Robust and Accurate Coronary Artery Centerline Extraction in CTA by Combining Model-Driven and Data-Driven Approaches

Abstract

Abstract. Various methods have been proposed to extract coronary artery centerlines from computed tomography angiography (CTA) data. Almost all previous approaches are data-driven, which try to trace a centerline from an automatically detected or manually specified coronary ostium. No or little high level prior information is used; therefore, the centerline tracing procedure may terminate early at a severe occlusion or an anatomically inconsistent centerline course may be generated. Though the connectivity of coronary arteries exhibits large variations, the position of major coronary arteries relative to the heart chambers is quite stable. In this work, we propose to exploit the automatically segmented chambers to 1) predict the initial position of the major coronary centerlines and 2) define a vessel-specific region-of-interest (ROI) to constrain the following centerline refinement. The proposed prior constraints have been integrated into a model-driven algorithm for the extraction of three major coronary centerlines, namely the left anterior descending artery (LAD), left circumflex artery (LCX), and right coronary artery (RCA). After extracting the major coronary arteries, the side branches are traced using a data-driven approach to handle large anatomical variations in side branches. Experiments on the public Rotterdam coronary CTA database demonstrate the robustness and accuracy of the proposed method. We achieve the best average ranking on overlap metrics among automatic methods and our accuracy metric outperforms all other 22 methods (including both automatic and semi-automatic methods).

Precise Lumen Segmentation in Coronary Computed Tomography Angiography

“…The graph-cut solver was implemented in C++ as described in the original work [1]. We used CTLs obtained from the automatic tracking algorithm described by Zheng et al [14]. Qualitative results are presented for pathological data from the training set in Fig.…”

CTA Coronary Labeling through Efficient Geodesics between Trees Using Anatomy Priors

“…We evaluated our labeling method on 20 patient cases of automatically detected coronary centerlines [11] that combines a model driven approach for the three major coronary arteries with a data driven approach for the side branches and distal parts of the main branches. Because the method of [11] knows which of the three main coronary branches it has recovered, the label of these branches is known as part of the segmentation.…”

Improving accuracy in coronary lumen segmentation via explicit calcium exclusion, learning-based ray detection and surface optimization

“…Our segmentation algorithm yields the 3D lumen surface of the coronary arteries in three main stages plus a pre-processing where a set of previously traced arterial centerlines [10] serves as input (see Fig. 2).…”

“…Optionally, there were sets of traced centerlines provided that could be used to initialize the segmentation. In our experiments we employed centerlines that were extracted by the combined model-driven and data-driven tracing algorithm described in reference [10].…”