Research on an Intelligent Pneumonia Diagnosis Model Based on Improved Convolutional Neural Networks Using Chest X-ray Images

Pneumonia, a common health concern today, requires early and accurate diagnosis. Chest X-ray examinations play a critical role in the early detection of pneumonia. To enhance diagnostic accuracy, this study utilizes a deep learning-based convolutional neural network (CNN) model, trained on a dataset of 5,216 chest X-ray images obtained from pediatric patients aged 1-5 years at Guangzhou Women and Children's Medical Center. Among these, 3,875 images show signs of pneumonia and 1,341 images are normal, serving as the training and testing data for the model. By incorporating Dropout techniques and Batch Normalization methods, the models robustness and generalization ability were significantly improved. Experimental results demonstrate that the model achieves a diagnostic accuracy of 97.83%, which will effectively alleviate physicians workload and holds substantial clinical application value.