Design And Implementation Of Medical Examination Results Generation System Based On Disease Prediction

With the development of medical science and technology,more and more intelligent systems are used to assist medical diagnosis.Physical examination system is an intelligent system,which usually has the functions of automatic generation,query and statistical analysis of physical examination results.At present,the automatic generation of physical examination results is mainly based on the knowledge base of medical experts.By setting the threshold of physical examination items,the system automatically matches and generates the corresponding physical examination templates when the results of physical examination items are abnormal.This method has the problem of inaccurate generation of physical examination results,which are greatly influenced by pre-set rules.In order to effectively solve the above problems,based on a large number of historical physical examination data,this paper carries out disease prediction through modeling and analysis,and carries out correlation analysis on the results of physical examination,and proposes a system solution of generating medical examination results based on disease prediction.The main contents of this paper include the following aspects.Firstly,the dataset comes from a third-class hospital in Henan Province.In the data preprocessing stage,several missing value filling methods are compared.Based on the accuracy and F1 evaluation criteria,the mean filling method is selected to fill the missing dataset,and the data are standardized.Pearson correlation coefficient method was used for feature screening,and the corresponding feature subset was screened by accuracy evaluation criteria.Six multi-label disease prediction models were constructed.Ensemble of Classifier Chains(ECC)algorithm was selected as multi-label disease prediction model by comparing and analyzing experimental data.Secondly,based on the relationship between disease and physical examination results in the medical examination dataset,Apriori algorithm is used to screen frequent items of medical examination results corresponding to diseases by confidence,and parameters are determined by evaluation criteria such as accuracy and recall rate,so as to build a model for generating medical examination results.Based on the predicted disease information and features,a medical examination result generation model based on features and diseases is constructed by using multi-label classification algorithm.Finally,we integrate the model of medical examination results generation based on Apriori method and the model of medical examination results generation based on disease and feature,and construct an integrated model of medical examination results generation by setting weight for each algorithm.Based on the medical examination result generation model of disease prediction,two servers are divided into business functions,cross-platform data transmission is carried out by interface,and the medical examination result generation system based on disease prediction is designed and implemented.