Advanced AI Algorithms for Automated Diagnosis in Healthcare: Pneumonia Detection using X-Ray and CT Scans

Pneumonia is a significant global health concern, responsible for millions of deaths annually, particularly among children and the elderly. Early and accurate detection of pneumonia is essential for timely treatment, which can drastically improve patient outcomes. This study presents a novel approach to automated pneumonia detection using a Hybrid DenseNet-MobileNet (HDM) model on chest X-ray and CT scan images. The HDM model combines DenseNet’s advanced feature extraction capabilities with MobileNet’s computational efficiency, providing a balance between accuracy and processing speed. To enhance performance, several preprocessing techniques—such as resizing, scaling, and denoising—were applied to normalize images. The Synthetic Minority Oversampling Technique (SMOTE) addressed class imbalance, while Mask R-CNN was utilized for precise lung segmentation. The HDM model achieved an accuracy of 97.99%, surpassing nine other state-of-the-art models, including DenseNet121 (93.12%) and ResNeXt50 (94.87%). Additional evaluations of precision, recall, and F1-score further confirmed the HDM model's effectiveness, with values of 96.98%, 96.45%, and 96.72%, respectively. Moreover, HDM's lightweight structure, with only 6.5 million parameters and an inference time of 40 ms per image, makes it highly suitable for real-time applications in resource-limited clinical settings. The proposed method demonstrates significant potential for improving early pneumonia diagnosis, thus enhancing patient outcomes and supporting healthcare providers with timely and accurate diagnostic tools.