COVID Detection Using Chest X-ray Images & Deep Learning CNN Mechanism

Abstract

As we know that how rapidly is corona virus spreading, starting from China, to all over the world. There are more than 5 million suspected cases and almost 0.33 million deaths according to the statistics of world meterinfo, and there are limited test kits available in hospitals because of cases are increasing rapidly on the daily bases. So, it is compulsory to build an authentic automatic detection system which gives maximum high performance to check whether the patient is Covid-19 suspect or not as an alternative to slow down the spread of coronavirus among people.In this research, we have used deep learning’s (DL) Convolutional neural network (CNN) and ResNet models. With a critical analysis, we conclude that every ResNet layered model has the high performance with error rate less than 3% on approximately all kinds of datasets of chest Xray images whether it includes rib shadow & clivade or after segmentation. We have proposed a new solution for existing model and to enhance the ResNet model by applying layered architecture style by adding more layers to our ResNet which will help to minimize the error rate. Further, to boost the performance of ResNet by tune up the batch size and learning rate, we achieve the learning rate 0.00001 that has higher accuracy as compared to the other learning rates 0.1, 0.01, 0.001 and 0.0001. The proposed study is promising framework for the covid detection that assist us to deal the COVID decease