Visual Analysis Of Data Quality And Patient Cohorts In Electronic Health Records

In the process of modern medical informatization,electronic health records have always been the foundation of digital medical treatment,and electronic health record data is widely used in medical research and diagnosis.People hope to explore the hidden medical patterns from the clinical data,and then provide more accurate and scientific help to diagnosis or research in other fields.However,the diagnosis process of different patients is not the same,which results in the complexity and lack of data reflected in the electronic medical record data.To this puzzle,this paper has designed an electronic health record data quality analysis module.This module separately classifies and evaluates the types of data anomalies and errors,and visually displays the results of the data evaluation on the system,and finally provides users with an interactive choice of methods to repair the problems of data quality.In order to explore the potential patterns in the electronic health record data,we selected the corresponding data analysis and visualization methods to establish an electronic health record visualization system.The system shows the differences and connections and distribution of different attributes between cohorts between different patient cohorts based on different characteristics including scatter plots and sankey plots through the connections and interactions between various views,and users can also explore interactively.This paper also explores the correlation of different attributes among the data,and finally verifies the practicability of the entire system through specific examples.