Forecast And Assessment Of Health Records Based On The Disease Warning And Monitoring System

With the popularity and deep development of the hospital informatization construction, almost all the domestic large and medium-size hospitals and lots of small scale hospitals have established hospital information system which has been operated for many years. The widely use of hospital information system in medical institutions has promoted the digitalization of medical information, meanwhile, the large quantity application of electronic cases and occurrences, the digitization of medical equipments and instruments made the expansion of the medical information incessantly. These precious medical resources are of great value to the diagnosis and treatment of diseases, as well as the medical research. In order to help the medical care staff diagnose and treat patents quickly and correctly with the existing information, how to make the full use of the data in the hospital information system to make precise statistical analysis of the large amount of medical data has became an important issue waiting to be solved. Based on Jilin University Disease Early Warning as well as Monitoring and Control System, the present thesis has developed the module of health records management, and studied the data mining approach in the application of medical field on the basis of large amount of cases and other medical data.The module of health records management is mainly used to collect and administrate the plenty of medical information in the Disease Early Warning as well as Monitoring and Control System, and to inspect and analyse each kind of medical materials which are characterized with the concept of group health and include related factors and behaviours that influence people's health through medical statistical method. In the Disease Early Warning as well as Monitoring and Control System, the statistical method is mainly divided into medical questionnaire and scan of physical examination records. It is effective to collect and analyse each kind of health information and related influential factors through medical questionnaires from different groups of various living habits. Plenty of health information such as records of outpatient clinic, records of inpatients and records of special disease also provided the Disease Early Warning as well as Monitoring and Control System with many medical data. On that basis, these data are analysed by related methods of medical statistics. Medical statistics is a subject which aims to study the collection, classification, analysis and inference of the numerical data in medical field using the principles and methods of probability theory and mathematical statistics. It is also an important tool for medical research and cognizing medical rules. It is possible to analyse the correlation of two or more kinds of medical phenomenon. Through the analysing work of all kinds of medical questionnaires, it is helpful for the medical staff to understand the patients' living habits and related factors that influence their health, from which they can get some related medical disciplines, and to provide advice and help to the health care and clinical decision of different groups. It can provide information support to the planning, policy designing, inspection and evaluation of region or group health care through analysing of physical examination data and public health data. For example, after analysing the incidence of the same infectious disease of two regions at the same time, it aims to find out the fundamental differences of the factors that influence the disease between the two regions to help the medical staff to design effective controls and protective measure.Traditional medical statistics plays an important role in analysing the medical information, however, with the rapid development of information technology, the increasing expanding of scale, scope and depth of database application, traditional medical statistics can not discover more instructive information and rules from these medical information which is of high quantity and covers a wide range. In order to find out these rules from the large amount of medical information, evaluate the health condition of the patients, instruct the clinically choosing medicine and diagnosis, predict more correctly and judge quickly to the occurrence and development of the infectious disease, it is necessary to adopt more advanced computer technology and database technology. Date mining technology is such a kind of science. It can discover the rules of disease infecting at different time, in different regions and groups, classify case information of different patients and find out the rules. It is significant for the disease control and early-warning. In addition, with the change of public health from disease monitoring to dangerous factors of behaviour, there will be more huge sets of data. And the data mining field will be in important application.This thesis aims to study the association analysis in data mining technology and classification technique applied in medical field on the basis of the medical information collected from Jilin University Disease Early Warning as well as Monitoring and Control System. Association analysis is an active branch in data mining which is mainly used to discover the significant correlations between the large amounts of data. The correction got from the association analysis can be showed in the form of association rules and frequent itemset. This thesis studied the case information that is from the First Hospital of Jilin University of the Coronary heart disease Patients in using association rules mining algorithm Apriori algorithm in association analysis. It is found out the frequent itemset and the association rules of cardinal symptom of coronary heart disease patients. Classification in data mining technology is to find out a group of models and functions that can describe the typical characteristics of the data set in order to classify the types and groups of unknown data. This thesis explores the decision tree classification models in classification algorithm. Decision trees technology is a method that is established on the basis of information theory. It uses information gain to find the maximum amount of information to build a node, establishes the branch of the tree using recursive fashion in order to construct the module the decision tree. The specific decision tree construction algorithm of this research is ID3 algorithm. This thesis provides the detailed process of reasoning in using ID3 algorithm of the classification application of the coronary heart disease patients, including calculation of all kinds of symptoms, the procedures of the algorithm of choosing decisions and examples of decision tree construction. It also explores the value judgment of rules derived from the algorithm in medical field.Overall, this thesis specifically introduces the design and implementation of the module of health records management through which plenty of information of physical examination is analysed on the basis of Jilin University Disease Early Warning as well as Monitoring and Control System. And it also studied the data mining technology in medical field on the information in Early Warning and Monitoring System, did experiments on some cases respectively using association analysis and classification algorithm. The knowledge derived from the algorithm is of practical value with verification in medical field.