A hybrid model for predicting COPD using CNN

The chronic inflammatory lung illness known as chronic obstructive pulmonary disease (COPD) restricts airflow from the lungs. The signs and symptoms include wheezing, coughing up mucus (sputum), and difficulty breathing. It is frequently brought on by prolonged exposure to irritant gases or particulates, most frequently from cigarettes. Heart disease, lung cancer, and a number of other diseases are more likely to occur in people with COPD. Convolutional neural networks (CNNs) have achieved successful outcomes in the categorization and analysis of medical image data using artificial intelligence (AI) approaches. In this study, a deep CNN architecture is proposed for the diagnosis of COPD based on the classification of a person’s breathing patterns and chest X-ray pictures. The dataset was preprocessed in phases using several methods to produce a useful training dataset for the proposed CNN model to achieve its best performance. This was done to deal with the difficulties of a dataset that was out of balance and had poor image quality. The proposed hybrid model was compared with the existing systems and a very promising accuracy was achieved in prediction.