Automatic border detection in intravascular iltrasound images for quantitative measurements of the vessel, lumen and stent parameters

The automated segmentation of lumen, vessel and stent boundaries in intravascular ultrasound (IMS) images will reduce the required analysis time and the subjectivity of the commonly used manual tracing procedure while the three-dimensional reconstruction permits an advanced assessment of the morphology. We describe a knowledge and model guided system for the (semi-) automatic contour detection of these borders. The different steps of a complete analysis of an IVUS pullback sequence are carried out such, that the different components assist each other. The stent detection assists the vessel detection and both detections assist the lumen detection. The results of the detection show a very good correlation with manually drawn contours. Due to the flexible use of more than two longitudinal cut planes and the advanced knowledge-guided contour detection approach, the new IVUS analysis system has proven to be suitable for clinical research-studies.