作者: Christian Wachinger , Nassir Navab
DOI: 10.1007/978-3-642-04268-3_96
关键词: Weighting 、 Matching (graph theory) 、 Signal 、 Ultrasound 、 Transducer 、 Ultrasound imaging 、 Phase (waves) 、 Feature (computer vision) 、 Artificial intelligence 、 Scattering 、 Viewing angle 、 Computer vision 、 Computer science
摘要: We address the problem of viewing-angle dependency ultrasound images for registration. The reflected signal from large scale tissue boundaries is dependent on incident angle beam. This applies an implicit weighting image, viewing-angle, which negatively affects registration process, especially when utilizing curved linear transducers. show that a simple reweighting images, considering common physical model imaging, not feasible. therefore introduce new matching function, separating reflectivity and scattering regions, are results two different types interactions beam with tissue. use local phase identifying regions reflectivity, consider it as one part our combining feature- intensity-based aspects. First experiments provide good this novel approach.