Tracking of cerebral vessels in MR angiography after highpass filtering

Maximum intensity projection (MIP), the commonly used technique for calculating MR-angiograms from threedimensional (3D) datasets, often fails to visualize small vessels due to superposition of background signal. An improved vascular depiction can be reached by application of a connectivity algorithm, but a very good delimitation of vessel signals from signals of stationary tissue is required for this technique. Results of intracranial vessel tracking could be improved by a 3D interpolation and a preceding highpass filtering of the 3D image dataset.