Chest X-rays are vital for diagnosing chest-related disorders. The global health threat posed by lung diseases is increasing day by day. This can be identified efficiently through the design of an effective deep learning model for accurate classification of illness from chest radiography images (CRIs). Conventional diagnostic procedures are limited, necessitating automated systems for precise illness classification. Leveraging pre-processing techniques such as Lerch transcendent functions (LTFs) for noise removal and CLAHE for image enhancement, the model employs intensified feature engineering, combining CNN-based feature extraction with Salps’ swarming mechanism for optimal feature selection. The novelty lies in this intensified feature engineering approach, integrating deep learning architectures, including ResNet152V2 and Bi-GRU, with enhanced Aquila Optimization. The proposed model demonstrates remarkable performance in accurately classifying various lung diseases, offering a promising automated diagnostic solution for improved healthcare outcomes.