Visualizing dynamics of patients in hospitals using devise locations

In this paper, we propose a method for visualizing the temporal change of patient distributions in a hospital as a preliminary step towards the process-dynamics mining. Firstly, we perform statistical analysis on process sequences (we call courses) in order to capture which types of processes were frequently used in outpatient services. Secondly, assuming that a patient presented himself or herself at the place where an action was recorded in the hospital information systems, we investigate the geographical distribution of patients in 30-minutes granularity during the open hour of outpatient services. Finally, we try to visualize the dynamics of patients by projecting the change of patient distribution onto hospital maps.