Classification of chest X-ray diseases using multi-layered neural network with feature selection based on AAPSO algorithm

In recent decades, the integration of Computer-Aided Diagnosis (CAD) systems is crucial in such contexts, offering a cost-effective alternative to expert medical assessment and addressing disease detection concerns in resource-constrained settings. This thesis presents an innovative machine learning model for multi-class chest X-ray diagnosis. Unlike previous binary classification studies, the classification of 15 distinct classes is uniquely addressed., demonstrating a broader and more complex application in disease detection. This work highlights the NIH chest X-ray dataset and the mechanism of integration with data preprocessing steps. For instance, image resizing, augmentation, and segmentation. On the other hand, deep feature extraction is carried out by applying the ResNet architecture, with feature selection optimized through AAPSO. An improved EfficientNetB0 design, expanded with 12 extra layers, was actualized to realize prevalent execution. Using exchange learning, an amazing precision of 87.9% was achieved, altogether boosting the effectiveness of chest X-ray infection classification. The proposed method illustrates the critical headway over conventional chest X-ray contamination classification methods, underscoring its capability to revolutionize diagnostics in settings with restricted assets.