Covid 19 identification model using Deep Learning techniques from thorax of lung X-ray images

Fidel H.-R.; Alejandro A.-T. · Proceedings of the LACCEI international Multi-conference for Engineering, Education and Technology (2024) · DOI: 10.18687/laccei2024.1.1.1491

Resumen

The pandemic, known as COVID-19, has preoccupied the world since December 2019. To date, it has had an impact on patients. In the fight against this disease, which causes pneumonia and provokes in symptoms from dry cough to fever, tools to urgent requires to provide experts with an efficient diagnosis to identify patients infected with COVID-19. The study aims to determine the diagnostic process model of COVID-19 by applying deep learning techniques from chest X-ray images of patients' lungs. By means of chest X-ray, Have 21165 patient images and those infected with COVID-19 are selected; subsequently, preprocessing and processing is carried out, using convolutional neural network models such as VGG19, DenseNet169, ResNet101 and EfficientNetB0. The performance of the algorithms is also evaluated and compared using algorithm performance metrics. They have been coded using the Python programming language. The best performance results of the four models evaluated are VGG19 DenseNet169 and EfficientNetB0, with 99% precision and recall in the implementation of the algorithms.

Instituciones: Universidad Nacional del Altiplano