ML-Powered Pneumonia Diagnosis: A CNN Approach for Chest X-Ray Analysis

Pneumonia one of the common respiratory disease that significantly increases morbidity and mortality worldwide, particularly in older adults and children under five. Conventional diagnostic methods, which depend on radiologists to interpret chest X-ray pictures, frequently lead to inconsistent accuracy and delays in diagnosis. This study presents an artificial intelligence (AI) method for automatically detecting pneumonia in chest radiographs using CNNs (convolutional neural networks). The suggested framework combines two complementary models: a segmentation model based on U-Net that identifies the infected lung regions and a CNN-driven classifier that differentiates between normal and pneumonia-infected images. While the segmentation model obtained a Dice score of 0.9587 and an IoU of 0.9225, the classification network achieved a 95% accuracy rate using a dataset of 20,000 chest X-rays. Results confirm that this deep learning system surpasses traditional machine learning models and performs on par with trained radiologists. By combining classification with visual localization, the model enhances both diagnostic reliability and interpretability, offering significant value in clinical settings.