Crackles detection method based on time-frequency features analysis and SVM

Crackles, which are a kind of abnormal lung sounds, are used as indicators for the diagnosis of pulmonary diseases. In this paper, an automatic and noninvasive method is presented for crackles detecting. This method mainly comprises three steps: preprocessing, features extracting and crackles detecting based on support vector machines(SVM). The features are f_min/f_max of the frequency limbic signal, standard deviation of the time limbic signal and the smoothing time limbic signal. The simulation result of the two testing groups shows that the accuracy(AC) are 97.14% and 100%, respectively, which indicates that this method could be an efficient way to detect crackles.