Determining lung sound characterization using Hjorth descriptor

Lung sounds provide important information about the health of the lungs and airways. Lung sounds have a special and distinguishable pattern related to abnormalities that might occur in the lungs or respiration tract. Automatic lung sound recognition is directed to reduce subjectivity in assessing lung sounds. Hjorth descriptor is one method used for observing natural biological signals. Hjorth descriptor was measured to reveal biological signal's complexity. In this paper, lung sound characteristics were measured using Hjorth descriptors. Hjorth descriptor will be used to look at complexity of lung sounds. Hjorth descriptors of lung sounds are measured in the time domain and the frequency domain, and clustered using the K-means clustering method. These parameters were tested as to whether they could function as features in automatic lung sound recognition. The experimental results show that the Hjorth descriptor in the time domain is promising in order to be used as a feature for lung sound classification.