A New Method to Improve the Quality of Embolic Ultrasound Signal Detection

Abstract

Emboli detection and denoising have been the focus of attention from many years. Emboli can be occurred because of the air bubbles, concentrated fat particles or blood clotted in a vessel. The detection of this issue can be possible by employing Doppler ultrasonic transducers. However, this method would be very sensitive to peripheral noise. In this paper for emboli denoising, CEEMDAN as one of the newest and effective methods is introduced. Then, an algorithm for emboli detection based on Bayesian estimation is proposed. The results are compared with a recent method, showing that the proposed method yields an SNR that this improved by 45% and an accuracy rate of 94%.