Separation of crackles and squawks from vesicular sounds using a wavelet-based filtering technique

An automated way of revealing the diagnostic character of discontinuous adv entitious sounds (DAS), i.e. crackles and squawks, by isolating them from v esicular sounds (VS), based on their nonstationarity, is presented in this paper. The proposed algorithm combines multiresolution analysis with hard t hresholding in order to compose a wavelet-based stationary-non-stationary f ilter (WTST-NST). Applying the WTST-NST filter to fine/coarse crackles and squawks, selected from three lung sound databases, the coherent structure o f the DAS is revealed and they are separated from VS. When compared to othe r separation tools, in noiseless case, the WTST-NST filter performed more a ccurately, objectively, and with lower computational cost. Owing to its sim ple implementation it can easily be used in clinical medicine.