Diabetes index evaluation framework based on data mining technology: a genetic factor involved solution for predicting diabetes risk

Abstract

With the development of data mining, scientists began to apply information technology to solve medical problems. In this context, the idea of auxiliary medical service emerged. The purpose of this study is to propose a new framework predicting the probability of suffering from diabetes via diabetes index (DI), which is defined as a score to assess the diabetes-related risk of the participant. DI is calculated based on a diabetic clinical dataset and the SVM model is applied as well. Particularly, genetic feature is innovatively introduced as an important factor in view of the fact that people with family history are more vulnerable to diabetes. The framework is applied to implement a diabetes auxiliary evaluation system. After a set of comprehensive experiments, the assessment result is supposed to identify risk of the disease at an early stage, which contributes to a deeper understanding of one's own health conditions.