Implementation and empirical analysis of Deep Learning models for COVID-19 Detection and Lung Disease Prediction

Human's inhale oxygen which is then transported to the lungs, where it is then delivered into the bloodstream. As a result, breathing entails inhaling oxygen and exhaling carbon dioxide, which a healthy individual accomplishes roughly 25000 times every day. COVID-19, Pneumonia, and Lung Opacity all cause the lungs to stop working properly, which can lead to respiratory failure. Some diseases can be fatal; COVID-19 is one of the deadly diseases that the world is now dealing with. For the detection of lung disease, different prediction models are developed using deep learning algorithms, and their performances are computed and assessed using various performance metrics. The proposed methodology comprises of analyzing the performance of the CNN model after it has been trained on a dataset of 8,462 images. During the performance training, the proposed model obtained an accuracy rate of 90.83 percent. The study also discusses how the pre-trained models VGG-16 and ResNet-50 were implemented and evaluated for the dataset. The Flask app has also been used to create a user interface that accepts the user's X-ray images as input and forecasts lung illness as an output.