Fallavollita Pascal, Cheriet Farida
Inst. of Biomed. Eng., Ecole Polytech. de Montreal, QUE, Canada.
Conf Proc IEEE Eng Med Biol Soc. 2006;2006:3037-40. doi: 10.1109/IEMBS.2006.260614.
A method is presented that aims at minimizing image processing time during X-ray fluoroscopy interventions. First, an automatic frame extraction algorithm is proposed in order to extract relevant image frames with respect to their cardiac phase (systole or diastole). Secondly, a 4-step filter is suggested in order to enhance vessel contours. The reciprocal of the enhanced image is used as an alternative speed function to initialize the fast marching method. The complete algorithm was tested on eight clinical angiographic data sets and comparisons with two other vessel enhancement filters (Lorenz and Frangi) are made for the centerline extraction procedure. In order to assess the suitability of our filter the extracted centerline coordinates are compared with the manually traced axis.
