure the energy and flow rate corresponding to the proton beam in the target practice phase and the irradiation duration to obtain the proton beam energy, proton beam current, and irradiation time; using the corresponding oxygen-18 water abundance, oxygen-18 water usage, oxygen-18 water transmission consumption in the closed target, proton beam target position, proton beam energy, proton beam current, and irradiation time as factors affecting fluoride ion yield;

[0084] In an embodiment of the present invention, the number of single or double target positions corresponding to the proton beam in the target shooting phase is obtained by utilizing the monitoring system provided by the cyclotron accelerator. The monitoring system determines the number of target positions by identifying the target position switching signal inside the accelerator. If the signal shows that only one target position is currently in operation, the proton beam target position is a single target position; if the activation signal of two target positions is detected, it is a double target position. The radiation monitoring sensor adopts an ionization chamber type detector, whose working principle is based on the characteristic that radiation ionizes gas to generate current, and accurately measures the energy, flow rate and irradiation time corresponding to the proton beam in the target shooting phase, and The detector is placed on the proton beam transmission path. The proton beam current is converted by measuring the ionization current. For example, if the measured current is 10 microamperes, the proton beam current is converted to 100 nanoamperes. A high-precision timer is used to record the irradiation time. For example, if the irradiation time is 30 minutes, the proton beam energy is measured by the energy detector. For example, the measured value is 16MeV. At the same time, the oxygen-18 water abundance, oxygen-18 water usage, oxygen-18 water transmission consumption in the closed target, proton beam target position, proton beam energy, proton beam current and irradiation time obtained previously are summarized as factors affecting fluoride ion yield.

[0085] Step S13: By 18 During the historical production of F-FDG, capacitive level sensors and weighing sensors were used to accurately measure the volume and mass of materials in the reaction tubes to obtain the amount of material required for production. Spectral analysis sensors were used to quickly and accurately detect the concentration and composition of production reagents to determine the reagent content, including trifluoromannose, anhydrous acetonitrile, K2.2.2 / K2CO3 eluent, and NaOH.

[0086] In the embodiment of the present invention, by 18During the synthesis phase of the F-FDG production process, capacitive level sensors were installed on QMA columns, long tC18 columns, and combination columns (including IC-H columns, AL2O3 columns, and tC18 columns) to measure the volume of materials in the reaction tubes. The volume of the materials in the columns was accurately measured by detecting changes in capacitance, and data was recorded every 15 seconds. At the same time, a Sartorius weighing sensor was installed at the bottom of each reaction tube to measure the mass of the materials using gravity sensing. Data was recorded every 20 seconds, and the volume and mass data were combined to obtain the volume before production preparation. For example, in the reaction tube, the capacitive liquid level sensor measures the material volume as 200 ml, and the weighing sensor measures the material mass as 220 g, thereby determining the production preparation column volume of the column. The spectral analysis sensor is used to detect the production reagents. By analyzing the reagents' absorption of light of specific wavelengths, the corresponding concentration and composition of the production reagents can be quickly and accurately detected. For example, for 1 ml of trifluoromannose or anhydrous acetonitrile reagent, the spectral analysis sensor determines that the concentration of trifluoromannose is 0.5 mol / L and the concentration of anhydrous acetonitrile is 0.5 mol / L based on the position and intensity of the absorption peak of light of specific wavelengths. The same method is used to detect the content of 1.5 ml of K2.2.2 / K2C03 eluent, NaOH and other reagents, and these data are summarized as the reagent content.

[0087] Step S14: Obtain the corresponding fluoride ion dosage after the nuclear reaction and the fluoride ion dosage after passing through the QMA column through the corresponding fluoride ion pipeline transmission process in the synthesis link to obtain the fluoride ion pipeline transmission consumption and the QMA full load; obtain the precursor purity, catalyst activity, reaction temperature, reaction time, 18 The number of F-FDG synthesis channels, including FDG1 and FDG2, is used to obtain the corresponding automated synthesis efficiency through two water removal and one nucleophilic reaction processes in the synthesis process. 18 F-FDG production; the corresponding production preparation precursor amount, reagent content, fluoride ion pipeline transmission consumption, QMA full load, precursor purity, catalyst activity, reaction temperature, reaction time, 18 F-FDG synthesis channel number, automated synthesis efficiency and 18 F-FDG production as 18 Factors affecting F-FDG synthesis.

[0088] In an embodiment of the present invention, during the fluoride ion pipeline transmission process in the synthesis link, radioactivity meters are installed at the starting end of the pipeline and after passing through the QMA column. The working principle is based on the scintillator detecting the flash generated by radioactive particles, and the flash is converted into an electrical signal by a photomultiplier tube for measurement, thereby obtaining the fluoride ion dose after the nuclear reaction and the fluoride ion dose after passing through the QMA column. The two are subtracted to obtain the fluoride ion pipeline transmission consumption. For example, if the starting end dose is 100MBq and the dose after passing through the QMA column is 90MBq, the transmission consumption is 10MBq. The full load of QMA is obtained by measuring the radioactivity after the QMA column is saturated with fluoride ions. In addition, by obtaining the corresponding reaction parameters in the synthesis process, the purity of the precursor is detected by high performance liquid chromatography (HPLC), the precursor sample is injected into the HPLC system, and the purity is determined by comparing it with the chromatogram of the standard. The catalyst activity is determined by a specific catalytic reaction experiment, such as measuring the catalytic rate of a catalyst for a certain reaction under certain conditions. The reaction temperature is measured using a thermocouple temperature sensor installed on the wall of the reaction vessel to display the temperature in real time, such as the reaction temperature is 60°C, and the reaction time is recorded by a timer. The time from the start to the end of the reaction is the reaction time, such as the reaction time is 20 minutes. In addition, it is also necessary to determine 18 The number of synthesis channels corresponding to F-FDG synthesis, such as FDG1 and FDG2 channels, is obtained by reading the relevant configuration parameters in the synthesis equipment control system. During the two water removals and one nucleophilic reaction in the synthesis process, the execution time and completion status of each step are recorded by the automatic control system. The corresponding reaction loss, reaction rate, raw material input and 18 The synthesis yield of F-FDG is obtained by automated synthesis rate. For example, the synthesis rate can also be obtained by transmitting 1000mci of F-18 ions, through chemical reactions in the synthesizer, purification, separation, and finally outputting 18 F-FDG, the activity meter measured 500mci, so the synthesis efficiency is 50%. 18 A high-precision weighing sensor is installed on the F-FDG product collection container to measure the collected 18 F-FDG mass, according to 18 The density of F-FDG is converted to volume, and we get 18 F-FDG production, at the same time, the production preparation precursor amount, reagent content, fluoride ion pipeline transmission consumption, QMA full load, precursor purity, catalyst activity, reaction temperature, reaction time, 18 F-FDG synthesis channel number, automated synthesis efficiency and 18 F-FDG production was summarized as 18 F-FDG synthesis impact factor, used to analyze the impact 18 Various factors in F-FDG synthesis.

[0089] Furthermore, the reaction tube described in step S13 includes a QMA column, a long tC18 column and a combination column, wherein the combination column includes an IC-H column, an AL2O3 column and a tC18 column.

[0090] Furthermore, the step S14 of obtaining the corresponding automated synthesis rate through two corresponding water removal processes and one corresponding nucleophilic reaction process in the synthesis link includes:

[0091] The corresponding water removal rate is obtained by performing two water removal processes on the reaction reagents in the synthesis process, and the water content is obtained by using a desiccant to absorb water and distill to remove water;

[0092] In an embodiment of the present invention, the dehydration rate is obtained by performing two dehydration processes on the reaction reagent in the synthesis process. First, a desiccant water absorption method is adopted, and a molecular sieve is selected as a desiccant. It is added to the reaction reagent in a certain ratio (such as 1:10, molecular sieve mass: reaction reagent volume), and the mixture is fully stirred for 30 minutes to allow the desiccant to fully contact and adsorb the water. The water content of the reaction reagent before and after adsorption is measured using a Karl Fischer moisture meter. If the water content before adsorption is 5% and the water content after adsorption is 3%, distillation is then performed to remove the water. Water, place the reaction reagent in a distillation apparatus, set a suitable distillation temperature (such as the boiling point of the reaction reagent + 10°C) and vacuum degree (such as 0.08MPa), and perform distillation for 30 minutes. Measure the water content after distillation with a Karl Fischer titrator again. Assuming it drops to 1%, the water removal rate is calculated as: (initial water content - final water content) ÷ initial water content × 100%, that is, (5% - 1%) ÷ 5% × 100% = 80%. Through such operations, the corresponding water removal rate is finally obtained, and the water content of each stage is determined.

[0093] Preferably, the water content is statistically analyzed based on the water removal rate to obtain 18 Residual water content before F-FDG synthesis reaction;

[0094] In the embodiment of the present invention, the water content is statistically analyzed based on the water removal rate to obtain 18 The residual water content before the F-FDG synthesis reaction is 80% of the water removal rate. The total amount of the initial reaction reagent is 100 ml, and the initial water content is 5%. Then the initial water content is 100×5%=5 ml. After two water removals, the remaining water content is the initial water content×(1-water removal rate), that is, 5×(1-80%)=1 ml. This 1 ml of water content is converted into a proportion of the total volume of the current reaction reagent, that is, 1÷(100-5+1)×100%≈1.02% (because the total volume of the reaction reagent changes slightly during the water removal process). This proportion is 18The residual water content before the F-FDG synthesis reaction can be accurately calculated through this calculation method based on the water removal rate and the initial water content.

[0095] Preferably, the corresponding nucleophilic reaction process in the synthesis link is carried out based on the residual water content. 18 F-FDG synthesis loss assessment to obtain the effect of reaction reagent water content on 18 F-FDG synthesis reaction affects the loss;

[0096] In the embodiment of the present invention, the corresponding nucleophilic reaction process in the synthesis link is analyzed based on the residual water content. 18 F-FDG synthesis loss evaluation to obtain the effect of reaction reagent water content on 18 The loss of F-FDG synthesis reaction is affected by the reaction loss model established based on historical experimental data. The model shows that the residual water content is positively correlated with the loss of synthesis reaction. Assuming that the residual water content increases by 0.1%, 18 The F-FDG synthesis yield loss is 5%. The current residual moisture content is known to be 1.02%. Compared with the ideal anhydrous state (assuming the ideal state loss is 0), the residual moisture content has increased by 1.02%. According to the model calculation, the loss of the synthesis reaction is (1.02% ÷ 0.1%) × 5% = 51%. By comparing the calculation with the historical data model, the effect of the moisture content of the reaction reagent on the 18 The impact of F-FDG synthesis reaction loss.

[0097] Preferably, the corresponding nucleophilic reaction temperature and nucleophilic reaction time are obtained during the nucleophilic reaction process in the synthesis link, and the nucleophilic reaction rate analysis is performed on the corresponding nucleophilic reaction process in the synthesis link based on the nucleophilic reaction temperature and the nucleophilic reaction time to obtain the corresponding reaction temperature and time. 18 F-FDG synthesis reaction rate;

[0098] In an embodiment of the present invention, a high-precision thermocouple temperature sensor is used to measure the nucleophilic reaction temperature in real time during a nucleophilic reaction in the synthesis process. The probe of the thermocouple temperature sensor is inserted into the reaction system to ensure that the reaction temperature can be accurately measured, and the temperature data is recorded every 10 seconds. At the same time, a timer in the automatic control system is used to record the nucleophilic reaction time, and the timing is triggered from the start of the reaction to the end of the reaction. For example, during a certain nucleophilic reaction, the thermocouple temperature sensor measures that the reaction temperature is 80°C at the beginning of the reaction, and gradually increases to 90°C and remains stable as the reaction proceeds. The reaction time from the beginning to the end is 120 minutes. Using the principle of chemical kinetics, according to the reaction temperature and time data, combined with the Arrhenius equation (where k is the reaction rate constant, A is the pre-exponential factor, Ea is the reaction activation energy, R is the gas constant, and T is the absolute temperature), and the corresponding reaction temperature and time are calculated. 18 The F-FDG synthesis reaction rate can be obtained by experimental determination or reference to relevant literature to obtain the pre-exponential factor A and reaction activation energy E of the nucleophilic reaction. a , convert the measured temperature data into absolute temperature and substitute it into the equation to calculate the reaction rate constant k at different times, and then obtain 18 The reaction rate of F-FDG synthesis is v = k × [reactant concentration], where k is the reaction rate constant at the corresponding reaction temperature and time. For example, if the reactant concentration before the reaction is 0.2 mol / L, and the reaction time is 120 minutes at 90°C, the reaction rate constant k is 0.0029 min. -1 , which means that the reaction rate at this temperature is a certain proportion of reactants reacted per minute, and then the corresponding reaction temperature of 90 ° C and the reaction time of 120 minutes are obtained. 18 The synthesis reaction rate of F-FDG is 0.0029 min -1 × 0.2mol / L = 0.00058mol / (L·min), and finally the corresponding reaction temperature and time are obtained. 18 F-FDG synthesis reaction rate.

[0099] Preferably, the corresponding raw material input amount and 18 F-FDG synthesis yield, and based on 18 The influence of F-FDG synthesis reaction loss and the corresponding reaction temperature and time 18 The F-FDG synthesis reaction rate is affected by the amount of raw materials input and 18 F-FDG synthesis yield 18 The F-FDG synthesis rate was quantified to obtain 18 The automated synthesis efficiency corresponding to F-FDG.

[0100] In the embodiment of the present invention, the amount of raw material input and the amount of 18 F-FDG synthesis yield, and based on 18 The influence of F-FDG synthesis reaction loss and the corresponding reaction temperature and time 18 The F-FDG synthesis reaction rate is affected by the amount of raw materials input and 18 F-FDG synthesis yield 18 Quantification of the F-FDG synthesis rate, given that 1000 mci of F-18 ions were delivered as raw material input, the final output 18The F-FDG activity is measured to be 500 mci. For example, under ideal conditions with no loss, the corresponding synthesis reaction loss is 0%, and the synthesis reaction rate under the corresponding reaction temperature and time is 0.00058 mol / (L·min). Assuming that the theoretical yield calculated according to the synthesis reaction rate and reaction time under ideal conditions (no loss) is 1000 mci, the actual synthesis rate is the actual yield ÷ (theoretical yield × (1-synthesis reaction loss)) × 100%, that is, 500 ÷ (1000 × (1-0%)) × 100% = 50%, and the final result is 18 The corresponding automated synthesis efficiency of F-FDG is 50% (calculated in mci).

[0101] Furthermore, the corresponding two water removal processes in the synthesis step are used to obtain the corresponding 18 The synthesis of F-FDG with water removal probability includes the following steps:

[0102] The corresponding drying water absorption reaction time and distillation water reaction time are obtained through the two corresponding water removal processes in the synthesis process;

[0103] In an embodiment of the present invention, during the two water removal processes in the synthesis step, for the desiccant water absorption process, a timer in the automated control system is started when the desiccant is put into use. When it is confirmed that the desiccant is saturated with water and no longer has a significant water absorption effect, the timer is stopped, and this period of time is recorded as the drying water absorption reaction duration. For example, the timer is started from the time the desiccant is added to the reaction system. After 30 minutes, the desiccant water absorption is determined to be complete by observing the color change of the desiccant (assuming that the color change of the desiccant indicates water absorption saturation) and the water content of the reaction system is no longer reduced using a Karl Fischer titrator. At this time, the drying water absorption reaction duration is recorded as 30 minutes. For the distillation water separation process, the timer in the automated control system is also started when the distillation apparatus is started. When the amount of water distilled reaches the expected amount or the water content at the discharge outlet is stabilized at the target low value as monitored by the online moisture sensor, the timer is stopped to obtain the distillation water reaction duration. For example, after 45 minutes after the distillation apparatus is started, the online moisture sensor shows that the water content at the discharge outlet is stabilized at 2%, reaching the expected low moisture content, and the distillation water reaction duration is recorded as 45 minutes.

[0104] Preferably, the drying water absorption process of the corresponding desiccant in the synthesis link is analyzed to obtain 18 F-FDG dry water absorption;

[0105] In the embodiment of the present invention, the moisture content of the desiccant before and after water absorption is detected by a high-precision Karl Fischer moisture meter during the desiccant water absorption process, and the moisture content of the desiccant is analyzed. 18The water absorption capacity of F-FDG during drying is as follows: before the desiccant is put into use, a desiccant sample of mass m1 gram is taken and its initial moisture content is measured using a Karl Fischer moisture meter, which is w1%. The initial moisture content is m1×w1% grams. After the desiccant has absorbed water, another desiccant sample of the same mass m1 gram is taken and its moisture content is w2%. The moisture content is m1×w2% grams. By calculating m1×(w2%-w1%), we can get 18 The water absorption capacity of F-FDG during drying is: assuming that 100g of desiccant sample is taken, the initial water content w1% = 0.1%, and the water content after water absorption w2% = 5%, then 18 The dry water absorption of F-FDG is 100×(5%-0.1%)=4.9 g, and the final product is 18 F-FDG drying water absorption.

[0106] Preferably, the corresponding distillation separation water process is obtained in the synthesis link 18 The amount of water migration from F-FDG distillation is calculated based on the drying water absorption reaction time and the distillation water reaction time for the two water removal processes. 18 F-FDG dry water absorption and 18 The F-FDG distillation water migration amount is synthesized and the water removal probability statistics are obtained to obtain the corresponding 18 F-FDG synthesis water removal probability.

[0107] In the embodiment of the present invention, during the process of water separation by distillation, the mass flow meter installed at the outlet of the distillation device is used to measure the mass of the distilled water flowing out, thereby obtaining 18 The amount of water migration during F-FDG distillation is, for example, that during the distillation process, the mass flow meter records that the mass of water flowing out from the beginning to the end of distillation is 50 grams, that is, 18 The amount of water migrated by F-FDG distillation is 50 g. At the same time, based on the drying water absorption reaction time t1 (such as 30 minutes) and the distillation water reaction time t2 (such as 45 minutes), combined with 18 F-FDG dry water absorption m 吸 (e.g. 4.9 g) and 18 F-FDG distillation water migration amount m 蒸 (e.g. 50g), calculate the probability of synthetic water removal, first calculate the total theoretical amount of water to be removed m 总理论 , assuming that it is 100 grams according to the reaction formula and chemical principles, the actual water removal amount m 实际 =m 吸 +m 蒸 =4.9+50=54.9 g, total reaction time 30+45=75 minutes, synthesis dehydration probability Assuming that the theoretical total reaction time is 90 minutes, P = 54.9 / 100×75 / 90×100%≈45.75%, and the corresponding 18F-FDG synthesis and water removal probability is finally obtained.

[0108] Furthermore, step S2 includes the following steps:

[0109] Step S21: Factors affecting fluoride ion yield and 18 Each factor in the F-FDG synthetic impact factor is combined with other factors to perform correlation measurement calculations to quantify the corresponding correlation coefficients between the two combinations of impact factors and form the corresponding impact factor correlation matrix to generate 18 F-FDG impact factor correlation matrix;

[0110] Step S22: Based on 18 The correlation matrix of F-FDG influencing factors on the influencing factors of fluoride ion yield and 18 The influence factors of F-FDG synthesis were constructed by the influence factor association grid to integrate the influence factors of fluoride ion yield and 18 Each impact factor in the F-FDG synthetic impact factor is used as a node, and according to 18 The correlation coefficient between the two influencing factor combinations in the F-FDG influencing factor correlation matrix determines the edges connecting them. If the correlation coefficient is greater than or equal to 0.85, the edges between the nodes are connected; if the correlation coefficient is less than 0.85, no processing is done and the corresponding 18 F-FDG impact factor association network;

[0111] Step S23: 18 The F-FDG impact factor association network is analyzed for node degree and betweenness centrality, in order to analyze the number of edges connected to the corresponding node as the corresponding degree of the node, and to analyze the betweenness centrality of the corresponding node to obtain the 18 The node degree and betweenness centrality corresponding to the F-FDG impact factor;

[0112] Step S24: Based on each 18 The node degree and betweenness centrality corresponding to the F-FDG impact factor have an impact on the fluoride ion yield factor and 18 Each factor in the F-FDG synthesis impact factor is divided into primary and secondary impacts. 18 If the node degree of the F-FDG impact factor exceeds 5 or the betweenness centrality exceeds 8, it is determined as the main impact factor, otherwise it is determined as the secondary impact factor to obtain 18 The main and secondary influencing factors corresponding to F-FDG efficiency;

[0113] Step S25: 18The main influencing factors and secondary influencing factors corresponding to F-FDG efficiency are determined by criterion weights to obtain the corresponding 18 F-FDG relative importance weight.

[0114] As an embodiment of the present invention, refer to Figure 3 As shown, Figure 1 Detailed step flow diagram of step S2 in the embodiment, step S2 includes the following steps:

[0115] Step S21: Factors affecting fluoride ion yield and 18 Each factor in the F-FDG synthetic impact factor is combined with other factors to perform correlation measurement calculations to quantify the corresponding correlation coefficients between the two combinations of impact factors and form the corresponding impact factor correlation matrix to generate 18 F-FDG impact factor correlation matrix;

[0116] In the embodiment of the present invention, the factors affecting the yield of fluoride ions and the 18 Each influencing factor in the F-FDG synthesis influencing factor is combined with each other for correlation measurement calculation. Assume that the fluoride ion yield influencing factor is stored in a pandas DataFrame object fluoride_factors_df, which contains columns such as the number of target shots and target water abundance. The 18F-FDG synthesis influencing factor is stored in 18F-FDG_synthesis_factors_df, which contains columns such as production preparation column volume and reagent content. The two DataFrames are merged into a new DataFrame all_factors_df, and the Pearson correlation coefficient between all factors is calculated by using the all_factors_df.corr() method. This method automatically performs correlation measurement calculations on each two factors. For example, the correlation coefficient between the number of target shots and the production preparation column volume is calculated to be 0.6, and the correlation coefficient between the target water abundance and the reagent content is 0.3. These correlation coefficients are organized into a two-dimensional matrix, and the corresponding influencing factor correlation matrix is constructed to generate the final result. 18 F-FDG impact factor correlation matrix, the rows and columns of the matrix correspond to each impact factor, and the matrix elements are the correlation coefficients between any two impact factor combinations.

[0117] Step S22: Based on 18 The correlation matrix of F-FDG influencing factors on the influencing factors of fluoride ion yield and 18 The influence factors of F-FDG synthesis were constructed by the influence factor association grid to integrate the influence factors of fluoride ion yield and 18Each impact factor in the F-FDG synthetic impact factor is used as a node, and according to 18 The correlation coefficient between the two influencing factor combinations in the F-FDG influencing factor correlation matrix determines the edges connecting them. If the correlation coefficient is greater than or equal to 0.85, the edges between the nodes are connected; if the correlation coefficient is less than 0.85, no processing is done and the corresponding 18 F-FDG impact factor association network;

[0118] In the embodiment of the present invention, the NetworkX library of Python is used to 18 The F-FDG impact factor correlation matrix is used to construct the impact factor correlation grid. First, an empty undirected graph object G = nx.Graph() is created to combine the fluoride ion yield impact factor and 18 Each factor in the F-FDG synthetic impact factor is added to the graph as a node. The corresponding operation process is as follows: G.add_nodes_from(all_factors_df.columns), traversing 18 F-FDG influencing factor correlation matrix, for each correlation coefficient value, determine whether it is greater than or equal to 0.85, if the correlation coefficient is greater than or equal to 0.85, such as the correlation coefficient between the number of target shots and a certain 18F-FDG synthesis influencing factor (assuming it is the automated synthesis rate) is 0.9, then use the G.add_edge('number of target shots','automated synthesis rate') method to connect the edges corresponding to these two nodes in the graph. If the correlation coefficient is less than 0.85, such as the correlation coefficient between the target water abundance and the production preparation column quantity is 0.5, then no processing is performed. After judging and processing all factor combinations, the corresponding edges are generated. 18 F-FDG impact factor association network, which intuitively shows the strong correlation between various impact factors.

[0119] Step S23: 18 The F-FDG impact factor association network is analyzed for node degree and betweenness centrality, in order to analyze the number of edges connected to the corresponding node as the corresponding degree of the node, and to analyze the betweenness centrality of the corresponding node to obtain the 18 The node degree and betweenness centrality corresponding to the F-FDG impact factor;

[0120] In the embodiment of the present invention, by continuing to use the NetworkX library 18The F-FDG impact factor association network is analyzed for node degree and betweenness centrality. For node degree analysis, the nx.degree(G) method is used. This method returns a dictionary containing each node and its corresponding degree (that is, the number of connected edges). For example, degree_dict = dict(nx.degree(G)), the degree of the target number node is 3, that is, the target number is connected to the other three impact factor nodes through edges. For betweenness centrality analysis, the nx.betweenness_centrality(G) method is used. This method calculates the betweenness centrality of each node. For example, betweenness_dict = nx.betweenness_centrality(G), the betweenness centrality of a certain impact factor node (assuming it is the automatic synthesis rate) is 9 (the actual calculation will vary according to the network situation). Through these calculations, we can finally get each 18 The node degree and betweenness centrality corresponding to the F-FDG impact factor.

[0121] Step S24: Based on each 18 The node degree and betweenness centrality corresponding to the F-FDG impact factor have an impact on the fluoride ion yield factor and 18 Each factor in the F-FDG synthesis impact factor is divided into primary and secondary impacts. 18 If the node degree of the F-FDG impact factor exceeds 5 or the betweenness centrality exceeds 8, it is determined as the main impact factor, otherwise it is determined as the secondary impact factor to obtain 18 The main and secondary influencing factors corresponding to F-FDG efficiency;

[0122] In the embodiment of the present invention, by 18 The node degree and betweenness centrality corresponding to the F-FDG impact factor have an impact on the fluoride ion yield factor and 18 Each influencing factor in the F-FDG synthetic influencing factor is divided into primary and secondary influences to traverse the dictionary of node degree and betweenness centrality. For each influencing factor node, it is judged whether its node degree exceeds 5 or whether its betweenness centrality exceeds 8. For example, if the node degree of the number of target shots node is 6, which exceeds 5, then the number of target shots is determined as the primary influencing factor. For another example, if the betweenness centrality of a certain influencing factor node (assuming it is a certain reaction tube quantity) is 5, which is less than 8, and the node degree is 3, which is less than 5, then the reaction tube quantity is determined as the secondary influencing factor. By judging all influencing factor nodes, we can get 18 The main and secondary influencing factors of F-FDG efficiency are used to clarify which factors are most important. 18 The impact of F-FDG efficiency is more critical.

[0123] Step S25: 18 The main influencing factors and secondary influencing factors corresponding to F-FDG efficiency are determined by criterion weights to obtain the corresponding 18 F-FDG relative importance weight.

[0124] In the embodiment of the present invention, the analytic hierarchy process (AHP) is used to determine 18 The criterion weights of the main influencing factors and secondary influencing factors corresponding to F-FDG efficiency are first constructed. For the main influencing factors, it is assumed that the main influencing factors include the number of target shots, the automatic synthesis rate, etc. According to expert experience or actual production data, the relative importance of these factors is compared two by two. For example, the number of target shots is slightly more important than the automatic synthesis rate. The corresponding element in the judgment matrix is set to 3 (1-9 scaling method, 1 means equal importance, 9 means absolute importance). After constructing the judgment matrix, the eigenvector method is used to calculate the weight vector. By calculating the maximum eigenvalue of the judgment matrix and its corresponding eigenvector, the eigenvector is normalized to obtain the weights corresponding to each main influencing factor. For the secondary influencing factors, the judgment matrix is constructed and the weights are calculated according to the above method. Finally, the corresponding weights of each influencing factor are obtained. 18 The relative importance weights of F-FDG can be used for subsequent 18 Optimization of F-FDG efficiency calculation model and development of early warning method.

[0125] Furthermore, step S25 includes the following steps:

[0126] Step S251: By 18 Obtain the corresponding F-FDG historical production process 18 F-FDG historical production efficiency and use it as the target layer;

[0127] In the embodiment of the present invention, by 18 In the F-FDG historical production data storage system, query and extract the corresponding data of each production process in the past period of time (such as the past year). 18 The data of F-FDG production and production time are calculated by the formula " 18 F-FDG historical production efficiency = 18 The total F-FDG production is calculated as "total production times ÷ total production times" 18 The historical average production efficiency of F-FDG, for example, a total of 174,000 mci was produced in the past year. 18 F-FDG, the total production time is 500 hours, then 18 The historical production efficiency of F-FDG is 174,000 ÷ 500 = 348 mci / hour. This efficiency value is used as the target layer for subsequent correlation analysis with influencing factors to determine the degree of influence of each factor on production efficiency.

[0128] Step S252: 18 The main influencing factors corresponding to F-FDG efficiency are used as the criterion layer, and 18 The secondary influencing factors corresponding to F-FDG efficiency are used as the indicator layer; the target layer, criterion layer and indicator layer are constructed into the corresponding expert evaluation system architecture according to the hierarchical structure;

[0129] In the embodiment of the present invention, by using the previously obtained 18 The main influencing factors corresponding to F-FDG efficiency (such as the number of targets, automated synthesis rate, etc.) are used as the criterion layer, and the secondary influencing factors (such as the amount of a certain reaction tube, the content of a certain component in the reagent, etc.) are used as the indicator layer. By using professional project management tools (such as Microsoft Project) or online hierarchical analysis tool websites (such as yaahp), the expert evaluation system architecture is constructed according to the hierarchical structure. In the tool, first create the target layer and enter 18 The relevant description and values of F-FDG historical production efficiency are then added under the target layer. The criterion layer node is added as a node. Then, the indicator layer nodes are added under the criterion layer node respectively, and the corresponding secondary influencing factors are associated with each major influencing factor to form a complete hierarchical system, providing a clear framework for expert scoring.

[0130] Step S253: Expert scoring is performed based on the correlation coefficients between each influencing factor corresponding to the criterion layer and the indicator layer within the expert evaluation system architecture, and a corresponding scoring judgment matrix is constructed based on the scoring results of each influencing factor, where the elements in the scoring judgment matrix represent the score ratio relationship between each influencing factor;

[0131] In the embodiment of the present invention, a team of experts in related fields (such as nuclear engineers, process researchers, etc.) is organized to score the correlation coefficients between each influencing factor corresponding to the criterion layer and the indicator layer in the expert evaluation system architecture, and a detailed scoring table is prepared. The table lists all the influencing factor combinations of the criterion layer and the indicator layer. For each group of influencing factors, the experts score the corresponding factors based on their own experience and understanding. 18 The in-depth understanding of the F-FDG production process was scored on a 1-9 scale. For example, for the influencing factors of the number of targets and the automated synthesis rate, if the expert believed that the number of targets had a significant impact on the production process, 18The impact of F-FDG production efficiency is slightly more important than the automated synthesis rate. Fill in 3 in the corresponding position in the scoring table. After collecting the scoring results of all experts, they are organized into a scoring judgment matrix. For example, assuming that there are 3 main influencing factors at the criterion level and 5 secondary influencing factors at the indicator level, the scoring judgment matrix is a (3+5)×(3+5) matrix. The matrix elements represent the score ratio relationship between each influencing factor. For example, the element in the i-th row and j-th column represents the importance score of the i-th influencing factor relative to the j-th influencing factor.

[0132] Step S254: Obtain the corresponding maximum eigenvalue and its corresponding eigenvector through the scoring judgment matrix, and perform consistency check calculation based on the maximum eigenvalue and its corresponding eigenvector to calculate the corresponding consistency index and random consistency index, and calculate the ratio of the consistency index and the random consistency index to obtain the corresponding consistency ratio;

[0133] In an embodiment of the present invention, the scoring judgment matrix is calculated by using the numpy library of Python, and the eigenvalues and eigenvectors of the scoring judgment matrix are calculated by the numpy.linalg.eig() function, and the maximum eigenvalue and its corresponding eigenvector are extracted. For example, assuming that the calculated maximum eigenvalue is 8.5, the corresponding eigenvectors are [0.3, 0.2, 0.1, 0.2, 0.1, 0.05, 0.05, 0.0], and then the consistency index (CI) is calculated. The formula is "CI = (maximum eigenvalue - matrix order) ÷ (matrix order - 1)", assuming that the matrix order is 8, then CI = (8.5-8) ÷ (8-1) ≈ 0.071, and the random consistency index (RI) can be obtained by referring to a pre-defined random consistency index table. For an 8-order matrix, the RI value is assumed to be 1.41. The consistency index CI is calculated by ratioing the random consistency index RI to obtain the consistency ratio (CR), that is, CR = CI ÷ RI ≈ 0.071 ÷ 1.41 ≈ 0.05, and finally the corresponding consistency ratio is obtained.

[0134] Step S255: If the consistency ratio is greater than or equal to 0.1, the corresponding scoring result is determined as the corresponding relative importance weight to obtain the corresponding 18 F-FDG relative importance weight; if the consistency ratio is less than 0.1, it will return to the expert scoring until the scoring judgment matrix passes the consistency test.

[0135] In the embodiment of the present invention, by comparing the calculated consistency ratio with 0.1, if the consistency ratio is greater than or equal to 0.1, such as the previously calculated consistency ratio is 0.05 less than 0.1, the scoring result of the scoring judgment matrix is discarded, the experts are reorganized to score, and the new scoring results are organized into a scoring judgment matrix again, and the calculation process of step S254 is repeated until the consistency ratio is greater than or equal to 0.1. When the consistency ratio meets the requirements, the eigenvector corresponding to the scoring judgment matrix at this time is normalized to obtain the corresponding eigenvectors of each influencing factor. Weights, for example, after normalizing the eigenvector [0.3, 0.2, 0.1, 0.2, 0.1, 0.05, 0.05, 0.0], we get [0.3÷1.0, 0.2÷1.0, 0.1÷1.0, 0.2÷1.0, 0.1÷1.0, 0.05÷1.0, 0.05÷1.0, 0.0÷1.0] = [0.3, 0.2, 0.1, 0.2, 0.1, 0.05, 0.05, 0.0]. These weight values are the relative importance weights of 18F-FDG corresponding to each influencing factor, which can be used for subsequent 18 Optimization of F-FDG efficiency calculation model and development of early warning method.

[0136] Furthermore, step S3 includes the following steps:

[0137] Step S31: Based on the corresponding 18 The relative importance weight of F-FDG was combined with multiple linear regression to analyze the 18 The main influencing factors and secondary influencing factors corresponding to F-FDG efficiency are used to train the efficiency calculation model to 18 The relative importance weight of F-FDG is calculated with the main impact factor and the secondary impact factor as input. 18 The F-FDG efficiency is used as output, generating 18 Calculate the initial model for F-FDG efficiency and output the corresponding prediction 18 F-FDG efficiency;

[0138] In this embodiment of the present invention, multiple linear regression is performed using the scikit-learn library of Python. 18 The efficiency calculation model training is carried out for the main and secondary influencing factors corresponding to the F-FDG efficiency, and the corresponding 18The relative importance weights of F-FDG are organized into a weight vector, for example, the weight vector is [0.3, 0.2, 0.1, 0.2, 0.1, 0.05, 0.05, 0.0], and the data of the main influencing factors (such as the number of target shots, the automated synthesis rate, etc.) and the secondary influencing factors (such as the amount of a certain reaction tube, the content of a certain component in the reagent, etc.) are organized into a feature matrix. Each row represents the data of a production process, and each column corresponds to an influencing factor. For example, a row of data in the feature matrix is [the number of target shots is 2, the automated synthesis rate is 80%, the anhydrous acetonitrile is 1.5 ml, the trifluoromannose content is 5 g / L...], and a multivariate linear regression model object is created by using the LinearRegression() function. The weights of each influencing factor in the weight vector are weighted calculated with the corresponding influencing factors in the feature matrix. 18 F-FDG efficiency is used as an output label, for example, the prediction is obtained by model calculation 18 The F-FDG efficiency is 1200 mci, generating 18 Calculate the initial model for F-FDG efficiency and output the corresponding prediction 18 F-FDG efficiency.

[0139] Step S32: Predict 18 F-FDG efficiency compared to historical standards 18 The F-FDG efficiency was compared and analyzed to obtain the deviation between the predicted efficiency and the standard efficiency;

[0140] In the embodiment of the present invention, by 18 Extraction of historical standards from the F-FDG historical production data storage system 18 F-FDG efficiency data, which were determined to be in compliance with the standards during the past production process 18 F-FDG production efficiency values, assuming historical standards 18 The F-FDG efficiency is 1000mci, and the predicted output is 18 F-FDG efficiency (e.g., 1200 mci) compared to historical standards 18 The F-FDG efficiency was compared and analyzed, and the deviation between the predicted efficiency and the standard efficiency was calculated. The formula was "deviation = | predicted 18 F-FDG efficiency - historical standard 18 F-FDG efficiency |÷historical standard 18 F-FDG efficiency × 100%". In this example, deviation = |1200-1000|÷2×100%=10%. The deviation value between the predicted efficiency and the standard efficiency is obtained, which is used to determine whether the model needs to be optimized.

[0141] Step S33: Compare and judge the deviation between the predicted efficiency and the standard efficiency according to the preset deviation threshold of 1%. If the deviation between the predicted efficiency and the standard efficiency is less than or equal to the deviation threshold of 1%, the model will not be optimized; if the deviation between the predicted efficiency and the standard efficiency is greater than the deviation threshold of 1%, the model will be re-optimized. 18 The corresponding initial model for F-FDG efficiency calculation 18 The relative importance weight of F-FDG is determined by the criterion weight, and the re-determined 18 F-FDG relative importance weight iterative optimization corresponding to 18 F-FDG efficiency calculation initial model to generate 18 F-FDG efficiency calculation model.

[0142] In the embodiment of the present invention, the deviation between the predicted efficiency and the standard efficiency (such as 10%) is compared with the preset deviation threshold of 1%. Since 10% is greater than 1%, it is necessary to re-calibrate the 18 The corresponding initial model for F-FDG efficiency calculation 18 The relative importance weight of F-FDG is used to determine the criterion weight, and relevant experts in related fields (such as nuclear engineers, process researchers, etc.) are reorganized to score the influencing factors of the criterion layer and the indicator layer again according to the method of step S25, and a scoring judgment matrix is constructed. The maximum eigenvalue and its corresponding eigenvector are calculated, and consistency tests are performed to obtain the re-determined 18 The relative importance weight of F-FDG was applied to the multiple linear regression model. 18 The initial model of F-FDG efficiency calculation is iteratively optimized. For example, the new weight vector is [0.25, 0.22, 0.12, 0.2, 0.1, 0.06, 0.04, 0.01]. The weighted calculation and model training are performed again to generate a new 18 F-FDG efficiency calculation model to improve the model 18 The accuracy of F-FDG efficiency prediction is finally generated 18 F-FDG efficiency calculation model.

[0143] Furthermore, step S4 includes the following steps:

[0144] Step S41: Get the current 18 The main and secondary influencing factors corresponding to the F-FDG production process;

[0145] In the embodiment of the present invention, by 18During the F-FDG production process, data is collected in real time through various sensors installed on the production equipment and the automated control system to obtain the corresponding major influencing factors and minor influencing factors. For major influencing factors, such as the number of target positions, mechanical sensors with position encoding functions are used to monitor the usage of single and double target positions in the targeting process in real time, and their number is counted as the number of target positions; the automated synthesis rate is directly read from the control system of the automated synthesis equipment, and for minor influencing factors, such as the volume of a certain reaction tube, capacitive liquid level sensors and weighing sensors are installed on reaction tubes such as QMA columns, long tC18 columns and combined columns to measure the volume and mass of the material in the column in real time to obtain the volume of the reaction tube; the content of a certain component in the reagent is obtained by detecting the production reagent through a spectral analysis sensor. For example, for 1ml of trifluoromannose / anhydrous acetonitrile reagent, the concentrations of trifluoromannose and anhydrous acetonitrile are determined by analyzing its absorption of light of a specific wavelength. These real-time collected major and minor influencing factor data are organized and stored for subsequent use. 18 F-FDG efficiency prediction.

[0146] Step S42: The current 18 The main and secondary influencing factors corresponding to the F-FDG production process are input into the trained 18 F-FDG efficiency calculation model 18 F-FDG efficiency prediction is to compare the corresponding main influencing factors and secondary influencing factors with 18 The weights corresponding to each influencing factor in the F-FDG efficiency calculation model are weighted to generate the current corresponding 18 F-FDG efficiency;

[0147] In the embodiment of the present invention, the current 18 The main and secondary influencing factors corresponding to the F-FDG production process are input into the trained 18 F-FDG efficiency calculation model 18 F-FDG efficiency prediction, assuming the trained 18 The F-FDG efficiency calculation model is stored in a Python pickle file. The pickle.load() function is used to load the model. The main influencing factors (such as the number of targets is 2, the automated synthesis rate is 75%, ...) and the secondary influencing factors (such as the volume of a reaction tube is 1 ml, the content of a certain component in the reagent is 6 g / L, ...) are organized into feature vectors. The weights corresponding to the influencing factors in the model (such as the weight vector is [0.25, 0.22, 0.12, 0.2, 0.1, 0.06, 0.04, 0.01]) are weighted and calculated. For example, by calculating (2×0.25+75%×0.22+1×0.12+6×0.2+...), the current corresponding18 F-FDG efficiency, assuming the current 18 The F-FDG dose was 1000 mcg.

[0148] Step S43: by presetting the corresponding first-level warning threshold of 85% and the second-level warning threshold of 70%;

[0149] In the embodiment of the present invention, by 18 In the configuration file of the F-FDG production early warning system, the corresponding first-level warning threshold is clearly set to 85% and the second-level warning threshold is 70%. These thresholds are determined after multiple analyses and tests based on historical production data and production process requirements. For example, through statistical analysis of a large amount of past production data, it is found that when 18 When the F-FDG efficiency is lower than 85%, although production can still be maintained, there are some potential factors that affect the production efficiency; 18 When the F-FDG efficiency is lower than 70%, 18 The F-FDG production will drop significantly, seriously affecting the production efficiency. These thresholds are stored in the system database for subsequent use in early warning judgments.

[0150] Step S44: Based on the preset first-level warning threshold of 85% and second-level warning threshold of 70%, the current corresponding 18 F-FDG efficiency 18 F-FDG production warning, if the current corresponding 18 When the F-FDG efficiency is lower than the first-level warning threshold of 85%, a corresponding minor warning signal will be issued to indicate that there is an impact. 18 Potential factors for F-FDG efficiency and alert production personnel to conduct further inspection and analysis in time; if the current corresponding 18 When the F-FDG efficiency is lower than the secondary warning threshold of 70%, a corresponding serious warning signal will be issued to indicate the occurrence of 18 The system will warn the production personnel to stop production immediately to avoid producing defective products, check whether the pipeline transmission system is leaking or blocked, and activate the preset emergency plan for radioactive material transmission pipeline leakage when necessary. 18 The corresponding spare targets and spare synthesis modules in the F-FDG production process are supplied first, and when the radiation dose in the target room drops to an acceptable range for personnel, the accelerator is inspected, maintained and troubleshooted to generate the corresponding 18 F-FDG production early warning adjustment measures.

[0151] In the embodiment of the present invention, the current corresponding 18 F-FDG efficiency18 F-FDG production warning, in the production warning system, set up a real-time data comparison module to continuously calculate the current 18 The F-FDG efficiency is compared with the two warning thresholds. If the current corresponding 18 If the F-FDG efficiency is 80%, which is lower than the first-level warning threshold of 85%, the system will immediately issue a slight warning signal, such as a low-frequency flashing light and a soft prompt sound through the sound and light alarm in the production workshop, and a prompt box will pop up on the production monitoring screen to inform that there is an impact. 18 The potential factors of F-FDG efficiency prompt production personnel to conduct further inspection and analysis of production parameters and equipment operating status in a timely manner. 18 The F-FDG efficiency drops to 60%, which is lower than the second-level warning threshold of 70%. The system issues a serious warning signal, the sound and light alarm emits a high-frequency strong light and a sharp alarm sound, and a striking red warning message is displayed on the production monitoring screen. 18 If there is a warning of a sharp drop in F-FDG production, the production staff will immediately stop production according to the established process and use pipeline detection equipment, such as pipeline leak detectors, to check whether the pipeline transmission system is leaking or blocked. If a pipeline leak is detected, the preset radioactive material transmission pipeline leak emergency plan will be immediately activated, such as activating the leak control device and evacuating nearby personnel. At the same time, quickly switch to 18 The corresponding spare targets and spare synthesis modules in the F-FDG production process are given priority in supply and use to ensure that the products can be supplied to patients. When the radiation dose in the target room is reduced to an acceptable range for personnel through ventilation and other measures, professional maintenance personnel are arranged to conduct a comprehensive inspection and maintenance of the accelerator, find out the cause of the fault, record the entire process, and finally generate the corresponding 18 F-FDG production early warning adjustment measures report.

[0152] Furthermore, the present invention also provides a method based on 18 The F-FDG efficiency calculation model is used to perform the above-mentioned early warning system based on 18 The early warning method of F-FDG efficiency calculation model is based on 18 The early warning system of the F-FDG efficiency calculation model includes:

[0153] 18 F-FDG impact factor analysis module is used to analyze the 18 The historical production process of F-FDG was used to obtain the factors affecting the yield of fluoride ions produced by cyclotron nuclear reactions and 18 The factors affecting F-FDG synthesis include the abundance of oxygen-18 water, the amount of oxygen-18 water used, the amount of oxygen-18 water consumed during transmission in a closed target, the proton beam target position, the proton beam energy, the proton beam current, and the irradiation time. 18The factors affecting F-FDG synthesis include the amount of precursors prepared for production, reagent content, fluoride ion pipeline transmission consumption, QMA full load, precursor purity, catalyst activity, reaction temperature, reaction time, 18 F-FDG synthesis channel number, automated synthesis efficiency and 18 F-FDG production;

[0154] Impact factor weight determination module is used to determine the impact factor weight based on 18 The factors affecting the historical production efficiency of F-FDG on the yield of fluoride ions and 18 The F-FDG synthesis influencing factors were divided into primary and secondary influences to obtain 18 The main and secondary influencing factors corresponding to F-FDG efficiency; 18 The main influencing factors and secondary influencing factors corresponding to F-FDG efficiency are determined by criterion weights to obtain the corresponding 18 F-FDG relative importance weight;

[0155] Efficiency calculation model training module is used to calculate the efficiency of the model based on the corresponding 18 F-FDG relative importance weight 18 The main influencing factors and secondary influencing factors corresponding to F-FDG efficiency are iteratively trained to generate the efficiency calculation model. 18 F-FDG efficiency calculation model;

[0156] 18 F-FDG production warning module is used to obtain current 18 The main and secondary influencing factors corresponding to the F-FDG production process are input into the trained 18 F-FDG efficiency calculation model 18 F-FDG efficiency prediction to generate the current corresponding 18 F-FDG efficiency; based on the preset first-level warning threshold and second-level warning threshold, the current corresponding 18 F-FDG efficiency 18 F-FDG production warning to generate corresponding 18 F-FDG production early warning adjustment measures.

[0157] The present invention is therefore intended to be illustrative and non-restrictive in all respects, with the scope of the invention being defined by the appended claims rather than the foregoing description, and all changes that come within the meaning and range of equivalents of the application documents are intended to be embraced therein.

[0158] The foregoing description is intended only to provide specific embodiments of the present invention, which will enable those skilled in the art to understand and implement the present invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention is not intended to be limited to the embodiments shown herein, but is to be construed in the widest possible manner consistent with the principles and novel features disclosed herein.

Claims

1. A method based on 18 The early warning method of the F-FDG efficiency calculation model is characterized by: The following steps are involved: Step S1: By 18 The historical production process of F-FDG was used to obtain the factors affecting the yield of fluoride ions produced by cyclotron nuclear reactions and 18 The factors affecting F-FDG synthesis include the abundance of oxygen-18 water, the amount of oxygen-18 water used, the amount of oxygen-18 water consumed during transmission in a closed target, the proton beam target position, the proton beam energy, the proton beam current, and the irradiation time. 18 The factors affecting F-FDG synthesis include the amount of precursors prepared for production, reagent content, fluoride ion pipeline transmission consumption, QMA full load, precursor purity, catalyst activity, reaction temperature, reaction time, 18 F-FDG synthesis channel number, automated synthesis efficiency and 18 F-FDG production; Step S2: Factors affecting fluoride ion yield and 18 The F-FDG synthesis influencing factors were divided into primary and secondary influences to obtain 18 The main and secondary influencing factors corresponding to F-FDG efficiency; right 18 The main influencing factors and secondary influencing factors corresponding to F-FDG efficiency are determined by criterion weights to obtain the corresponding 18 F-FDG relative importance weight; Step S3: Based on the corresponding 18 F-FDG relative importance weight 18 The main influencing factors and secondary influencing factors corresponding to F-FDG efficiency are iteratively trained to generate the efficiency calculation model. 18 F-FDG efficiency calculation model; Step S4: Get the current 18 The main and secondary influencing factors corresponding to the F-FDG production process are input into the trained 18 F-FDG efficiency calculation model 18 F-FDG efficiency prediction to generate the current corresponding 18 F-FDG efficiency; Based on the preset first-level warning threshold and second-level warning threshold, the current corresponding 18 F-FDG efficiency 18 F-FDG production warning to generate corresponding 18 F-FDG production early warning adjustment measures.

2. The method according to claim 1 18 The early warning method of the F-FDG efficiency calculation model is characterized by: Step S1 includes the following steps: Step S11: Using a mass spectrometer to obtain the corresponding abundance of oxygen-18 water produced by the cyclotron nuclear reaction, and using a flow sensor and a liquid level sensor to monitor the corresponding flow and storage amount of oxygen-18 water in the target practice process in real time, so as to calculate the corresponding oxygen-18 water consumption based on the corresponding flow and storage amount of oxygen-18 water; and by monitoring each target transfer stop point or target transfer position on the target transfer track during the target practice process, obtain the corresponding oxygen-18 water transmission consumption in the closed target; Step S12: using a cyclotron to monitor and obtain the number of single or double target positions corresponding to the proton beam in the target practice phase to obtain the proton beam target position; using a radiation monitoring sensor to accurately measure the energy and flow rate corresponding to the proton beam in the target practice phase and the irradiation duration to obtain the proton beam energy, proton beam current, and irradiation time; using the corresponding oxygen-18 water abundance, oxygen-18 water usage, oxygen-18 water transmission consumption in the closed target, proton beam target position, proton beam energy, proton beam current, and irradiation time as factors affecting fluoride ion yield; Step S13: By 18 During the historical production of F-FDG, capacitive level sensors and weighing sensors were used to accurately measure the volume and mass of materials in the reaction tubes to obtain the amount of material required for production. Spectral analysis sensors were used to quickly and accurately detect the concentration and composition of production reagents to determine the reagent content, including trifluoromannose, anhydrous acetonitrile, K2.2.2 / K2CO3 eluent, and NaOH. Step S14: Obtain the corresponding fluoride ion dosage after the nuclear reaction and the fluoride ion dosage after passing through the QMA column through the corresponding fluoride ion pipeline transmission process in the synthesis link to obtain the fluoride ion pipeline transmission consumption and the QMA full load; obtain the precursor purity, catalyst activity, reaction temperature, reaction time, 18 The number of F-FDG synthesis channels, including FDG1 and FDG2, is used to obtain the corresponding automated synthesis efficiency through two water removal and one nucleophilic reaction processes in the synthesis process. 18 F-FDG production; the corresponding production preparation precursor amount, reagent content, fluoride ion pipeline transmission consumption, QMA full load, precursor purity, catalyst activity, reaction temperature, reaction time, 18 F-FDG synthesis channel number, automated synthesis efficiency and 18 F-FDG production as 18 Factors affecting F-FDG synthesis.

3. The method according to claim 2 18 The early warning method of the F-FDG efficiency calculation model is characterized by: The reaction tube described in step S13 includes an activated QMA column, a long tC18 column and a combined column, wherein the combined column includes an IC-H column, an AL2O3 column and a tC18 column.

4. The method according to claim 2 18 The early warning method of the F-FDG efficiency calculation model is characterized by: The step S14 of obtaining the corresponding automated synthesis efficiency through two corresponding water removal processes and one corresponding nucleophilic reaction process in the synthesis link includes: The corresponding water removal rate is obtained by performing two water removal processes on the reaction reagents in the synthesis process, and the water content is obtained by using a desiccant to absorb water and distill to remove water; The moisture content is statistically analyzed based on the water removal rate to obtain the moisture residual 18 Residual water content before F-FDG synthesis reaction; Based on the residual water content, the corresponding nucleophilic reaction process in the synthesis link was analyzed. 18 F-FDG synthesis loss assessment to obtain the effect of reaction reagent water content on 18 F-FDG synthesis reaction affects the loss; The corresponding nucleophilic reaction temperature and nucleophilic reaction time are obtained during the nucleophilic reaction process in the synthesis link, and the nucleophilic reaction rate analysis is performed on the corresponding nucleophilic reaction process in the synthesis link based on the nucleophilic reaction temperature and nucleophilic reaction time to obtain the corresponding nucleophilic reaction rate at the corresponding reaction temperature and time. 18 F-FDG synthesis reaction rate; Obtain the corresponding raw material input amount and 18 F-FDG synthesis yield, and based on 18 The influence of F-FDG synthesis reaction loss and the corresponding reaction temperature and time 18 The F-FDG synthesis reaction rate is affected by the amount of raw materials input and 18 F-FDG synthesis yield 18 The F-FDG synthesis rate was quantified to obtain 18 The automated synthesis efficiency corresponding to F-FDG.

5. The method according to claim 4 18 The early warning method of the F-FDG efficiency calculation model is characterized by: The method of obtaining the corresponding water content of the reaction reagent by performing two corresponding water removal processes in the synthesis step includes the following steps: The corresponding drying water absorption reaction time and distillation water reaction time are obtained through the two corresponding water removal processes in the synthesis process; The drying water absorption capacity of the corresponding desiccant in the synthesis process is analyzed to obtain the drying water absorption capacity of the reaction reagent; The corresponding distillation water migration amount of the reaction reagent is obtained by the corresponding distillation water separation process in the synthesis link, and the reaction reagent drying water absorption amount and the reaction reagent distillation water migration amount corresponding to the two dehydration processes are synthesized based on the drying water absorption reaction time and the distillation water reaction time to obtain the corresponding reaction reagent moisture content.

6. The method according to claim 1 18 The early warning method of the F-FDG efficiency calculation model is characterized by: Step S2 includes the following steps: Step S21: Factors affecting fluoride ion yield and 18 Each factor in the F-FDG synthetic impact factor is combined with other factors to perform correlation measurement calculations to quantify the corresponding correlation coefficients between the two combinations of impact factors and form the corresponding impact factor correlation matrix to generate 18 F-FDG impact factor correlation matrix; Step S22: Based on 18 The correlation matrix of F-FDG influencing factors on the influencing factors of fluoride ion yield and 18 The influence factors of F-FDG synthesis were constructed by the influence factor association grid to integrate the influence factors of fluoride ion yield and 18 Each impact factor in the F-FDG synthetic impact factor is used as a node, and according to 18 The correlation coefficient between the two influencing factor combinations in the F-FDG influencing factor correlation matrix determines the edges connecting them. If the correlation coefficient is greater than or equal to 0.85, the edges between the nodes are connected; if the correlation coefficient is less than 0.85, no processing is done and the corresponding 18 F-FDG impact factor association network; Step S23: 18 The F-FDG impact factor association network is analyzed for node degree and betweenness centrality, in order to analyze the number of edges connected to the corresponding node as the corresponding degree of the node, and to analyze the betweenness centrality of the corresponding node to obtain the 18 The node degree and betweenness centrality corresponding to the F-FDG impact factor; Step S24: Based on each 18 The node degree and betweenness centrality corresponding to the F-FDG impact factor have an impact on the fluoride ion yield factor and 18 Each factor in the F-FDG synthesis impact factor is divided into primary and secondary impacts. 18 If the node degree of the F-FDG impact factor exceeds 5 or the betweenness centrality exceeds 8, it is determined as the main impact factor, otherwise it is determined as the secondary impact factor to obtain 18 The main and secondary influencing factors corresponding to F-FDG efficiency; Step S25: 18 The main influencing factors and secondary influencing factors corresponding to F-FDG efficiency are determined by criterion weights to obtain the corresponding 18 F-FDG relative importance weight.

7. The method according to claim 6 18 The early warning method of the F-FDG efficiency calculation model is characterized by: Step S25 includes the following steps: Step S251: By 18 Obtain the corresponding F-FDG historical production process 18 F-FDG production efficiency and use it as the target layer; Step S252: 18 The main influencing factors corresponding to F-FDG production efficiency are used as the criterion layer, and 18 The secondary influencing factors corresponding to F-FDG production efficiency are used as the indicator layer; the target layer, criterion layer and indicator layer are constructed into the corresponding expert evaluation system architecture according to the hierarchical structure; Step S253: Expert scoring is performed based on the correlation coefficients between each influencing factor corresponding to the criterion layer and the indicator layer within the expert evaluation system architecture, and a corresponding scoring judgment matrix is constructed based on the scoring results of each influencing factor, where the elements in the scoring judgment matrix represent the score ratio relationship between each influencing factor; Step S254: Obtain the corresponding maximum eigenvalue and its corresponding eigenvector through the scoring judgment matrix, and perform consistency check calculation based on the maximum eigenvalue and its corresponding eigenvector to calculate the corresponding consistency index and random consistency index, and calculate the ratio of the consistency index and the random consistency index to obtain the corresponding consistency ratio; Step S255: If the consistency ratio is greater than or equal to 0.1, the corresponding scoring result is determined as the corresponding relative importance weight to obtain the corresponding 18 F-FDG relative importance weight; if the consistency ratio is less than 0.1, it will return to the expert scoring until the scoring judgment matrix passes the consistency test.

8. The method according to claim 6 18 The early warning method of the F-FDG efficiency calculation model is characterized by: Step S3 includes the following steps: Step S31: Based on the corresponding 18 The relative importance weight of F-FDG was combined with multiple linear regression to analyze the 18 The main influencing factors and secondary influencing factors corresponding to F-FDG efficiency are used to train the efficiency calculation model to 18 The relative importance weight of F-FDG is calculated with the main impact factor and the secondary impact factor as input. 18 The F-FDG efficiency is used as output, generating 18 Calculate the initial model for F-FDG efficiency and output the corresponding prediction 18 F-FDG production efficiency; Step S32: Predict 18 F-FDG efficiency compared to historical standards 18 The F-FDG efficiency was compared and analyzed to obtain the deviation between the predicted efficiency and the standard efficiency; Step S33: Compare and judge the deviation between the predicted efficiency and the standard efficiency according to the preset deviation threshold of 1%. If the deviation between the predicted efficiency and the standard efficiency is less than or equal to the deviation threshold of 1%, the model will not be optimized; if the deviation between the predicted efficiency and the standard efficiency is greater than the deviation threshold of 1%, the model will be re-optimized. 18 The corresponding initial model for F-FDG efficiency calculation 18 The relative importance weight of F-FDG is determined by the criterion weight, and the re-determined 18 F-FDG relative importance weight iterative optimization corresponding to 18 F-FDG efficiency calculation initial model to generate 18 F-FDG efficiency calculation model.

9. The method according to claim 1 18 The early warning method of the F-FDG efficiency calculation model is characterized by: Step S4 includes the following steps: Step S41: Get the current 18 The main and secondary influencing factors corresponding to the F-FDG production process; Step S42: The current 18 The main and secondary influencing factors corresponding to the F-FDG production process are input into the trained 18 F-FDG efficiency calculation model 18 F-FDG efficiency prediction is to compare the corresponding main influencing factors and secondary influencing factors with 18 The weights corresponding to each influencing factor in the F-FDG efficiency calculation model are weighted to generate the current corresponding 18 F-FDG efficiency; Step S43: by presetting the corresponding first-level warning threshold of 85% and the second-level warning threshold of 70%; Step S44: Based on the preset first-level warning threshold of 85% and second-level warning threshold of 70%, the current corresponding 18 F-FDG efficiency 18 F-FDG production warning, if the current corresponding 18 When the F-FDG efficiency is lower than the first-level warning threshold of 85%, a corresponding minor warning signal will be issued to indicate that there is an impact. 18 Potential factors for F-FDG efficiency and alert production personnel to conduct further inspection and analysis in time; if the current corresponding 18 When the F-FDG efficiency is lower than the secondary warning threshold of 70%, a corresponding serious warning signal will be issued to indicate the occurrence of 18 The system will warn the production personnel to stop production immediately to avoid producing defective products, check whether the pipeline transmission system is leaking or blocked, and activate the preset emergency plan for radioactive material transmission pipeline leakage when necessary. 18 The corresponding spare targets and spare synthesis modules in the F-FDG production process are supplied first, and when the radiation dose in the target shooting room drops to an acceptable range for personnel, the accelerator is inspected, maintained and troubleshooted to generate the corresponding 18 F-FDG production early warning adjustment measures.

10. A method based on 18 The early warning system of the F-FDG efficiency calculation model is characterized by: For executing the method according to claim 1 18 The early warning method of F-FDG efficiency calculation model is based on 18 The early warning system of the F-FDG efficiency calculation model includes: 18 F-FDG impact factor analysis module is used to analyze the 18 The historical production process of F-FDG was used to obtain the factors affecting the yield of fluoride ions produced by cyclotron nuclear reactions and 18 The factors affecting F-FDG synthesis include the abundance of oxygen-18 water, the amount of oxygen-18 water used, the amount of oxygen-18 water consumed during transmission in a closed target, the proton beam target position, the proton beam energy, the proton beam current, and the irradiation time. 18 The factors affecting F-FDG synthesis include the amount of precursors prepared for production, reagent content, fluoride ion pipeline transmission consumption, QMA full load, precursor purity, catalyst activity, reaction temperature, reaction time, 18 F-FDG synthesis channel number, automated synthesis efficiency and 18 F-FDG production; Impact factor weight determination module is used to determine the impact factor weight based on 18 The factors affecting the historical production efficiency of F-FDG on the yield of fluoride ions and 18 The F-FDG synthesis influencing factors were divided into primary and secondary influences to obtain 18 The main and secondary influencing factors corresponding to F-FDG efficiency; 18 The main influencing factors and secondary influencing factors corresponding to F-FDG efficiency are determined by criterion weights to obtain the corresponding 18 F-FDG relative importance weight; Efficiency calculation model training module is used to calculate the efficiency of the model based on the corresponding 18 F-FDG relative importance weight 18 The main influencing factors and secondary influencing factors corresponding to F-FDG efficiency are iteratively trained to generate the efficiency calculation model. 18 F-FDG efficiency calculation model; 18 F-FDG production warning module is used to obtain current 18 The main and secondary influencing factors corresponding to the F-FDG production process are input into the trained 18 F-FDG efficiency calculation model 18 F-FDG efficiency prediction to generate the current corresponding 18 F-FDG efficiency; based on the preset first-level warning threshold and second-level warning threshold, the current corresponding 18 F-FDG efficiency 18 F-FDG production warning to generate corresponding 18 F-FDG production early warning adjustment measures.

Images