Title :
Identifying vascular features with orientation specific filters and B-spline snakes
Author :
Klein, Andreas ; Egglin, Thomas K. ; Pollak, Jeffrey S. ; Lee, Forrester ; Amini, Amir A.
Author_Institution :
Dept. of Electr. Eng., Yale Univ., New Haven, CT, USA
Abstract :
The authors present several new techniques for determining vessel boundaries, or enhancing their centerline features. A bank of S-Gabor filter pairs are used to create an energy field. Vessels are represented by B-spline snakes, and depending on whether vascular boundaries or centerlines are to be determined, they are optimized on appropriate filter outputs. The optimization engine is a new dynamic programming algorithm which determines the parametric curve of least energy, and can be used to optimize B-spline curves of any order. Results are demonstrated on a number of common angiograms including preand post-angioplasty coronary angiograms, and a carotid arteriogram
Keywords :
angiocardiography; diagnostic radiography; dynamic programming; feature extraction; filters; medical image processing; optimisation; splines (mathematics); S-Gabor filter pairs bank; carotid arteriogram; common angiograms; contrast angiography; dynamic programming algorithm; filter outputs; least energy parametric curve; medical diagnostic imaging; optimization engine; postangioplasty coronary angiogram; preangioplasty coronary angiogram; vascular features identification; vessel boundaries determination; Biomedical imaging; Dynamic programming; Engines; Filter bank; Filtering; Frequency; Gabor filters; Medical diagnostic imaging; Reproducibility of results; Spline;
Conference_Titel :
Computers in Cardiology 1994
Conference_Location :
Bethesda, MD
Print_ISBN :
0-8186-6570-X
DOI :
10.1109/CIC.1994.470236