Transfer learning based COVID-19 detection Using Radiological Images

ABSTRACT

Artificial Intelligence is a process that enables machines to imitate human behaviour. Both Machine Learning and Deep Learning are subsets of AI. The basic difference between ML(Machine Learning) and DL(Deep Learning) is that in Machine Learning manually defining of features is done to get the desired outcome whereas in deep learning the neural network learns of its own and publishes the result. In the present crisis due to COVID-19 pandemic the contagious power of virus has led to huge encounter of cases on daily basis. This stimulates the need for specialised and accurate methods to detect COVID-19 cases. The contribution of deep learning to this problem has been significant. The application of deep learning concepts has shown its emence importance and utility in medical domain for detection of COVID-19 cases using CT scan and X-Ray images of lungs. Our proposed method compares the accuracy of multiple pretrained models in predicting COVID-19 infected cases for a specific dataset of radiological images using three distinct optimizers for each model. This research aims to determine which model, together with its associated optimizer, is most suitable for identifying COVID-19 infected cases from radiological lungs images. © 2021 IEEE.