Machine Learning-Based Predictive Modeling of Complications of Chronic Diabetes.
Ilia V. Derevitskii, Sergey V. Kovalchuk
Abstract
Chronic diabetes is one of the most common chronic diseases in the world. For health care, this type of diabetes is one of the highest priority problems. This disease is related to many concomitant diseases, that lead to early disability and increased cardiovascular risk. T2DM patients have an increased risk of various complications. For such patients, medical specialists needed practical tools for calculating the future risks of complications. In this study, we propose a method for calculating the set of risks of the most common T2DM complications. These complications include nephropathy, neuropathy, chronic heart failure, and atrial fibrillation. This method includes a set of models for calculating the risk of a specific complication during a six-month follow-up. These models are based on Machine Learning methods. The method has a high quality of the prediction for different complications. Therefore, medical specialists can use this model as a practical tool. This tool can be used as part of the Support and Decision System for working with T2DM patients.