Abstract: The non-destructive detection of circulating emboli with Doppler ultrasound technique is of active signifi- cance in the clinical application.In order to eliminate the drawbacks of artifacts brought by the movement of probes or patients and detect emboli accurately,relevant feature parameters are extracted from two angles of the wavelet transform of Doppler signals.Firstly,to analyze the singularity of the signal waveform through its wavelet scalogram,transverse and longitudinal parameters are extracted to represent the scalogram characteristics.A novel method is then proposed based on the adaptive wavelet packet basis,from which several parameters such as energy,scale,etc.are extracted to represent the optimized signal approximation features.With all the feature parameters in the two aspects,a classification system is established for Doppler Ultrasound embolic signals by solving the generalized Fisher discriminant plane.From experiments on 300 cases simulated and 298 cases clinical Doppler ultrasound signals of cerebral arteries,it is shown that the proposed system can achieve the emboli detection rates of 99.0% and 98.5% for the training set and the testing set respectively.Therefore this method makes great improvement of emboli detection compared to traditional methods and has the possibility to be applied in the automatic detection of clinical emboli.