hide
Free keywords:
-
Abstract:
Identification of reference points or reference time behaviors is a crucial issue in artery/vein separation algorithms, because it is both time consuming and a possible source of uncertainties. For this study, three-dimensional datasets are acquired during the first pass of a contrast agent bolus; the time courses of the brightnesses of all voxels are separated into clusters by a modified k-means algorithm, and center points of the clusters (centroids) are automatically selected as reference time courses for a subsequent artery/vein separation algorithm. In all considered datasets, reference were correctly identified and correct artery/vein separation could be performed.