| 摘要 |
PROBLEM TO BE SOLVED: To provide a technique for automatically, efficiently creating a prediction model that enables a user to easily grasp a causal correlation and a condition of transition of a clinical state so as to be able to practice efficient and effective healthcare services.SOLUTION: A medical data analysis system includes a health guidance assistance feature that creates and visualizes a graphical model while setting items of medical checkup information and receipt information as probability variables on the basis of the medical checkup information and the receipt information, and predicts a medical-expenses reduction effect by health guidance, thereby selecting effective health guidance contents and health guidance object persons. By setting up nodes (items) of the graphical model on the basis of degrees of dependence among the items and influences of the items on medical expenses, it is possible to simultaneously visualize a dependence relation among the items and the influences on the medical expenses. As a result, it is possible to graphically represent association of life habits with test values, association of the test values with onset of diseases, and association of increase in severity after the onset of diseases in an easy-to-understand manner. Furthermore, risks and factors of the onset of diseases and the increase in severity can be intuitively grasped. |
