Pneumonia Disease detection using the convolutional neural network

Abstract

A bacterial illness called pneumonia causes inflammation in the air passages with one or even both lungs. The disease can range from mild to life-threatening. Diagnosing the disease at an earlier stage is crucial for the successful recovery of the patient. In this study, we analyze and compare various deep learning algorithms for lung illness identification and propose an updated model for pneumonia detection. The model is implemented to test its efficacy. The convolutional neural network is fed 5856 chest X-ray images split into 3 categories: training, test, and validation. Two chest conditions, namely pneumonia and normal, were detected and classified. The CNN model, trained with these datasets, achieved 94.66% training accuracy and 91.83% validation accuracy. Moreover, we also run some pre-trained models. They are: Resnet50, Inceptionv3, EfficientNet B0, Xception and VGG16,EfficientNet B6. We gained 68.91%, 83.71%, 62.50%, 91.35%, 90.75% and 62.50% accuracy respectively from them. Hence, We can observe that what was suggested. In these experimental results, the CNN model fared better than them.