Detection and classification of lung diseases using deep learning

Lung disease is one of the diseases that can be cured when the disease is spotted in its early stages before getting accumulated. Early diagnosis can be promising to curable. But the most people fail to detect their disease before it comes to chronic. It leads to an increase in the death toll all around the world. Image Classification is the process of extracting information from an image that plays a major role in medical image classification. Convolutional Neural Network (CNN) model a type of deep learning architecture introduced to achieve the correct classification of lung disease. The proposed technique experimented on the Covid19 CXR dataset from the GitHub repository. The dataset has been gathered which comprises normal CXR and Covid19 CXR. The method employed is ResNet50 and DenseNet as a pre-trained classifier model along with softmax and sigmoid activation function respectively. The ResNet50 model provides promising results with a validation accuracy of 86.67 % and the DenseNet provides a validation accuracy of 98.33 %. The performance of each model is compared based on validation loss and validation accuracy. The classification concludes that the DenseNet model is more efficient than the ResNet50 model.