One-click detection of coronary artery ostia from three-dimensional CTA data

In this paper, an one-click method for coronary artery ostia extraction from computed tomography angiography (CTA) scans is presented and a strict definition of coronary artery ostia in geometric graphics is given. Due to the presence of diverse anatomical structures surrounding the heart of cardiacCTA data, computerized segmentation of the coronary arteries is a challenging procedure. Many coronary segmentation algorithms need start point(s). And the proposed method can automatically locate artery ostia as start points by simply selecting any point inside the aorta of the top slice. It is much efficient andconvenient so that it is not need to search the start point(s) by slices. Region growing algorithm is firstly used to acquire the region of interest. Then the multi-scale filtering and morphologic opening operation are used, whose responds act as the input of the proposed ostia identification function. The method is developed and tested on 32 clinical datasets.And the result showed that our method is able to detect the ostia with a high detection rate of 96.88%.