Enhancing Pneumonia Detection Through Deep Learning: A Comparative Study of CNN

Pneumonia is a severe casing of respiratory sickness that requires immediate demonstration so that correct intervention can be applied. More specifically, with the advance in Deep Learning technology in recent years, specifically Convolutional Neural Networks (CNNs), automatic extraction of pneumonia from the chest X-ray images has become almost a possibility. In the process of this research, the authors explore the deep learning techniques utilising CNN-based models in the identification of pneumonia. Thus, we have given the performance comparisons of several CNN architectures with the help of graphs displaying Training & Testing Accuracy, Precision, Recall and F1 Score. The results demonstrate much better performance of the models in comparison with diagnostic accuracy and, thus high precision and recall values that can be acknowledged as the potential ability of the models in clinical decision support. This work contributes to knowledge in the area of healthcare AI and offers a strong argument for how CNNs can help in the acceleration of pneumonia identification; the general objective being to advance patients’ quality of life. The current study focus on applying deep learning models to diagnose pneumonia on chest X-ray images. In this research we have used CNN & VGG-16 architecture and the testing accuracy of these using the studied dataset is stated in the paper. Therefore, the study proves that deep learning methodologies improve the ability to recognize pneumonia and offer an imperceptible positive improvement to identify viral forms.