Lung Sounds Ventilation Cycle Segmentation and Classify Healthy, Asthma and COPD
Author(s): Gunes Harman
The identification of the ventilation cycle holds significant importance in clinical cases, as it provides crucial insights into respiratory patterns. Each breath- ing cycle possesses unique characteristics closely related to specific pathological information. This phenomenon plays a vital role in the accurate diagnosis of respi- ratory diseases and aids in making informed decisions regarding the overall health status of the subject. In this particular study, the classification of subjects as either healthy or pathological Asthma or COPD was performed using two popu- lar machine learning algorithms: Artificial Neural Network (ANN) and K-Nearest Neighbor (KNN). These algorithms were employed to analyze the inspiration and expiration phases of the respiratory cycle, extracting valuable information for the classification task. By leveraging the distinctive features observed during these phases, the algorithms aimed to accurately categorize individuals based on their respiratory health. The classification results obtained from the ANN and KNN algorithms were evaluated using sensitivity, specificity, and accuracy met- rics. Additionally, this study compares the actual lung sound signals that have been labeled and categorized by medical professionals (Class 1) and automatically segmented lung sound signals (Class 2). These signals are obtained through an automated segmentation process that extracts specific portions of the recorded lung sounds based on predefined criteria.