作者: Zeynettin Akkus , Diego DB Carvalho , Stijn CH van den Oord , Arend FL Schinkel , Wiro J Niessen
DOI: 10.1016/J.ULTRASMEDBIO.2014.10.004
关键词:
摘要: Abstract Carotid plaque segmentation in B-mode ultrasound (BMUS) and contrast-enhanced (CEUS) is crucial to the assessment of morphology composition, which are linked vulnerability. Segmentation BMUS challenging because noise, artifacts echo-lucent plaques. CEUS allows better delineation lumen but contains lacks tissue information. We describe a method that exploits combined information from simultaneously acquired images. Our consists non-rigid motion estimation, vessel detection, lumen–intima media–adventitia segmentation. The evaluation was performed training (n = 20 carotids) test (n = 28) data sets by comparison with manually obtained ground truth. average root-mean-square errors were comparable for (411 ± 224 393 ± 239 μm) (362 ± 192 388 ± 200 μm), inter-observer variability. To best our knowledge, this first perform fully automatic carotid using CEUS.