REGISTRATION OF ULTRASOUND IMAGES USING AN INFORMATION-THEORETIC FEATURE DETECTOR

摘要: In this paper, we present a new method for ultrasound image registration. For each to be registered, our first applies an ultrasound-specific information-theoretic feature detector, which is based on statistical modeling of speckle and provides that robustly delineates important edges in the image. These images are then registered using differential equations, solution locally optimal transformation brings into alignment. We describe experimental results demonstrating its effectiveness, particularly low contrast, speckled images. Furthermore, compare standard gradient-based techniques, show more susceptible misregistration.