Level Set-Based Semiautomatic Bolus Segmentation in Bolus-Chasing CTA

CT angiography (CTA) is a useful tool for diagnosis on vessel diseases. To get high quality images, a dose of high contrast material (bolus) is injected into a patient´s body. In recent years, bolus-chasing CTA has been introduced to capture better images. The CT table in bolus-chasing CTA can be moved according to the real speed of bolus. Segmenting the bolus boundary in each cross-section image automatically is a task to solve. Level set methods have become increasingly popular as a general framework for image segmentation since the early 90´s. A level set based semiautomatic segmentation is presented in this paper to capture bolus boundary. Experimental results show that the semiautomatic segmentation can get better results in a short time. Though it is not real-time now, the segmentation speed can be accelerated using parallel computation technique in the future.