Respiratory Analysis –Detection of Various Lung diseases using Audio Signals

Chronic lung disease kills many people every year, regardless of age. Lung sound analysis, Heart sound analysis, and cough sound analysis has been critical diagnostic tool in pulmonary diseases detection precisely. Prior to the advent of automatic screening, manual screeningwas unable to detect lung diseases resulting from a va riety of reasons, such as hearing loss and differences in the way other physicians perceive different sounds. With the advent of computerized analysis, patients suffering from various lung diseases, such as asthma, bronchitis, emphysema, tuberculosis, and pneumonia, are treated much more effectively. These disorders include Asthma, Bronchitis, Emphysema, Tuberculosis, and Pneumonia. Symptoms include wheezing, shortness of breath, bronchi, and chronic coughing. In this work, we use a respiratory audio dataset to predict various diseases such as Asthma, Pneumonia, bronchiectasis, and many more. We took the disease diagnosis and respiratory audio datasets to implement this work, extracted features from all audio datasets, and trained a convolution neural network (CNN) algorithm model. After the training model, we can upload any new test data to predict its disease.