Detection Of Pandemic Virus Covid-19 Using CNN
Abstract
The recent pandemic which originated from Wuhan, China late in 2019 has wrecked the daily lives, economic activities and also escalated the death toll to devastating figures. In the absence of a vaccine or a medicine to curb the number of positive cases, it is very important to detect the number of positive cases earlier. Since there lacks an availability of accurate testing kits to diagnose the COVID-19, there requires a need for an alternative .Artificial intelligence clubbed with radiology adds it to the call with the help of the chest X-Rays. This technique can actually reduce the need for radiologists to be available to assess the imagery. It also consumes less time for assessing and way lesser time than the present testing method; PCR. In this study the ResNet 50 architecture is made in use to automatically detect the COVID -19 positive cases. The architecture is able to achieve an accuracy of 95.95 percent for binary classification (COVID Positive vs. Negative).The usage of artificial intelligence to detect COVID-19 has proved to be in par with manual assessment by a radiologist , less time consuming and also facilitates the usage of this technique where professionals are unavailable.