Design And Implementation Of Internet Of Things Management Platform For Intelligent Medical Treatment

In recent years,the aging of the population has been accompanied by news of sudden death and cancer among young people.With the percentage of sub-health population increased year by year,our medical resources far cannot satisfy the needs of people existing,let everyone go to the formal medical institutions detection is not realistic,At the same time,at the national planning level,China also added the development policy of medical Internet of Things to the smart medical Internet medical policy system based on the master plan of Health China in 2019.Therefore,the design of intelligent medical platform combining Internet of Things technology is of great significance to both individuals and countries.In view of the above background,this thesis first carries out the overall feasibility analysis of the platform,determines the functional requirements and non-functional requirements of the platform,and analyzes the five functional requirements and three non-functional requirements of the platform respectively.It laid a good foundation for the follow-up work.Secondly,the thesis designs the overall architecture of the intelligent medical IoT platform,and focuses on the design and implementation of the application layer of the platform.Based on the actual needs of the platform,the thesis designs the micro-service architecture for the application layer of the platform.The architecture is based on Spring Cloud,Spring Boot,Vue.js and other technologies.Then the connection mode of database and the communication mechanism of front and back end are designed and the application layer system of the platform is realized by using the development mode of front and back end separation.After the system test,the functional test and non-functional test of the platform have passed and met the expectation.Finally,the intelligent classification algorithm of ECG signals is studied and applied in this thesis.In this thesis,the "inter-patient" method is adopted to segment the MIT-BIT arrhythmia data set,and the algorithm model is studied with this data set.Based on the current situation of limited medical data and the need to protect the privacy of users,a Resnet+GRU algorithm is proposed,which uses federated learning at the same time.Experimental results show that the algorithm can protect users’privacy while ensuring the accuracy of the model.Then the Resnet+GRU algorithm model is deployed to the lower computer of the platform,and the intelligent prediction function is added to the application layer of the platform,so as to further verify the scalability of the platform and the availability of the algorithm.