COVID–19 Detection from Chest X-Ray Images Using Deep Learning Techniques

Across the world, many people are affected with the outbreak of Coronavirus or COVID–19, since early 2020. The major organ where the viral strain infects is primarily the lungs which brings about a shortness of breath as reported in cases worldwide. At this juncture, we have not been able to find the vaccine or even a clinically approved cure that would help curb the same. Many cases were reported recently, which has grown in an exponential fashion across the world and claimed many lives too. We worked with a substantially small medically approved dataset with clinical correlation of the patient’s condition as was available and tried a deep learning model, a network using the VGG16 network with weights pre-trained on ImageNet dataset and fine-tuned the architecture with our training data so as to suit our need of detection based on the chest X-ray images as to whether the concerned individual has been affected with COVID–19 or not. With limited amount of data that we trained the model with, an accuracy of about 90% was achieved which may well vary as a larger sized training dataset, once available, could be fed into the proposed convolutional neural network architecture model.