Three-dimensional reconstruction of the arterial lumen.
Item
-
Title
-
Three-dimensional reconstruction of the arterial lumen.
-
Identifier
-
AAI9020814
-
identifier
-
9020814
-
Creator
-
Suardiaz, Manuel.
-
Contributor
-
Adviser: Joseph Barba
-
Date
-
1990
-
Language
-
English
-
Publisher
-
City University of New York.
-
Subject
-
Engineering, Biomedical | Engineering, Electronics and Electrical
-
Abstract
-
An accurate method for the three-dimensional (3-D) reconstruction of the arterial lumen from two orthogonal digitally subtracted X-ray angiograms was developed. 3-D reconstruction was obtained by multiple two-dimensional (2-D) reconstructions of vessel cross-sections at a sequence of transverse cuts across the vessel main axis. The arterial lumen cross-section was reconstructed using an iterative probabilistic algorithm which introduces multiple constraints to reduce the ambiguity in the reconstruction. Also, two algorithms for the automatic determination of arterial center line position in 3-D space were developed. While the first 3-D positioning algorithm is suitable for center lines with single-valued projections, the second one can work with both single and multi-valued projected center lines. These 3-D positioning algorithms use cubic spline interpolation (CSI) techniques to obtain a piecewise functional representation of the projected center line on each of the two orthogonal views. Cross-sectional reconstructions of computer simulated stenotic arterial segments with crescentic lumen shapes are presented. The accuracy of the cross-sectional reconstruction algorithm when applied to computer simulated density profiles in the presence of additive Gaussian noise of different magnitudes was investigated. The sensitivity of this method to boundary constraint error was also evaluated. Moreover, the cross-sectional reconstruction error of the algorithm when applied to actual X-ray images of a physical model with a crescent-shaped lumen is presented. The accuracy of the 3-D positional algorithms when reconstructing different computer simulated center line shapes was also examined. In addition, 3-D positional accuracy in the presence of both input data error and a stenosed segment of increasing magnitude was also studied. Application of these algorithms to biplane images of coronary arteries may permit increased understanding of three-dimensional arterial lumen shape and consequent hemodynamic significance of atherosclerotic lesions.
-
Type
-
dissertation
-
Source
-
PQT Legacy CUNY.xlsx
-
degree
-
Ph.D.
