Covid Envisage

Abstract

Covid-19 is comprehended as one of the most lethal viruses ever existed. Starting in 2019, Covid-19 being highly virulent, killed over 200 million people around the globe. The alarming situation stimulated the need for better ways of testing. The alternative to standard lab tests is some assistive tools. Covid-19 vigorously affects the human lungs. It is detectable through Chest X-rays or CT scans. Artificial Intelligence provides us with deep neural networks that make detection a frequent task. For this, the state of art techniques of deep learning allows predictive nature of modelling. Our project uses the VGG-16 model for predicting the virus. This study consists of three dimensions. First dimension enables the prediction of Covid-19 using Chest X-rays. The deep learning algorithms measures the result accuracy. The second dimension enables the prediction of Covid-19 using HRCT samples. Lastly, the system allows prediction using user symptoms. VGG-16 is a CNN architecture that is the best vision model to date. We are addressing the needs of radiologists by providing lung masking. It allows to highlight infected parts and make predictions effortless and recurring. The positive classified images indicate the virus. The training results with an accuracy of 86% emphasize the need for artificial intelligence in medical assistance.