Title :
Derivative-based edge detection in quantitative coronary angiography is not independent of vessel size
Author :
Herrington, D.M. ; Siebes, M. ; Walford, G.D. ; Selzer, R.H.
Author_Institution :
Johns Hopkins Hospital, Baltimore, MD, USA
Abstract :
Current applications of derivative-based detection in quantitative coronary angiography typically rely on the assumption that a scale factor derived from the image of an object of known diameter (e.g. the catheter) is applicable over a wide range of vessel sizes. To test this assumption, the authors measured pixel diameters and absolute diameters of 10 plexiglass phantoms imaged under conditions simulating clinical angiography, using a quantitative-coronary-angiography system with a derivative-based edge operator. They found a curvilinear relationship between true and pixel diameters, suggesting that the relationship between density profile gradients and the locations of the true edge is not invariant, but rather changes with size of the vessel
Keywords :
cardiology; diagnostic radiography; picture processing; catheter; curvilinear relationship; density profile gradients; derivative-based edge detection; pixel diameter; plexiglass phantoms; quantitative coronary angiography; vessel size; Angiography; Catheters; Hospitals; Image edge detection; Imaging phantoms; Laboratories; Lesions; Propulsion; Quantum cellular automata; Time measurement;
Conference_Titel :
Computers in Cardiology, 1988. Proceedings.
Conference_Location :
Washington, DC
Print_ISBN :
0-8186-1949-X
DOI :
10.1109/CIC.1988.72633