Application of tonal index to pulmonary wheezes detection in asthma monitoring

In the paper a new application of tonal index, well-known signal feature from MPEG-audio psychoacoustic model, to pulmonary wheezes detection at patients with asthma diseases is proposed. Efficiency of this approach is compared with application of some other acoustic signal features which have already been tested in asthma monitoring (e.g. spectral peaks entropy, frequency ratio and kurtosis) and superiority of the tonal index is shown. Experiments have been conducted on artificially generated wheezes-like signal that have being embedded in real pulmonary noise (hybrid signal) as well as on recorded real wheezes. The SVM classifier was used in recognition process.