“…However, ultrasound image is characterized by the multiplicative speckle noise and low signal-to-noise ratio (SNR), wherefore traditional edge detection operators are not suitable for its interpretation. 3 Over time a variety of anti-noise edge detection methods have been proposed, such as multiscale-based methods, [4][5][6] fuzzy rules-based methods, 7,8 adaptive threshold-based methods, 9,10 and machine learning-based methods, 11,12 which could improve the detection accuracy and noise robustness to a certain extent. However, most of these methods applies additive operators, which are not suitable for ultrasound images with multiplicative speckle noise.…”