Electronic Medical Record Visual Web Research

With the continuous development and progress of information technology, the electronic medical records system (EMR) has become important in the construction of medical and health information technology core and logos. Electronic medical records system as an important part of the hospital information system (HIS), electronic medical records system, has a practical significance and research value.EHR data visualization is the most direct means to assist medical professionals in comprehensively analyzing the data, to extract useful information from the patient’s massive data and to found hiding laws. Existing electronic medical record data visualization technology is mainly through the classification of treatment behavior and establishing the form of the time axis to organize and browse the electronic medical record data. The time axis visually displays the historical events data, but the analysis of data is still very time-consuming. And the traditional time axis lack visual expression of the relationship between medical data. However, in the actual medical environment, it often has some significant correlations among the clinical manifestations, disease symptoms and clinical interventions.The main works and innovative points of this thesis are as follows:(1) In order to make relationship between the data clearly, to make data of weak relationship display separately and the data of strong association display closely, it first need to classify the data. It research the minimum spanning tree algorithm, and combine the degree of association of nodes to come up with the clustering algorithm based on the degree of association of nodes and minimum spanning tree algorithm。 The main theory is:it proposes a function which computes the degree of association of nodes base on the relations of the structure of the nodes in the data set, it sets the threshold base on the characteristics of the scatter diagram of the degree of association of nodes. Then, it clusters the nodes of data set base on the threshold. What’s more, it clusters the nodes of data set by the minimum spanning tree algorithm based on the relations of the structure of the nodes. The improved algorithm makes clustering result more reasonable.(2) In order to make the diagram clearly shows the inherent network information, and make the display of data not mess, it needs to layout the data points.It proposes an optimized initial layout through analyzing the Fruchterman Reingoll (FR) layout algorithm. It decreases the iteration of the FR layout algorithm properly. It improved layout algorithm which makes lay outing process quickly and lay outing result clearly. Using the same data set, it compares the results of FR algorithm and improved FR algorithm to illustrate the advantages of the improved FR algorithm.(3)Regarding the improved algorithm as the core, it use JavaScript and python language, writes the codes of electronic medical record network visual module, and in troduces the main function of each module of the system.