Volume 13, Number 4

Marie: Validation of the Artificial Intelligence Model for Covid-19 Detection

  Authors

Valdirene Bento¹, Bruno Frederico Salaroli² and Paula Santos³, ¹Radiologist responsible at UBS, Brazil, ²General practitioner in Itapeva, Brazil, ³University of São Paulo, Brazil

  Abstract

Lung X-ray images, if processed using statistical and computational methods, can distinguish pneumonia from COVID-19. The present work shows that it is possible to extract lung X-ray characteristics to improve the methods of examining and diagnosing patients with suspected COVID-19, distinguishing them from malaria, tuberculosis, and Streptococcus pneumonia. More precisely, an intelligent computational model was developed to process lung X-ray images and classify whether the image is of a patient with COVID-19. In partnership with the municipality of Itapeva, Minas Gerais, we carried out patient analysis and, at the same time, we evolved the algorithms to meet the needs of health professionals and also expand support in tracking COVID-19 in the municipality. In this project we will describe cases and even signs and symptoms that were similar to the clinical performed by the doctor. The average recognition accuracy of COVID-19 was 0.97,4 ± 0.043.

  Keywords

Probabilistic Models, Machine Learning and Computer Vision and case studies.