Estimation of Coronary Artery Dimensions by Median Filtering Approach

Isam M. EL-NAQA  Mohammad J. MISMAR  Isam H. ZABALAWI  

IEICE TRANSACTIONS on Fundamentals of Electronics, Communications and Computer Sciences   Vol.E79-A   No.7   pp.1034-1043
Publication Date: 1996/07/25
Online ISSN: 
Print ISSN: 0916-8508
Type of Manuscript: PAPER
Category: Digital Signal Processing

Full Text: PDF(934.3KB)
>>Buy this Article

A new algorithm based on the median filtering approach is presented for computerized assessment of the severity of coronary artery obstruction from single-view angiograms. The median filtering approach is used because median filters have excellent properties in the presence of noise and preserve the fine details of the image. These features are exploited to optimize the estimation of coronary vessel dimensions. The proposed algorithm consists of four distinct stages: image enhancement, contour tracking, edge detection, and post-processing smoothing. The algorithm is tested on synthetic data, vessel phantoms, and real coronary angiograms. The experimental outcomes of the proposed median-based algorithm has demonstrated excellent performance and encouraging diagnostic results in comparison with currently existing methods.