Detection of Doppler Microembolic Signals Using High Order Statistics

Figure 2

(a) The Doppler energy signal. An empirical threshold is applied to obtain the microembolic standard detection. (b) Skewness signal calculated from the windowed energy signal. A data-based threshold is applied to complete the microembolic detection. The mean value of the skewness signal is 0.7. (c) Kurtosis signal calculated from the windowed energy signal. A data-based threshold is applied to complete the microembolic detection. The mean value of the kurtosis signal is 3.2. Moreover, we choose in (b) and (c) three time positions: during which an embolus is present and and when no embolus is present. We detect, in the case of absence of embolus, and , while in the presence of embolus we detect and

(a)

(b)

(c)