Pneumonia Detection Using CNN

One of the common and fatal infectious diseases around the world is pneumonia, and it is one of the main causes of death. According to the World Health Organization, one-third of total deaths is recorded as due to this infectious disease. Pneumococci are a unit, a bacterial strain belonging to the species Streptococcus pneumoniae that remains the most important causal pathogen. Swift diagnosis, especially in faraway regions, is indispensable since it is an essential condition of proper treatment. This paper will present an automated system for the detection and classification of pneumonia with chest X-ray images according to three classes: origin, Normal or viral pneumonia or bacterial pneumonia. The model applies the method of convolutional-neural-networks and a pretrained architecture design of ResNet50 in order to make its functioning. In this study, a trained ResNet50 model was used for the classification of chest X-ray images. This will aid the radiologist in making accurate medical decisions because the model will classify the specific kind of pneumonia present. Some of the performance metrics that can be used to evaluate the model’s usefulness include test accuracy, F1 score, recall, precision, and support. Off this pneumonia dataset, the model depicts a very high test F1 score, which means it could label cases of pneumonia with high accuracy. The results show that the created model can be potentially applied as a useful tool in declaring pneumonia assistance for the radiologist in scenarios characterized by scarce resources. Pneumonia, CNN, X-ray images, ResNet50, convolutional neural network, precision, F1 measure, recall, accuracy.