Detection And Diagnosis of Covid-19 Using Pnuemonia

Abstract

The corona virus 2019 (COVID-19), originated in Wuhan city, China, with its rapid spread around the globe,
has become a pandemic, drastically affecting the daily lives, welfare and global economy. So, it becomes
critical for quick and primary detection of the virus in the system to prevent further spread and loss of lives.
The dominantly used screening technique for detection is Polymerase Chain Reaction (PCR). With rapid
growth in the number of cases the scientists and scholars are looking into an substitute and also more
effective ways for screening, as sometimes these PCR kits turn out to be faulty with false results. This calls
for a more automated method for initial screening of patients also the aspect of expense to be considered
for the same. So we propose a fully automated end to end system for first screening and detection of covid-
19 as well pneumonia via X-ray scans of the individual and it will be done using Deep Learning techniques. In
this project we have surveyed the various datasets available from different open source resources which are
often updated by scientists and radiologists. This data set are used for Training as well as Testing of the data,
with the various image modalities like X-ray scans and CT scans the CNN model is trained, for primary purpose
the X-ray scans are used as image modalities and data. The Model is made of various CNN layers, Pooling is
done for feature extraction, and as we have multi classifications such as Covid-19, Pneumonia and Nonfindings
the Softmax layer is used for obtaining the results.