A method is disclosed for identification of a contrasted
blood vessel in digital 3D image data. The method includes: a) one image
voxel, which is associated with the
blood vessel, in the 3D image data is selected as a seed point, b) a
current threshold value for the intensity of image voxels is selected as the intensity of the seed point, and the seed point is associated with the threshold value, c) a test criterion for finding new seed points is used to search for new seed points in the image voxels which are adjacent to the seed point, and are associated with specific threshold values as a function of their intensities and the
current threshold value, d) step c) is repeated until no further seed points can be found with the
current threshold value, e) the seed points which have been found are stored as vessel voxels associated with that
blood vessel, and are associated with the current threshold value, f) the current threshold value is reduced, g) steps c) to f) are repeated until a termination criterion is satisfied, h) the blood vessel is identified as the set of all vessel voxels in the image data, in which prior knowledge exists, on the basis of which a corresponding method step specifically linked to the blood vessel to be identified can lead to incorrect identification, and at least one of the method steps a) to h) is modified or extended, taking into account the prior knowledge.