Deep Learning System Detection of COVID-19

Main Article Content

Pooja. A, Shaalini. R, Lathamary. A, Sivapriya. P, Shruthi Parathasarathy

Abstract

Infectious disease like corona virus (COVID-19) surveillance is crucial. Artificial intelligence (AI) is well adapted for identifying patterns to curb worst effects of COVID-19. Corona infection becoming an common complication nowadays can be diagnosed using deep learning system by the analysis of a CT(Computed Tomography) scan in less than sixty seconds with accuracy as high as 92% and a recall rate of about 95% on test data sets as compared to rapid diagnostic tests. It is proceeded with the help of an open-source deep learning platform Google Colab, where Convolutional Neural Network (CNN) methodology analyzed CT images for the detection of lung disorders. This work elucidates about different architectures like scratch program, Densenet, VGG-16 that are programmed for classifying normal cases, positive COVID-19 cases and positive pneumonia cases providing different results in terms of accuracy.

Article Details

Section
Articles