Medical ultrasound image registration is an essential component in an increasing number of applications, and has therefore been the subject of many studies in the literature. These applications use either generic registration algorithms or pixel-to-pixel comparison based ultrasound-specific methods. Hence, they are not well suited for the case of speckled images resulting from different realizations of a random process. To better handle the speckle, this work proposes an information-theoretic feature detector-based registration approach. Using speckle modeling based on the distributions of Rayleigh or normalized Fisher-Tippett, a speckle-specific information-theoretic feature detector is constructed and applied to provide feature images. Those feature images are then registered using differential equations, whose solution provides a transformation to bring the images into alignment. Compared to standard gradient-based techniques, the experimental results demonstrate the effectiveness of the proposed method, particularly for low contrast ultrasound images. It can be readily applied in the healthcare industry.
展开▼