Examinando por Autor "Prada Robles, Diana Carolina"
Mostrando 1 - 1 de 1
Resultados por página
Opciones de ordenación
- ÍtemDiseño de un modelo de aprendizaje automático para la predicción de casos de infección por SARS-CoV-2 a partir de parámetros clínicos de laboratorio(Fundación Universitaria Los Libertadores. Sede Bogotá., ) Prada Robles, Diana Carolina; González Veloza, José John FredyThe diagnosis of COVID-19 is crucial for the identification, isolation and treatment of contagious individuals, in order to mitigate the increase in cases as much as possible, and classify and prioritize them, according to the complexity of the disease manifestation. Although there are highly sensitive diagnostic tests, not all health institutions have the infrastructure or technology to perform them, consequently, the process must be outsourced, lengthening the diagnosis itself. Therefore, the present study focuses on design a machine learning model that allows predict cases of SARS-CoV-2 infection from clinical laboratory parameters, in the hospitalization service of a health institution in eastern of Colombia. Methodology: With the data of some biomarkers from the clinical laboratory, those that had a significant association with the spread of SARS-CoV-2 were evaluated, developing different machine learning algorithms, using PYTHON language libraries. Results: The Random Forest classifier was obtained as the best model with an AUCROC of 0.79, a sensitivity of 78% and an accuracy of 72%. Conclusion: The use of some blood biomarkers linked with machine learning algorithms can be useful tools for the prognosis of many diseases, including COVID-19.