Abstract | ||
---|---|---|
A pre-segmentation algorithm for computed tomography angiograms (CTA) is described. It is based on adaptive thresholding. The median axis of the binarized image provides arterial centerlines and estimates of the local radii. Boundary location is refined by analyzing radial intensity profiles in the vicinity of their intersections with the thresholds. The thresholding rule was learned on CTAs of 60 patients. The method was successfully evaluated in 10 patients. |
Year | DOI | Venue |
---|---|---|
2004 | 10.1007/1-4020-4179-9_52 | COMPUTER VISION AND GRAPHICS (ICCVG 2004) |
Keywords | Field | DocType |
Medical imaging, 3D images, segmentation, centerline | Computer vision,Segmentation,Artificial intelligence,Computed tomography,Thresholding,Mathematics | Conference |
Volume | Citations | PageRank |
32 | 4 | 0.63 |
References | Authors | |
2 | 3 |
Name | Order | Citations | PageRank |
---|---|---|---|
Leonardo Florez-Valencia | 1 | 15 | 5.70 |
Fabrice Vincent | 2 | 10 | 1.29 |
Maciej Orkisz | 3 | 315 | 24.14 |