HMM based cough sound analysis for classifying pneumonia and asthma in pediatric population

摘要: Separating pediatric asthma from pneumonia is one of the major issues in remote areas. These diseases have overlapping symptoms, but require drastically different treatments. Existing guidelines for classification resource poor regions The World Health Organization call use bronchodilator test to separate pneumonia. However, an expensive conduct and not easily available In this study, we propose innovative novel technique using cough sound analysis cases asthma. work paper analyzed data 20 subjects (10 10 patients). Using mathematical features sounds, HMM classifier was trained identify pneumonic asthmatic cough. Then by computing Pneumonic Cough Index each patient classified as either into or Proposed method achieved accuracy 90% (sensitivity = 100% specificity 80%) classifying patients. Our results indicate that carry critical information which can be used patients shows potential become alternative areas world.