Early diagnosis of respiratory system diseases (RSD) using deep convolutional neural networks

The evaluation of respiratory system disorders and their classification has been to be one of the most significant investigated topics in recent years. Medical scan dataset sizes are expanding quickly in order to capture diseases in hospitals, along with a variety of medical imaging applications in hospitals, diseases, and diagnostic centers. Even though there have been numerous studies on this particular subject, the field is still unclear and difficult. This paper illustrates the use of a Deep Convolution Neural Network (DCNN) probabilistic model to analyze Respiratory System Disease (RSD). It entails a methodology for classification and estimation based on the recognition of normal and diagnostically significant CXR image features related to the RSD. The proposed DCNN was compared to standard classifiers with the introduced Composite Hybrid Feature Selection (CHFS) extraction model. The longitudinal trial with actual evidence shows the feasibility and prospects for the above-mentioned cause of the proposed solution. The suggested DCNN structure achieved an accuracy of 98.20% for potential events, which corresponds to the regular classifiers’ RSD events.