Diagnosis of Lung Diseases from Chest X-Ray Images Using Different Fusion Techniques

Lung diseases refer to a group of disorders that affect the lungs and respiratory system. Several factors, such as genetics, environmental pollution, infections, and smoking can these. Lung diseases include coronavirus (COVID-19), pneumonia, chronic obstructive pulmonary disease (COPD), and asthma. Lung diseases cause significant damage to lung function and lead to respiratory failure or even death. The symptoms of lung diseases can range from mild difficulty breathing to severe ones, including chest pain, bloody coughing, and shortness of breath. Early detection can increase the chances of successful treatment and improve the overall outcome for affected individuals. Artificial intelligence (AI) has demonstrated considerable potential for detecting and diagnosing lung diseases through machine learning algorithms and deep learning models. The detection of lung diseases using chest X-rays (CXRs) is demonstrated in this paper by applying feature-level fusion (FLF) and decision-level fusion (DLF) techniques. FLF involves concatenating the features from two models before the classification process. In comparison, DLF is executed after training the two models and then concatenating the results to make a single decision. The two models are DenseNet-169 and Vision Transformer (ViT-L32). On the COVID-19 Radiography database, the proposed models have been tested and trained. The data has been preprocessed using data augmentation and a blurring method. An ’Adam’ optimizer is used while compiling the model. The accuracy of the DLF is 93.3%, while the FLF achieved an accuracy of 94.54%, which is better than the accuracy of the models without fusion.