Machine Learning for COPD Diagnosis Through Breathing Patterns Using Complex Network

Respiratory diseases are escalating swiftly in the present era, causing much suffering among many. Lung sounds emerge as crucial indicators of human endurance. Breathing, an auditory manifestation, may be harnessed to discern various respiratory conditions. In this research, we present a Machine Learning approach for extracting discernible attributes from breathing sounds to diagnose Chronic Obstructive Pulmonary Disease (COPD), employing Complex Network analysis and Wavelet Transforms. A comparative analysis of six distinct Machine Learning algorithms determines the most effective model. The Wavelet family demonstrates optimal classification performance, achieving a peak accuracy of 99.67%.