3D Coronary Vessel Tracking in X-Ray Projections

Abstract

Fusing pre-operative CT angiography with per-operative angiographic and fluoroscopic images is considered by physicians as a potentially useful tool for improved guidance. To be adopted, this tool requires the development of tracking methods adapted to the deformations of the arteries caused by the cardiac motion. Here, we propose a 3D/2D temporal tracking of one coronary vessel, based on a spline deformation, using pairings with a controlled 2D stretching or contraction along the paired curves and a preservation of the length of the 3D curve. Experiments were conducted on a database of 10 vessels from 5 distinct patients, with dedicated metrics assessing both the global registration and the local coherency of the position along the vessel. The proposed results demonstrate the efficiency of the proposed method, with an average standard deviation of 2 mm for the localization of landmarks.

3D Coronary Vessel Tree Tracking in X-Ray Projections

“…Fig. 3, (a) and (b), shows the results obtained thanks to the registration of only the vessel of interest (in blue), and the results obtained Evalutation Previous method Actual method sp 26% 31% sp cl 45% 70% lt 6.7mm 4.3mm Table 1: Average results for each evaluation, for the previous method presented in [7] and the actual method presented in this work.…”

“…Selection is based on the available angiographic views and the vessels are selected as the ones that could be the object of an interventional procedure. In the following, we propose to compare the trackings on the vessels of interest made thanks to the registration of the single vessel of interest as in [7], and the registration of the vessel of interest made thanks to the registration of the vessel tree as described in this work. Fig.…”

“…The aim of the proposed tracking method is to track the tree T in all the consecutive phases of the cardiac cycle, which necessitates to deform it. A spline description [7] is a tool suited for a one vessel deformation which can be represented by the optimization of its control points. Nevertheless, the tree structure implies a vessel connectivity preservation.…”

“…The representation of this tree shall support deformation of each 3D curves while keeping coherency between them when they share extremities. As seen in [7], the spline functions support a compact and smooth description of curves which can be continuously deformed by changing the position of the control points. We thus fit an approximating cubic (C 2 continuity) spline curve C k for each edge A k as in [6], using a centripetal method such that, A k ≈ {C k (u) | u ∈ [0, 1]}.…”

“…This strategy raises the question of the validity of the generic model. [7] proposed to focus the attention of the algorithm to the vessel of interest and to track it along the cardiac sequences. It takes profit of a spline representation of the vessel of interest.…”