Reconstruction of vessels from MR-angiography data

Today pathologic changes in the blood vessel system belongs to the most frequent causes physical disturbances (heart attack, stroke etc.). Therefore it is very important for the medical diagnostic to registrate arteriosclerotic vessel stenosis.

Imageproviding techniques for the representation of blood vessels are in general summerized under the generic term Angiography.

The purpose of this project was to research the possibilities of knowledge- and modell-based methods in digital image processing and pattern recognition for MR-angiography and to create a larger MR-angiographical evaluating program. After a prototype had been finished the system was tested in a clinical environment. Hereafter, the positive results led to an integration into the post-processing tool of SCR (Siemens Corporate Research, Princeton, New Jersey) under Windows NT.

total view segmentation
Figure 1: Volume rendered data set (screenshot provided by SCR) Figure 2: Segmented vessel tree with centerlines (screenshot provided by SCR)

Project staff

Project partners