SEMIAUTOMATIC SEGMENTATION OF VASCULAR NETWORK IMAGES USING A ROTATING STRUCTURING ELEMENT (ROSE) WITH MATHEMATICAL MORPHOLOGY AND DUAL FEATURE THRESHOLDING
Bd. Thackray et Ac. Nelson, SEMIAUTOMATIC SEGMENTATION OF VASCULAR NETWORK IMAGES USING A ROTATING STRUCTURING ELEMENT (ROSE) WITH MATHEMATICAL MORPHOLOGY AND DUAL FEATURE THRESHOLDING, IEEE transactions on medical imaging, 12(3), 1993, pp. 385-392
Citations number
17
Categorie Soggetti
Engineering, Biomedical","Radiology,Nuclear Medicine & Medical Imaging
Investigation of the physical dimensions of vascular network component
s can enhance an understanding of the network's structure. The current
study demonstrates a method for measuring the spatial concentration o
f specific categories of vessels in a vascular network consisting of v
essels of several diameters, lengths, and orientations. It is shown th
at a combination of the mathematical morphology operation, opening, wi
th a linear rotating structuring element (ROSE), and dual feature thre
sholding can semi-automatically segment categories of vessels in a vas
cular network. Specifically, capillaries and larger vessels (arteriole
s and venules) are segmented in order to assess their spatial concentr
ations. The ROSE algorithm generates the initial segmentation, and dua
l feature thresholding provides a means of eliminating the nonedge art
ifact pixels. The subsequent gray scale histogram of only the edge pix
els yields the correct segmentation threshold value. This image proces
sing strategy is demonstrated on micrographs of vascular casts. By adj
usting the structuring element and rotation angles, this image process
ing strategy could be applied to other network structures where a segm
entation by network component categories is advantageous, but where th
e objects can have any orientation.