Vascular Territory Segmentation Using Mutual Clustering in Image and Label Space

Keywords

able actually applied arisen arteries artery assumption assumptions audience automated basal basis becoming biomedical blue branches branching capability centroid centroids cerebral circle clinicians cluster clustered clustering clusters complicated computed connectivity contaminated contiguous continuous coordinates cortical depending description detection determine diameters distance distances divided editing enables encoded engineering estimation even finally five flight forest ganglia generate generated green growing guarantee identified includes inferior inspection institute intended intensity knowledge label labeling lateral laterally leaving limited location locations magma manual manually mapped mapping maps mass medicine middle moving must mutual needs neighboring noise north offset often orange overlaid overlay perfusion placement potentially priori projection propose proposed pseudo purple radiology regarded rejected rejects required resolution resolve respectively school science scientists segment segmentation segmented segmenting separate separation smaller source sources space spatial spread stage stem step steps superior supply surviving target territories territory thalamus translational turn turns upper validated vascular verification vessel volume wake wide wise