Improving pneumonia diagnosis with high-accuracy CNN-Based chest X-ray image classification and integrated gradient

Pneumonia remains a significant global health challenge, contributing to high morbidity and mortality rates, particularly in vulnerable populations. Early and accurate diagnosis is critical for effective treatment, yet manual interpretation of chest X-rays can be time-consuming and prone to errors due to inter-observer variability. To address these challenges, this research focuses on automating pneumonia diagnosis by leveraging deep learning techniques. X-ray imaging, being a key diagnostic tool, is at the center of this study, where we aim to improve diagnostic accuracy using a carefully designed Convolutional Neural Network (CNN) architecture. Our model, trained on a curated dataset of 5,856 chest X-ray images, integrates the 'inception_v3 ' layer for feature extraction with dense layers to build a high-performing binary classification tool. To further enhance model transparency and clinical utility, we incorporate Integrated Gradients (IG) for interpreting the model's decision-making process. With 22.9 million trainable parameters, the proposed model achieves an accuracy of 97.23%. The integration of IG provides additional insights by identifying critical image regions that influence predictions, boosting user confidence in the system's diagnostic recommendations. This research advances medical image analysis by presenting a reliable, interpretable CNN model for pneumonia detection, offering promising utility for clinical practice.