作者: Mingqing Chen , Hua Zhong , Yefeng Zheng , Gareth Funka-Lea
DOI: 10.1109/ISBI.2015.7164113
关键词:
摘要: Heart isolation (separating the heart from neighboring tissues, e.g., lung, liver, and rib cage) is a prerequisite to generate 3D volume visualization as an intuitive view for coronary disease diagnosis treatment planning. Previously, we proposed component-carving based approach by removing unwanted background tissues (e.g., non-cardiac structures, left atrial appendage, pulmonary veins/arteries) sequentially. However, final mask usually has many small extra pieces due difficulty model segment all tissues. In this paper, propose component-composition approach, which starts empty then adds in wanted components (coronary arteries, four chambers, aorta) segmented original image. The method can much cleaner than since only structures are added segmentation of chambers aorta more robust accurate their greater predictability.