Deep Learning Model for Chronic Obstructive Pulmonary Disease through Breathing Sound

The lung condition known as Chronic Obstructive Pulmonary Disease (COPD) results in breathlessness, wheezing, or a persistent cough. Lack of oxygen causes the lungs to release extra carbon dioxide. The audio of the respiratory sounds is useful in the prediction of COPD since early detection is important for respiratory disorders. Initially, the diagnosis of COPD exacerbations has a lot of potential to benefit from deep learning approaches. Deep learning is an effective machine learning algorithm that combines supervised training with Convolutional Neural Network (CNN) and Long Short-Term Memory (LS TM) applications and unsupervised training to reduce optimization. The dataset of breathing sounds is from people with COPD and other six lung diseases, as well as those who are healthy. Feature extraction is done in order extract necessary features from the audio files. An artificial Convolutional Neural Network (CNN) was used to extract and classify features from the preprocessed dataset. Exacerbation forecasting could lessen their severe effects and the substantial expenditures related to COPD patients. Deep learning in breathing sounds is used to forecast whether an output will be COPD or non-COPD related. The findings demonstrate that the deep learning model was highly accurate in differentiating between people without COPD and those who have it. Also, this technique is a financially advantageous choice for long-term remote monitoring in medical applications.