Ensemble CNN Models for Accurate Pneumonia Classification in Chest X-Ray Images

Pneumonia is a life-threatening respiratory disease, particularly in children and the elderly. Early detection through chest X-rays enables faster and more accurate treatment. This study presents a method using Convolutional Neural Networks (CNN) to classify lung X-ray images as either normal or pneumonia-infected. We compare the performance of two CNN models, InceptionResNetV2 and MobileNetV2, and introduce an ensemble approach that combines both models. The ensemble model significantly improves classification accuracy, achieving 92.47%, compared to 89.58% for InceptionResNetV2 and 90.38% for MobileNetV2. Additionally, the ensemble model enhances precision, recall, and F1 score. These results suggest that using an ensemble approach in medical image classification is an effective strategy for improving the accuracy of pneumonia diagnosis, contributing to the advancement of technology in disease diagnosis through medical imaging.