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.
