Feature Extraction with Using S Transform for Classification of Chronic Obstructive Pulmonary Disease

There are some people all over the world who suffer from lung diseases. Chronic Obstructive Pulmonary Disease (COPD) is one of the most fatal diseases that kills a significant number of people each year. It is the third leading cause of death worldwide. Therefore, early detection of COPD and controlling the early stages of the disease can have a great impact on reducing the mortality of this diseases. firstly, it is necessary to mention that COPD contains 5 levels from COPD0 to COPD4. In this study, the S transform was used as a tool to extract the features of the lung signals. After extracting the features by using S transform, in order to reduce the number of features, the statistical methods are used. Then, in order to classify the different levels of this disease, SVM classification with K-Fold Cross validation is used, which provided accuracy, Sensitivity and Specificity of 92.59%, 83.33% and 95.23%, respectively.