An optimal solution based on image fusion for blood vessel diameter measurements in cerebral angiographies

Abstract

The main goal of this study is an accurate extraction of the brain vascular structure to improve detection and to enhance the blood vessels in the human brain. Data provided by Canny and Frangi edge detection algorithms are evaluated and then these disparate data are combined with a fusion algorithm for comparison purposes. This investigation is performed in an edge detection framework. The performance of the proposed methods is evaluated by conducting measurements on the diameter of the brain blood vessels in cerebral angiography images. The average diameters of the blood vessels provided by Frangi, Canny and the fusion algorithms are measured. To highlight the improvements brought by data fusion, an objective evaluation measure based on edge structural similarity and error rate have been made. Also, the superiority of the fusion algorithm in terms of image quality is proved.