Analysis of Doppler Embolic Signals

STFT or CWT for the detection of Doppler ultrasound embolic signals

“…Those parameters will respectively convey information about the blood velocity, the turbulence of the flow, and nature and the quantity of scatterers in the sample volume . From different alternative approaches reported in the literature, the Wang and Fish simulator was chosen because of the demonstrated reliable simulation of MCA background signals for a low computational effort .…”

“…The duration or the number of samples in the signal with increased power due to emboli can be measured directly from the analysis of the power variation over time, and the mean velocities of the signals from the mean frequency waveform. Those parameters can be extracted directly from the time‐domain TCD signal or from its spectral representation over time . Although signal acquisition by the TCD instrumentation is a requirement and therefore takes a fundamental role in the system, it can be defined as the starting point for the system.…”

“…size and speed of emboli), and are especially relevant in the detection of small emboli. Other time‐frequency approaches for representing Doppler blood flow signals with and without emboli have been broadly used (namely Cohen's class distributions), usually leading to the discussion of the trade‐off between time and frequency resolutions . Even so, STFT was reported to achieve good results when compared with other Cohen's class distributions .…”

“…Other time‐frequency approaches for representing Doppler blood flow signals with and without emboli have been broadly used (namely Cohen's class distributions), usually leading to the discussion of the trade‐off between time and frequency resolutions . Even so, STFT was reported to achieve good results when compared with other Cohen's class distributions . The wavelet transform (WT) has the main advantage of varying the time‐frequency aspect ratio, having good frequency and time localization for respectively low and high frequencies.…”

“…The amount of possible combination of alternatives strongly limits the optimal tuning and comparison of the methods. Furthermore, different authors use different criteria, and that may hinder comparability . The results obtained in the analysis of Doppler embolic signals have proven that values above 90% for sensibility, specificity and accuracy are achievable in ordinary conditions.…”