®®®® SIIA Público

Título del libro: 2019 Ieee International Autumn Meeting On Power, Electronics And Computing, Ropec 2019
Título del capítulo: Artificial intelligence system to support the clinical decision for influenza

Autores UNAM:
MARIA DEL CARMEN EDNA MARQUEZ MARQUEZ;
Autores externos:

Idioma:

Año de publicación:
2019
Palabras clave:

Diagnosis; Intelligent systems; Patient treatment; Polymerase chain reaction; Public health; Support vector machines; Artificial intelligence systems; Clinical decision; Clinical diagnosis; Historical data; Machine learning methods; Molecular analysis; Respiratory tract; Sensitivity and specificity; Learning systems


Resumen:

Influenza is an acute disease of the respiratory tract. In Mexico, the influenza every year occurs with temporality and intensity is not known exactly. This disease also represents a high cost for national public health. The goal of this research is to create an intelligent system to support by diagnosis of influenza using the relevant factors based on historical data of the Mexican population.Until now all patients who arrive at a health institution with symptoms of influenza are diagnosed like disease of Influenza and the drug for influenza is applied, for a minimal cases sends the molecular analysis using the real-time PCR technique to corroborate the diagnosis and in many cases the result is negative. The consequences are the amount of false positive, which implies a high cost for the Health Sector, the shortage in the market of the medicine that could not be enough for the whole season and, for the patient, also the secondary disorders that arise due to the use of the medication.We propose support the first clinical diagnosis with machine learning methods, we use 3473 registers of people plus 6 years old were taken care in Mexico City by public health institutions and treated by probable influenza and after taken a sample to PCRRT test to confirm the result. Our better machine learning method was support vector machine with a sensitivity and specificity of.9715 and.9285 respectively. © 2019 IEEE.


Entidades citadas de la UNAM: