Share Email Print

Proceedings Paper

Erode/dilate analysis of micro-CT images of porcine myocardial microvasculature
Author(s): Timothy L. Kline; Yue Dong; Mair Zamir; Erik L. Ritman
Format Member Price Non-Member Price
PDF $17.00 $21.00

Paper Abstract

Analysis of 3D images of vascular trees presents a major logistic and multi-scale imaging challenge. One approach that greatly reduces the image analysis difficulty is to apply an 'erode/dilate' approach to a binarized, segmented, image so as to progressively eliminate branches of increasing diameter. Although this provides useful data for detecting some changes in branching geometry, it eliminates information about the hierarchical structure of the vascular tree. To quantify the impact of this loss of branching hierarchy information we analyzed 3D micro-CT images (4μm and 20μm isotropic voxels) of porcine myocardial "biopsies" obtained in control animals and in animals after 100μm diameter microspheres were injected into the coronary artery perfusing the site of subsequent biopsy. After the in vivo embolization, the vascular tree was injected with radiopaque Microfil and "biopsies" of the myocardium harvested. The analysis of the micro-CT images of the biopsies involved erode/dilate analysis of the opacified vessels in the entire biopsy and also of isolated vascular trees (isolated via a 'connect' function) within the biopsy. The isolated trees were also analyzed by dimensional measurement of the individual interbranch segment lengths and volumes, results that were then put into the same form as those produced by the erode/dilate method. In the embolized specimens the volume-loss of vessels below 60μm diameter closely matched for (i) erode/dilate of entire biopsy, (ii) erode/dilate of isolated tree, and (iii) direct measurement of isolated tree. The erode/dilate method quantifies the effects of a microsphere embolization, indicating what diameter interbranch segments trap a microsphere of a given size.

Paper Details

Date Published: 9 March 2010
PDF: 9 pages
Proc. SPIE 7626, Medical Imaging 2010: Biomedical Applications in Molecular, Structural, and Functional Imaging, 762620 (9 March 2010); doi: 10.1117/12.845611
Show Author Affiliations
Timothy L. Kline, Mayo Clinic College of Medicine (United States)
Yue Dong, Mayo Clinic College of Medicine (United States)
Mair Zamir, The Univ. of Western Ontario (Canada)
Erik L. Ritman, Mayo Clinic College of Medicine (United States)

Published in SPIE Proceedings Vol. 7626:
Medical Imaging 2010: Biomedical Applications in Molecular, Structural, and Functional Imaging
Robert C. Molthen; John B. Weaver, Editor(s)

© SPIE. Terms of Use
Back to Top
Sign in to read the full article
Create a free SPIE account to get access to
premium articles and original research
Forgot your username?