The computer image analysis for the 2-D coronary arteriograms

A new method for the automatic identification of coronary border is presented. It mainly utilizes mathematical morphology to segment the blood vessels and derive a refined skeleton for tracking. Finally, a tracking algorithm incorporated with a branching-point model, match filter, skeletons of blood vessels is applied to detect the arterial border and estimate the lumen width of blood vessels. Experiments reveal that the authors' approach can not only eliminate background variations and noise, but accurately define the boundary of coronary angiograms.<
>