Vaccination one-stop informatization management platform based on AI intelligent monitoring

Abstract

The invention discloses a vaccination one-stop informatization management platform based on AI intelligent monitoring, and relates to the technical field of medical data management. The platform comprises a database module, a data collection and feature extraction module, a model training module, an inoculation reaction prediction module, a vaccine refrigeration environment monitoring module, a vaccine purchase and inventory management module, an inoculation reservation and execution management module and an inoculator tracking management module. Unified management of vaccinator information, vaccination data and vaccine information is realized, the data utilization rate is improved, the vaccination reaction and the vaccine refrigeration environment temperature are accurately predicted, the refrigeration environment is monitored in real time, vaccine purchase and inventory management are optimized, vaccination reservation and execution processes are facilitated, and vaccinator tracking management is enhanced; according to the invention, the problems of non-standard data management, low utilization rate, inaccurate prediction, improper monitoring and the like in the traditional vaccination management are effectively solved, and the efficiency and safety of vaccination are remarkably improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of medical data management, and more specifically, relates to a one-stop information management platform for vaccination based on AI intelligent monitoring. Background Art

[0002] There are many problems with traditional vaccination management methods, such as irregular information management, low data utilization, inaccurate predictions of vaccination reactions, and inadequate monitoring of the vaccine refrigeration environment. These problems seriously affect the efficiency and safety of vaccination.

[0003] The vaccine recipient information, vaccination data, vaccine information, etc. are stored in a scattered manner and lack a unified management platform, which makes data retrieval, updating and backup difficult; historical vaccination data, vaccination reaction data, etc. are not effectively utilized, and cannot provide data support for vaccination reaction prediction and vaccine refrigeration environment monitoring; the existing vaccination reaction prediction methods mainly rely on doctors' experience, lack scientificity and accuracy, and cannot provide timely response methods for vaccine recipients; the vaccine refrigeration environment temperature monitoring methods are backward and cannot adjust the temperature in real time to ensure vaccine quality. Summary of the invention

[0004] 1. Technical issues to be resolved In response to the problems in the related technologies, the present invention provides a one-stop information management platform for vaccination based on AI intelligent monitoring. Through artificial intelligence, machine learning and other technologies, it solves the problems of irregular data management, low utilization rate, inaccurate prediction of vaccination reactions, inadequate monitoring of the refrigerated environment, inefficient procurement and inventory management, inconvenient appointment and execution management, and insufficient tracking and management of vaccine recipients.

[0005] (II) Technical solution To solve the above technical problems, the present invention is implemented through the following technical solutions: a one-stop information management platform for vaccination based on AI intelligent monitoring, including the following modules: A database module for storing and managing vaccine recipient information, vaccination data, vaccine information, cold storage environment data, purchase and inventory data, and vaccination reaction data; The data collection and feature extraction module is used to collect historical vaccination data, vaccination reaction data, real-time vaccination feature data, vaccine refrigeration environment temperature data, and vaccine procurement and inventory data from the database; extract information including but not limited to vaccine recipient information, vaccine type, vaccination time features, and construct a historical vaccination feature matrix and a historical vaccination feature label matrix; Model training module, used to build and train logistic regression models and MLP models to predict vaccination reactions and vaccine refrigeration environment temperature; The vaccination reaction prediction module is used to combine the final logistic regression model to predict the real-time physical reaction data of the vaccine recipients and match them with the coping methods in the database; The vaccine cold storage environment monitoring module is used to collect real-time vaccine cold storage environment temperature data, and predict the vaccine cold storage environment temperature data at future times based on the MLP model to adjust the vaccine cold storage environment temperature; Vaccine Procurement and Inventory Management module for electronic procurement process, purchase order tracking and vaccine inventory management; Vaccination appointment and execution management module, used to manage the vaccine appointment process and vaccine execution process; The vaccine recipient tracking management module is used to remind vaccine recipients of precautions and collect their feedback.

[0006] Preferably, the data collection and feature extraction module is used to collect historical vaccination data, vaccination reaction data, real-time vaccination feature data, vaccine refrigeration environment temperature data, and vaccine procurement and inventory data from the database; extract features including but not limited to vaccine recipient information, vaccine type, and vaccination time, and construct a historical vaccination feature matrix and a historical vaccination feature label matrix. The specific steps are as follows: S11. Collect historical vaccination data from the database, including information about the vaccine recipient, vaccine type, vaccination time, etc., obtain historical vaccination data, extract features, including the vaccine recipient's age, gender, health status, vaccine batch, etc., and construct a historical vaccination feature matrix A ,as follows, ; in, A im Represents the historical vaccination feature matrix i The first vaccination m Features, n Represents the total number of vaccinations in the historical vaccination feature matrix; S12. Collect the vaccination response data of each vaccination in the historical vaccination feature matrix from the database, extract the features, and construct the historical vaccination feature label matrix B ,as follows, ; in, B ip Represents the historical vaccination feature matrix i The first vaccination, the first p Features S13, arranging a temperature sensor in the vaccine cold storage environment, the temperature sensor is used to collect real-time temperature data of the vaccine cold storage environment, and upload the temperature data to the database; S14, collecting historical vaccine cold storage environment data from the database and constructing a historical vaccine cold storage environment data matrix; Through the above steps, historical vaccination data, vaccination reaction data, real-time vaccination feature data, vaccine refrigeration environment temperature data, and vaccine procurement and inventory data were accurately collected from the database, and multidimensional features including vaccine recipient information, vaccine type, and vaccination time characteristics were deeply extracted; by constructing a historical vaccination feature matrix and a historical vaccination feature label matrix, refined analysis and prediction of vaccination reactions were achieved. At the same time, real-time monitoring of the vaccine refrigeration environment temperature and construction of a historical data matrix provided data support for the optimization of vaccine storage conditions.

[0007] Preferably, the model training module is used to construct and train a logistic regression model and an MLP model, and the specific steps of predicting the vaccination reaction and the vaccine refrigeration environment temperature are as follows: S21, constructing an initial logistic regression model, setting the initial weight of the initial logistic regression model to a 11 , the initial bias is a 12 The initial learning rate is a 13 , set the training ratio and test ratio of the initial logistic regression model to be d 1 and d 2 , set the training accuracy threshold to e , the test accuracy threshold is f ; S22, according to the training ratio of the initial logistic regression model d 1 and test ratio d 2 Divide the historical vaccination feature matrix into a historical vaccination feature training matrix and a historical vaccination feature test matrix, and divide the historical vaccination feature label matrix into a historical vaccination feature training label matrix and a historical vaccination feature test label matrix; S23, using a grid search combined with a cross-validation algorithm, combining a historical vaccination feature training matrix and a historical vaccination feature training label matrix to train an initial logistic regression model, adjusting an initial weight, an initial bias, and an initial learning rate of the initial logistic regression model, and obtaining a trained logistic regression model; S24, inputting the historical vaccination feature training label matrix and the historical vaccination feature test label matrix into the trained logistic regression model for testing, and after the test optimization is completed, obtaining the final logistic regression model; S25, using the historical vaccine cold storage environment training data matrix and the historical vaccine cold storage environment test data matrix, training and testing the initial MLP model and optimizing it to obtain the final MLP model; Through the above steps, the logistic regression model and the MLP model were constructed and trained, and the vaccination reaction and the vaccine refrigeration environment temperature were accurately predicted; the high accuracy and stability of the model were ensured, and the accuracy of the vaccination reaction prediction and the real-time monitoring of the vaccine refrigeration environment temperature were effectively improved; through scientific data analysis and model training, the present invention provides strong technical support for vaccination management, further ensures the safety and effectiveness of vaccination, optimizes resource allocation, and improves overall management efficiency.

[0008] Preferably, the specific steps of S23 are as follows: S231, inputting the historical vaccination feature training matrix and the historical vaccination feature training label matrix into the initial logistic regression model to train the initial logistic regression model; S232, set the parameter search space to p , the maximum number of random searches is q , set the cross validation fold to X ; S233, initializing the optimal parameter combination , the initialization optimal performance is b , initial training accuracy h ; Randomly select parameter combinations , build a logistic regression model based on the selected parameter combination and divide the training data into X Fold, for each fold x , ; Use except x All training data outside the fold are used to train the logistic regression model. x The training data of the fold is used to evaluate the model and obtain the performance indicator set , calculate the mean of the performance indicators of all folds and get the average performance g , the calculation formula is as follows, ; If the average performance g > Best performance b , then the updated best performance is b , update the optimal parameter combination to , and update the training accuracy h ; Otherwise, the original best performance, optimal parameter combination and training accuracy are maintained; S234, repeat S233, when the maximum number of random searches is reached p or training accuracy h ≥ Training accuracy thresholde When , stop the iteration and get the weight a 21 , biased to a 22 , the learning rate is a 23 The optimal parameter combination is used as the parameter of the logistic regression model to obtain the trained logistic regression model; The above steps enable fine training and parameter tuning of the logistic regression model. By inputting the historical vaccination feature training matrix and label matrix, combined with the parameter search space setting, the maximum number of random searches and the cross-validation fold, the optimal parameter combination is initialized and iteratively updated, and finally a logistic regression model with optimal weights, bias and learning rate is obtained. This not only improves the training accuracy of the model, but also ensures the stability and reliability of the model in predicting vaccination reactions.

[0009] Preferably, the specific steps of S24 are as follows: S241, input the trained logistic regression model to test, obtain the test output result, compare the test output result with the historical vaccination feature test label matrix, and obtain the test accuracy k ; S242, when k ≥ f , taking the trained logistic regression model as the final logistic regression model; S243, when k < f When , the maximum number of random searches in the grid search combined with cross validation is increased, and the increased maximum number of random searches is o ; S244, use the increased maximum number of random searches o , continue to train the trained logistic regression model through grid search combined with cross validation, and test after each round of training; S245, repeat S244, when the random search times reaches the maximum random search times o or k ≥ f When , the iteration is stopped and the final logistic regression model is obtained; The above steps ensure the high accuracy and stability of the logistic regression model in predicting vaccination reactions; the test data is input into the trained model for testing and compared with the true labels to calculate the test accuracy; if the test accuracy reaches the preset threshold, the model is confirmed as the final logistic regression model; if the threshold is not reached, the model training and testing continues by increasing the maximum number of random searches in the grid search; this iterative process continues until the maximum number of searches is reached or the test accuracy meets the requirements; the quality of the model and the prediction accuracy are improved, providing a more reliable technical guarantee for the prediction of vaccination reactions.

[0010] Preferably, the vaccination reaction prediction module is used to predict the real-time physical reaction data of the vaccinee in combination with the final logistic regression model, and the specific steps of matching the data with the coping methods in the database are as follows: S31. Collect the characteristic data of real-time vaccination from the database, combine the characteristic data of real-time vaccination with the final logistic regression model, predict the physical reaction data of the real-time vaccinee, match the physical reaction data of the real-time vaccinee with the coping methods in the database, if the coping methods are matched, return the physical reaction data of the real-time vaccinee and the coping methods to the database, if the coping methods are not matched, send the physical reaction data of the real-time vaccinee to the doctor, and after the doctor provides the coping methods, save the coping methods in the database; The above steps, combined with the final logistic regression model, achieve accurate prediction of the real-time physical reaction data of the vaccine recipients, and intelligently match them with the coping methods in the database, providing personalized health management services for the vaccine recipients. At the same time, it accumulates rich coping method data, providing a reference for handling similar situations in the future; the doctor intervention mechanism ensures professional guidance in complex situations and further improves the safety of vaccination.

[0011] Preferably, the vaccine refrigeration environment monitoring module is used to collect real-time vaccine refrigeration environment temperature data, and predict the vaccine refrigeration environment temperature data at future moments according to the MLP model. The specific steps for adjusting the vaccine refrigeration environment temperature are as follows: S41, using the historical vaccine cold storage environment training data matrix and the historical vaccine cold storage environment test data matrix, the initial MLP model is trained and tested and optimized to obtain the final MLP model; S42, setting a reasonable range of vaccine refrigeration environment temperature, collecting real-time vaccine refrigeration environment temperature data, if the real-time vaccine refrigeration environment temperature is within the reasonable range of vaccine refrigeration environment temperature, then there is no need to adjust the vaccine refrigeration environment temperature, if the real-time vaccine refrigeration environment temperature is not within the reasonable range of vaccine refrigeration environment temperature, adjust the temperature of the vaccine refrigeration environment; S43, input the real-time vaccine refrigeration environment temperature data into the final MLP model to obtain the vaccine refrigeration environment temperature data at the future time. If the future vaccine refrigeration environment temperature is within the reasonable range of the vaccine refrigeration environment temperature, there is no need to adjust the vaccine refrigeration environment temperature. If the future vaccine refrigeration environment temperature is not within the reasonable range of the vaccine refrigeration environment temperature, adjust the temperature of the vaccine refrigeration environment in advance. The above steps realize intelligent monitoring and adjustment of the vaccine refrigeration environment by collecting real-time vaccine refrigeration environment temperature data and using the final MLP model to predict the vaccine refrigeration environment temperature in the future. It can not only monitor the refrigeration environment temperature in real time to ensure that it is within a reasonable range, but also predict future temperature changes and make adjustments in advance, thereby effectively avoiding the impact of temperature fluctuations on vaccine quality. It improves the safety and stability of vaccine storage, ensures the effectiveness of vaccines, reduces vaccine waste due to improper temperature, and improves the overall management level of vaccination.

[0012] Preferably, the vaccine procurement and inventory management module is used for electronic procurement process, purchase order tracking and vaccine inventory management. The specific steps are as follows: S51, determine the purchase demand, click the purchase application button in the vaccine purchase and inventory management module, and select the purchase quantity and supplier name; after approval, obtain the purchase information data, and upload the logistics information of the purchase order to the vaccine purchase and inventory management module to track the purchase order; S52. Set a vaccine inventory threshold. After the vaccine is used and put into storage, update the vaccine inventory. When the vaccine inventory is less than or equal to the vaccine inventory threshold, the system sends an inventory warning notification to the manager. The above steps achieve efficient and accurate control of vaccine procurement and inventory through electronic procurement processes, purchase order tracking and vaccine inventory management; improve the efficiency and accuracy of vaccine procurement and inventory management, reduce management costs, and ensure the timely supply and reasonable distribution of vaccines.

[0013] Preferably, the specific steps of the vaccination appointment and execution management module for managing the vaccine appointment process and the vaccine execution process are as follows: S61. Set the number threshold of the appointment time period. The vaccine recipient clicks the vaccine appointment button in the vaccine appointment module and enters his / her personal information according to the prompts to complete the appointment, including the name of the appointment recipient, identity information, appointment time period, etc. When the number of people in the appointment time period is equal to the number threshold of the appointment time period, the appointment channel for the appointment time period is closed; S62. One day before the appointment date, inform the vaccine recipient via SMS to arrive at the hospital during the appointment time period; S63. After the vaccine recipient arrives at the hospital, he / she clicks the "Arrived at the Hospital" button in the vaccination execution module and selects the current location of the vaccine recipient to plan a route for the vaccine recipient to reach the vaccination site; S64. After the vaccination is completed, the identity information and vaccination information of the vaccinee are stored in the database, including the gender, year, physical condition, and batch information of the vaccinee; The above steps achieve reasonable control of the number of vaccinations, avoiding problems such as crowding and long waiting times; vaccine recipients can easily complete personal information input and appointments, and the system automatically closes the appointment channel during full time periods, ensuring the fairness and orderliness of appointments; it also provides route planning functions to the vaccination site, providing vaccine recipients with more humane services; after the vaccination is completed, the system automatically saves the vaccine recipient's identity information and vaccination information, providing complete data support for subsequent data analysis and tracking management.

[0014] Preferably, the specific steps of the vaccine recipient tracking management module for reminding vaccine recipients of matters needing attention and collecting vaccine recipients' feedback are as follows: S71. After the vaccine recipient completes the vaccination, the system automatically extracts the vaccine recipient's physical reaction data and coping method data from the database and sends it to the vaccine recipient; S72, the vaccinee clicks the feedback button of the vaccinee tracking management module for the physical reaction after the vaccination, and inputs the physical reaction data into the system. After receiving the physical reaction data, the system stores the physical reaction data and the vaccination data in the database; The above steps achieve high care and effective management of the vaccine recipients by automatically extracting and sending the vaccine recipients' physical reaction data and coping methods; after the vaccination is completed, the system immediately provides the vaccine recipients with personalized precautions and coping suggestions, enhancing the vaccine recipients' self-management ability; it provides a convenient feedback channel, and the vaccine recipients can easily enter physical reaction data by clicking the feedback button. After receiving the data, the system saves it to the database together with the corresponding vaccination data, forming a complete data record; this tracking and management method not only enhances the vaccine recipients' trust and satisfaction with the vaccination process, but also accumulates valuable feedback data, which provides important data support for subsequent vaccination reaction analysis and vaccine improvements, and further optimizes the vaccination service and management level.

[0015] (III) Beneficial effects The present invention has the following beneficial effects: The present invention standardizes data management and improves data utilization; through a unified database module, it realizes the centralized storage, retrieval, update and backup of vaccine recipient information, vaccination data, vaccine information, etc., solving the problems of decentralized data storage and irregular data management in traditional methods; at the same time, through the data collection and feature extraction module, it effectively utilizes historical vaccination data, vaccination reaction data, etc., provides solid data support for vaccination reaction prediction and vaccine cold storage environment monitoring, and improves the utilization value of data.

[0016] The present invention improves the accuracy of vaccination reaction prediction and the real-time monitoring of the cold storage environment; utilizes the logistic regression model and the MLP model to scientifically predict vaccination reactions and the temperature of the vaccine cold storage environment, thereby improving the accuracy and timeliness of the prediction; the vaccination reaction prediction module can combine the real-time vaccination feature data and the final logistic regression model to predict the physical reaction of the vaccinee, and match it with the coping methods in the database, thereby providing the vaccinee with timely and effective coping methods; the vaccine cold storage environment monitoring module can collect and adjust the vaccine cold storage environment temperature in real time, thereby ensuring the quality of the vaccine and guaranteeing the safety of vaccination.

[0017] The present invention optimizes vaccine procurement and inventory management, facilitates vaccination appointment and execution processes, and strengthens vaccine recipient tracking management; the vaccine procurement and inventory management module realizes electronic procurement process, real-time order tracking and inventory management, and improves the efficiency and accuracy of procurement and inventory management; the vaccination appointment and execution management module simplifies the appointment process, provides flexible execution management, and provides vaccine recipients with convenient appointment services and efficient vaccination experience; the vaccine recipient tracking management module can automatically remind vaccine recipients of precautions, collect vaccine recipient feedback, realize full-process tracking management of vaccine recipients, and further improve the quality and effectiveness of vaccination services.

[0018] Of course, any product implementing the present invention does not necessarily need to achieve all of the advantages described above at the same time. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] In order to more clearly illustrate the technical solutions of the embodiments of the invention, the drawings required for describing the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the invention. For ordinary technicians in this field, they can also obtain drawings based on these drawings without paying creative work.

[0020] Figure 1 This is a module schematic diagram of a one-stop information management platform for vaccination based on AI intelligent monitoring in the present invention. DETAILED DESCRIPTION

[0021] The following will be combined with the drawings in the embodiments of the invention to clearly and completely describe the technical solutions in the embodiments of the invention. Obviously, the described embodiments are only part of the embodiments of the invention, not all of the embodiments. Based on the embodiments in the invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the invention.

[0022] In the description of the present invention, it is necessary to understand that the terms "opening", "upper", "lower", "top", "middle", "inside" and the like indicating orientation or positional relationship are only for the convenience of describing the invention and simplifying the description, and do not indicate or imply that the components or elements referred to must have a specific orientation, be constructed and operate in a specific orientation, and therefore cannot be understood as limiting the invention.

[0023] See also Figure 1 The present invention discloses a one-stop information management platform for vaccination based on AI intelligent monitoring, including the following modules: A database module for storing and managing vaccine recipient information, vaccination data, vaccine information, cold storage environment data, purchase and inventory data, and vaccination reaction data; The data collection and feature extraction module is used to collect historical vaccination data, vaccination reaction data, and vaccine refrigeration environment temperature data from the database; extract features, and construct a historical vaccination feature matrix and a historical vaccination feature label matrix; Model training module, used to build and train logistic regression models and MLP models to predict vaccination reactions and vaccine refrigeration environment temperature; The vaccination reaction prediction module is used to combine the final logistic regression model to predict the real-time physical reaction data of the vaccine recipients and match them with the coping methods in the database; The vaccine cold storage environment monitoring module is used to collect real-time vaccine cold storage environment temperature data, and predict the vaccine cold storage environment temperature data at future times based on the MLP model to adjust the vaccine cold storage environment temperature; Vaccine Procurement and Inventory Management module for electronic procurement process, purchase order tracking and vaccine inventory management; Vaccination appointment and execution management module, used to manage the vaccine appointment process and vaccine execution process; The vaccine recipient tracking management module is used to remind vaccine recipients of precautions and collect their feedback.

[0024] The data collection and feature extraction module is used to collect historical vaccination data, vaccination reaction data, real-time vaccination feature data, vaccine refrigeration environment temperature data, and vaccine procurement and inventory data from the database; extracting but not limited to vaccine recipient information, vaccine type, vaccination time features, and constructing a historical vaccination feature matrix and a historical vaccination feature label matrix. The specific steps are as follows: S11. Collect historical vaccination data from the database, including information about the vaccine recipient, vaccine type, vaccination time, etc., obtain historical vaccination data, extract features, including the vaccine recipient's age, gender, health status, vaccine batch, etc., and construct a historical vaccination feature matrix A ,as follows, ; in, A im Represents the historical vaccination feature matrix i The first vaccination m Features, n Represents the total number of vaccinations in the historical vaccination feature matrix; S12. Collect the vaccination response data of each vaccination in the historical vaccination feature matrix from the database, extract the features, and construct the historical vaccination feature label matrix B ,as follows, ; in, B ip Represents the historical vaccination feature matrix i The first vaccination, the first vaccination reaction of the vaccine recipient p Features S13, arranging a temperature sensor in the vaccine cold storage environment, the temperature sensor is used to collect real-time temperature data of the vaccine cold storage environment, and upload the temperature data to the database; S14. Collect historical vaccine refrigeration environment data from the database and construct a historical vaccine refrigeration environment data matrix.

[0025] The model training module is used to construct and train a logistic regression model and an MLP model, and the specific steps of the prediction package for vaccination reaction and vaccine refrigeration environment temperature are as follows: S21, constructing an initial logistic regression model, setting the initial weight of the initial logistic regression model to a 11 , the initial bias is a 12 The initial learning rate is a 13 , set the training ratio and test ratio of the initial logistic regression model to be d 1 and d 2 , set the training accuracy threshold to e , the test accuracy threshold is f ; S22, according to the training ratio of the initial logistic regression model d 1 and test ratiod 2 Divide the historical vaccination feature matrix into a historical vaccination feature training matrix and a historical vaccination feature test matrix, and divide the historical vaccination feature label matrix into a historical vaccination feature training label matrix and a historical vaccination feature test label matrix; S23, using a grid search combined with a cross-validation algorithm, combining a historical vaccination feature training matrix and a historical vaccination feature training label matrix to train an initial logistic regression model, adjusting an initial weight, an initial bias, and an initial learning rate of the initial logistic regression model, and obtaining a trained logistic regression model; The specific steps of S23 are as follows: S231, inputting the historical vaccination feature training matrix and the historical vaccination feature training label matrix into the initial logistic regression model to train the initial logistic regression model; S232, set the parameter search space to p , the maximum number of random searches is q , set the cross validation fold to X ; S233, initializing the optimal parameter combination , the initialization optimal performance is b , initial training accuracy h ; Randomly select parameter combinations , build a logistic regression model based on the selected parameter combination and divide the training data into X Fold, for each fold x , ; Use except x All training data outside the fold are used to train the logistic regression model. x The training data of the fold is used to evaluate the model and obtain the performance indicator set , calculate the mean of the performance indicators of all folds and get the average performance g , the calculation formula is as follows, ; If the average performance g > Best performance b , then the updated best performance is b , update the optimal parameter combination to , and update the training accuracy h ; Otherwise, the original best performance, optimal parameter combination and training accuracy are maintained; S234, repeat S233, when the maximum number of random searches is reached p or training accuracy h ≥ Training accuracy threshold e When , stop the iteration and get the weight a 21 , biased toa 22 , the learning rate is a 23 The optimal parameter combination is used as the parameter of the logistic regression model to obtain the trained logistic regression model; S24, inputting the historical vaccination feature training label matrix and the historical vaccination feature test label matrix into the trained logistic regression model for testing, and after the test optimization is completed, obtaining the final logistic regression model; The specific steps of S24 are as follows: S241, input the trained logistic regression model to test, obtain the test output result, compare the test output result with the historical vaccination feature test label matrix, and obtain the test accuracy k ; S242, when k ≥ f , taking the trained logistic regression model as the final logistic regression model; S243, when k < f When , the maximum number of random searches in the grid search combined with cross validation is increased, and the increased maximum number of random searches is o ; S244, use the increased maximum number of random searches o , continue to train the trained logistic regression model through grid search combined with cross validation, and test after each round of training; S245, repeat S244, when the random search times reaches the maximum random search times o or k ≥ f When , the iteration is stopped and the final logistic regression model is obtained; S25. Use the historical vaccine refrigeration environment training data matrix and the historical vaccine refrigeration environment test data matrix to train and test the initial MLP model and optimize it to obtain the final MLP model.

[0026] The specific steps of the vaccination reaction prediction module for predicting the real-time physical reaction data of the vaccinee in combination with the final logistic regression model and matching it with the response methods in the database are as follows: S31. Collect the characteristic data of real-time vaccination from the database, combine the characteristic data of real-time vaccination with the final logistic regression model, predict the physical reaction data of the real-time vaccinee, match the physical reaction data of the real-time vaccinee with the coping methods in the database, if the coping methods are matched, return the physical reaction data of the real-time vaccinee and the coping methods to the database, if the coping methods are not matched, send the physical reaction data of the real-time vaccinee to the doctor, and after the doctor provides the coping methods, save the coping methods in the database; The vaccine cold storage environment monitoring module is used to collect real-time vaccine cold storage environment temperature data and predict the vaccine cold storage environment temperature data at future moments according to the MLP model. The specific steps for adjusting the vaccine cold storage environment temperature are as follows: S41, using the historical vaccine cold storage environment training data matrix and the historical vaccine cold storage environment test data matrix, the initial MLP model is trained and tested and optimized to obtain the final MLP model; S42, setting a reasonable range of vaccine refrigeration environment temperature, collecting real-time vaccine refrigeration environment temperature data, if the real-time vaccine refrigeration environment temperature is within the reasonable range of vaccine refrigeration environment temperature, then there is no need to adjust the vaccine refrigeration environment temperature, if the real-time vaccine refrigeration environment temperature is not within the reasonable range of vaccine refrigeration environment temperature, adjust the temperature of the vaccine refrigeration environment; S43, input the real-time vaccine refrigeration environment temperature data into the final MLP model to obtain the vaccine refrigeration environment temperature data at the future time. If the future vaccine refrigeration environment temperature is within the reasonable range of the vaccine refrigeration environment temperature, there is no need to adjust the vaccine refrigeration environment temperature. If the future vaccine refrigeration environment temperature is not within the reasonable range of the vaccine refrigeration environment temperature, adjust the temperature of the vaccine refrigeration environment in advance. The vaccine procurement and inventory management module is used for electronic procurement process, purchase order tracking and vaccine inventory management. The specific steps are as follows: S51, determine the purchase demand, click the purchase application button in the vaccine purchase and inventory management module, and select the purchase quantity and supplier name; after approval, obtain the purchase information data, and upload the logistics information of the purchase order to the vaccine purchase and inventory management module to track the purchase order; S52. Set a vaccine inventory threshold. After the vaccine is used and put into storage, update the vaccine inventory. When the vaccine inventory is ≤ the vaccine inventory threshold, the system sends an inventory warning notification to the manager.

[0027] The specific steps of the vaccination appointment and execution management module for managing the vaccine appointment process and the vaccine execution process are as follows: S61. Set the number threshold of the appointment time period. The vaccine recipient clicks the vaccine appointment button in the vaccine appointment module and enters his / her personal information according to the prompts to complete the appointment, including the name of the appointment recipient, identity information, appointment time period, etc. When the number of people in the appointment time period is equal to the number threshold of the appointment time period, the appointment channel for the appointment time period is closed; S62. One day before the appointment date, inform the vaccine recipient via SMS to arrive at the hospital during the appointment time period; S63. After the vaccine recipient arrives at the hospital, he / she clicks the "Arrived at the Hospital" button in the vaccination execution module and selects the current location of the vaccine recipient to plan a route for the vaccine recipient to reach the vaccination site; S64. After the vaccination is completed, the identity information and vaccination information of the vaccinee are stored in the database, including the gender, year, physical condition, and batch information of the vaccinee; The specific steps of the vaccine recipient tracking management module for reminding vaccine recipients of matters needing attention and collecting vaccine recipients' feedback are as follows: S71. After the vaccine recipient completes the vaccination, the system automatically extracts the vaccine recipient's physical reaction data and coping method data from the database and sends it to the vaccine recipient; S72. The vaccinee clicks the feedback button of the vaccinee tracking management module regarding the physical reaction after vaccination, and inputs the physical reaction data into the system. After receiving the physical reaction data, the system stores the physical reaction data and the vaccination data together in the database.

[0028] In the description of this specification, the description with reference to the terms "one embodiment", "example", "specific example", etc. means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the invention. In this specification, the schematic representation of the above terms does not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described can be combined in any one or more embodiments or examples in a suitable manner.

[0029] The preferred embodiments of the invention disclosed above are only used to help explain the invention. The preferred embodiments do not describe all the details in detail, nor do they limit the invention to the specific implementation methods described. Obviously, many modifications and changes can be made according to the content of this specification. This specification selects and specifically describes these embodiments in order to better explain the principles and practical applications of the invention, so that those skilled in the art can understand and use the invention well.

Claims

1. A one-stop information management platform for vaccination based on AI intelligent monitoring, characterized in that: Includes the following modules: A database module for storing and managing vaccine recipient information, vaccination data, vaccine information, cold storage environment data, purchase and inventory data, and vaccination reaction data; The data collection and feature extraction module is used to collect historical vaccination data, vaccination reaction data, and vaccine refrigeration environment temperature data from the database; Extract features and construct historical vaccination feature matrix and historical vaccination feature label matrix; Model training module, used to build and train logistic regression models and MLP models to predict vaccination reactions and vaccine refrigeration environment temperature; The vaccination reaction prediction module is used to combine the final logistic regression model to predict the real-time physical reaction data of the vaccine recipients and match them with the coping methods in the database; The vaccine cold storage environment monitoring module is used to collect real-time vaccine cold storage environment temperature data, and predict the vaccine cold storage environment temperature data at future times based on the MLP model to adjust the vaccine cold storage environment temperature; Vaccine Procurement and Inventory Management module for electronic procurement process, purchase order tracking and vaccine inventory management; Vaccination appointment and execution management module, used to manage the vaccine appointment process and vaccine execution process; The vaccine recipient tracking management module is used to remind vaccine recipients of precautions and collect their feedback.

2. A one-stop information management platform for vaccination based on AI intelligent monitoring according to claim 1, characterized in that: The specific steps of the data collection and feature extraction module are as follows: S11, collecting historical vaccination data from the database, obtaining historical vaccination data, extracting features, and constructing a historical vaccination feature matrix; S12, collecting vaccination response data of each vaccination and the vaccinee in the historical vaccination feature matrix from the database, extracting features, and constructing a historical vaccination feature label matrix; S13, arranging a temperature sensor in the vaccine cold storage environment, the temperature sensor is used to collect real-time temperature data of the vaccine cold storage environment, and upload the temperature data to the database; S14. Collect historical vaccine refrigeration environment data from the database and construct a historical vaccine refrigeration environment data matrix.

3. According to claim 1, a one-stop information management platform for vaccination based on AI intelligent monitoring, characterized in that: The specific steps of the model training module are as follows: S21. Build the initial logistic regression model, set the initial weight, initial bias, initial learning rate, set the training ratio and test ratio, set the training accuracy threshold, and test accuracy threshold f ; S22. Dividing the historical vaccination feature matrix into a historical vaccination feature training matrix and a historical vaccination feature test matrix according to the training ratio and the test ratio of the initial logistic regression model, and dividing the historical vaccination feature label matrix into a historical vaccination feature training label matrix and a historical vaccination feature test label matrix; S23, using a grid search combined with a cross-validation algorithm, combining a historical vaccination feature training matrix and a historical vaccination feature training label matrix to train an initial logistic regression model, adjusting an initial weight, an initial bias, and an initial learning rate of the initial logistic regression model, and obtaining a trained logistic regression model; S24, inputting the historical vaccination feature training label matrix and the historical vaccination feature test label matrix into the trained logistic regression model for testing, and after the test optimization is completed, obtaining the final logistic regression model; S25. Use the historical vaccine refrigeration environment training data matrix and the historical vaccine refrigeration environment test data matrix to train and test the initial MLP model and optimize it to obtain the final MLP model.

4. A one-stop information management platform for vaccination based on AI intelligent monitoring according to claim 3, characterized in that: The specific steps of S23 are as follows: S231, inputting the historical vaccination feature training matrix and the historical vaccination feature training label matrix into the initial logistic regression model to train the initial logistic regression model; S232, set the parameter search space, the maximum number of random searches, and the cross-validation fold X ; S233, initialize the optimal parameter combination, initialize the best performance, initialize the training accuracy; randomly select a parameter combination, build a logistic regression model based on the selected parameter combination, and divide the training data into X Fold, for each fold x , ; Use except x All training data outside the fold are used to train the logistic regression model. x The training data of the fold is used to evaluate the model, and a set of performance indicators is obtained. The mean of the performance indicators of all folds is calculated to obtain the average performance. If the average performance is greater than the best performance, then update the best performance, update the best parameter combination, and update the training accuracy; otherwise, keep the original best performance, best parameter combination, and training accuracy; S234, repeat S233, when the maximum number of random searches is reached or the training accuracy is ≥ the training accuracy threshold, stop the iteration, obtain the optimal parameter combination and use it as the parameters of the logistic regression model, and obtain the trained logistic regression model.

5. According to claim 3, a one-stop information management platform for vaccination based on AI intelligent monitoring, characterized in that: The specific steps of S24 are as follows: S241, input the trained logistic regression model to test, obtain the test output result, compare the test output result with the historical vaccination feature test label matrix, and obtain the test accuracy k ; S242, when k ≥ f , taking the trained logistic regression model as the final logistic regression model; S243, when k < f When the maximum number of random searches in the grid search combined with cross validation is increased to o ; S244, continue to train the trained logistic regression model through grid search combined with cross validation, and perform testing after each round of training; S245, repeat S244, when the random search times reaches the maximum random search times o or k ≥ f When , the iteration stops and the final logistic regression model is obtained.

6. According to claim 1, a one-stop information management platform for vaccination based on AI intelligent monitoring, characterized in that: The specific steps of the vaccination response prediction module are as follows: S31. Collect characteristic data of real-time vaccination from the database, combine it with the final logistic regression model, predict the physical reaction data of the real-time vaccinee, and match it with the coping methods in the database. If the coping method is matched, the coping method is returned to the database. If the coping method is not matched, the physical reaction data of the real-time vaccinee is sent to the doctor. After the doctor provides the coping method, the coping method is saved in the database.

7. According to claim 1, a one-stop information management platform for vaccination based on AI intelligent monitoring, characterized in that: The specific steps of the vaccine refrigeration environment monitoring module are as follows: S41, setting a reasonable range of vaccine refrigeration environment temperature, collecting real-time vaccine refrigeration environment temperature data, if the real-time vaccine refrigeration environment temperature is within the reasonable range of vaccine refrigeration environment temperature, then there is no need to adjust the vaccine refrigeration environment temperature, otherwise, adjust the temperature of the vaccine refrigeration environment; S42. Input the real-time vaccine refrigeration environment temperature data into the final MLP model to obtain the vaccine refrigeration environment temperature data at the future moment. If the vaccine refrigeration environment temperature at the future moment is within the reasonable range of the vaccine refrigeration environment temperature, there is no need to adjust the vaccine refrigeration environment temperature. Otherwise, adjust the temperature of the vaccine refrigeration environment in advance.

8. According to claim 1, a one-stop information management platform for vaccination based on AI intelligent monitoring, characterized in that: The specific steps of the vaccine procurement and inventory management module are as follows: S51. Determine the purchase demand, click the purchase application button, and select the purchase quantity and supplier name; after approval, obtain the purchase information data, and upload the logistics information of the purchase order to the vaccine procurement and inventory management module to track the purchase order; S52. Set a vaccine inventory threshold. After the vaccine is used and put into storage, update the vaccine inventory. When the vaccine inventory is ≤ the vaccine inventory threshold, the system sends an inventory warning notification to the manager.

9. The one-stop information management platform for vaccination based on AI intelligent monitoring according to claim 1, characterized in that: The specific steps of the vaccination appointment and execution management module are as follows: S61. Set the number threshold of the appointment time period. The vaccine recipient clicks the vaccine appointment button in the vaccine appointment module and enters his / her personal information according to the prompt to complete the appointment. When the number of people in the appointment time period is greater than or equal to the number threshold of the appointment time period, the appointment channel for the time period is closed. S62. One day before the appointment date, inform the vaccine recipient via SMS to arrive at the hospital during the appointment time period; S63. After the vaccine recipient arrives at the hospital, he / she clicks the "Arrived at the Hospital" button in the vaccination execution module and selects the current location of the vaccine recipient to plan a route for the vaccine recipient to reach the vaccination site; S64. After the vaccination is completed, the identity information of the vaccinee and the vaccination information are stored in the database.

10. A one-stop information management platform for vaccination based on AI intelligent monitoring according to claim 1, characterized in that: The specific steps of the vaccine recipient tracking management module are as follows: S71. After the vaccine recipient completes the vaccination, the system extracts the vaccine recipient's physical reaction data and coping method data from the database and sends it to the vaccine recipient; S72. The vaccinee clicks the feedback button of the vaccinee tracking management module regarding the physical reaction after vaccination, and inputs the physical reaction data into the system. The system saves the physical reaction data corresponding to the vaccination data in a database.

Images