Predicción del diagnostico de diabetes a partir de perfiles clínicos de pacientes utilizando aprendizaje automático

Abstract

Diabetes in Colombia is one of the leading causes of death in most of the country's departments, according to the Ministry of Health. The World Health Organization recognizes three main types of diabetes: type I, type II, and gestational. One of the main causes of death from diabetes is that when the patient is diagnosed, the disease is already advanced and therefore difficult to treat. Therefore, it is very important to make a diagnosis in time, so that the factors that derive from this event can be minimized, such as: serious complications (such as: amputations, heart attacks, eye damage, foot ulcer, among others.); monetary expenses (such as: hospital, personal, state); time invested, among others. One of the methods used and making use of technology is the prediction of the risk of developing diabetes using machine learning (ML), where the prognosis of the disease is obtained as a result and with it, prevention fatal results and reduction of financial expenses. This process has already been carried out over time and there are several studies in which an attempt is made to predict the diagnosis of diabetes using machine learning.