Implementación de una interfaz software para establecer grupos de riesgo de pacientes sospechosos de Tuberculosis Pulmonar

Abstract

Pulmonary tuberculosis (TB) is an infectious disease that usually affects the lungs, this disease is curable and preventable, however, delayed diagnosis and treatment can cause death. In 2019 an estimated 10 million people worldwide fell ill with tuberculosis and a total of 1.4 million people died according to the World Health Organization (WHO) in 2020 [1]. In Colombia, in 2020 11390 people became ill, of which 10632 were new cases, and on average during the last 5 years, there have been 1077 deaths per year due to tuberculosis, according to the National Public Health Surveillance System (SIVIGILA) in 2021 [2]. At the national level the high demand for patients hinders the priority and quality of care by the Health Service Provider Institutions (IPS). These problems originate due to the poor management of resources and procedures, generating as a consequence the progression of the disease or lethality due to TB, particularly in isolated regions. This project contributes to the decision-making of health specialists under precarious conditions from an automated system, which consists of a software interface that allows establishing the high or low risk group to which suspected tuberculosis patients belong by entering a series of social, demographic and health status characteristics of the patient. Three previously trained and validated automatic learning algorithms Fuzzy C-means, K-means and Perceptron Multilayer were worked on, of the three algorithms the Perceptron Multilayer network was chosen to perform the risk prediction because it has the highest sensitivity with a value of 95 %.